Title VII of the Civil Rights Act of 1964 prohibits employers from intentionally discriminating against employees because of their race or gender. It also prohibits employers from adopting even facially-neutral employment practices that have a “disparate impact” on women or racial minorities. But what exactly is a “disparate impact”? Does this term refer to any disparity that is “statistically significant”—i.e., any disparity that is unlikely to be a result of mere chance? Or does this term refer only to those disparities that are both statistically significant and large enough to matter in “practical” terms?
Neither Congress nor the Supreme Court has definitively answered this question. The text of Title VII does not separately define the term “disparate impact.” And the Supreme Court has generally spoken of “significant” or “substantial” disparities without reducing those terms to any mathematical formula. Unsurprisingly, then, the lower courts have split on this question.
This Note argues that courts should evaluate disparate impact claims using a pure statistical significance standard. The current articulation of the “practical significance” standard is too vague to apply consistently or intelligibly. And courts can use statistical significance and other legal doctrines to protect employers from incurring unfettered liability under Title VII. In the alternative, courts can presume that any statistically significant disparity is actionable unless proven otherwise. This would incorporate some of the benefits of a pure statistical significance standard in cases where statistical and practical significance overlap while retaining some of the protections of the practical significance in cases where these two concepts diverge.