Statistical Discrimination in Labor Markets: An Experimental Analysis

Statistical discrimination occurs when distinctions between demographic groups are made on the basis of real or imagined statistical distinctions between the groups. While such discrimination is legal in some cases (e.g., insurance markets), it is illegal and/or controversial in others (e.g., racial profiling and gender-based labor market discrimination). First moment statistical discrimination occurs when, for example, female workers are offered lower wages because females are perceived to be less productive, on average, than male workers. Second moment discrimination occurs when risk averse employers offer female workers lower wages based not on lower average productivity but on a higher variance in their productivity. Empirical work on statistical discrimination is hampered by the difficulty of obtaining suitable data from naturally-occurring labor markets. This article reports results from controlled laboratory experiments designed to study second moment statistical discrimination in a simulated labor marker setting. Since decision-makers may not view risk in the same way as economists or statisticians (i.e., risk=variance of distribution), we also examine two possible alternative measures of risk: the support of the distribution, and the probability of earning less than the expected (maximum) profits for the employer. Our results indicate that individuals do respond to these alternative measures of risk, and employers made statistically discriminatory wage offers consistent with loss-aversion in our full sample (though differences between male and female employers can be noted). If one can transfer these results outside of the laboratory, they indicate that labor market discrimination based only on first moment discrimination is biased downward. The public policy implication is that efforts and legislation aimed at reducing discrimination of various sorts face an additional challenge in trying to identify and limit relatively hidden, but significant, forms of statistical discrimination.