Which Wage Distributions Are Consistent with Statistical Discrimination?

We derive a non-parametric test for statistical discrimination that can be applied to cross-sectional wage data. Specifically, we show that the wage distributions for two groups (with identical observable characteristics) are consistent with a general reduced-form model of statistical discrimination if, and only if, neither wage distribution first-order stochastically dominates the other. Our model allows us to interpret a rejection of this condition as evidence of bias.