The Statistical Definition of Fairness in the Federal Selection Guidelines and Its Implications for Minority Employment

The definition of fairness in the guidelines is the same as Cleary's regression model of test bias, which results in a lower proportion of new hires that is minority than the proportion of qualified applicants that is minority. This article explains the disparity, reviews alternative definitions that reduce the disadvantage, and discusses legal arguments. Apparently the guidelines definition is widely disregarded in practice–a color‐blind model seems to be the definition generally used by employers. That definition puts minorities at less of a disadvantage than the federal guidelines definition does.