Disparate impact theory

The 1964 Civil Rights Act covered two categories of employment discrimination liability—disparate treatment and disparate impact. Disparate treatment refers to direct and intentional discrimination, unequal treatment, and prejudiced practices against individuals. Disparate impact is often indirect or unintentional and involves the unequal treatment of an entire class of workers who belong to a protected class. Disparate treatment and impact cases involve an employer’s intentional and unintentional discrimination against an individual or a group because of race, color, religion, sex, or national origin. Despite the 1964 legislation, discrimination has not disappeared from American society; citizens are still being prevented from enjoying the full benefits of equal opportunity in employment, hiring, termination, compensation, and other employment conditions. The 1991 Civil Rights Act modified the 1964 act to facilitate discrimination litigation.

Disparate impact theory, instead of focusing on employer intent, examines discrimination caused by specific employer practices. The US Supreme Court case of Griggs v. Duke Power Company (1971) provided the basis for the concept of disparate impact. In this case, the central issue was whether Duke Power, the defendant, used a test not relevant to job performance to screen out Black applicants. The court ruled that the plaintiff must show that they applied for an available job, were rejected based on the employer’s customary practice, and that this practice discriminated against a protected class (that it had a disparate impact on a protected class). If the plaintiff established this, then the defendant had to prove that the practice was required by business necessity. If the defendant met this requirement, the employee could still prevail if they could show that some other less discriminatory employment practice would serve the employer’s business interest.

The 1971 Supreme Court ruling was incorporated into the Civil Rights Act of 1991 so that it could not be weakened by subsequent Court decisions. The 1991 act allows the plaintiff to treat the employer’s entire decision-making process as an entity in cases in which the plaintiff cannot show separately which requirement caused the discrimination.

Under the 1991 act, the employer carries a greater burden of proof in establishing a reason for rejecting a potential employee. The act still allows the use of employment tests but limits how the scores can be used and prohibits race norming. According to this act, if race, color, religion, sex, or national origin were motivating factors in an employment decision, then unlawful discrimination exists, even if lawful factors were also used. The 80 percent rule is often used as a standard in hiring decisions to detect disparate impact in a protected group. If a company’s hiring rate of a particular group of protected individuals occurs at 80 percent or more when compared to selecting people at random, the company is generally thought to be protected from legal action. However, some practitioners have moved away from this rule in the twenty-first century.

AI-powered hiring decisions can promote diversity and inclusion if used properly, but as technology develops, the risk of disparate impact is unknown in some populations and can be risky for some organizations. Recognizing the potential for bias in AI technology and continually assessing and modifying processes is critical in the age of technology.

Bibliography

Anderson, April J. “What Is Disparate-Impact Discrimination?” Congress.gov, 9 July 2025, www.congress.gov/crs-product/IF13057. Accessed 16 Jan. 2026.

Charles, Guy-Uriel E., et al. Race, Reform, and Regulation of the Electoral Process: Recurring Puzzles in American Democracy. Cambridge UP, 2011.

Fredman, Sandra. Discrimination Law. 2nd ed., Oxford UP, 2011.

Guerin, Lisa. Employment Law: The Essential HR Desk Reference. Nolo, 2011.

Mack, Olga. "Promoting AI Fairness: The Application of Disparate Impact Theory." MIT Computational Law Report, 31 Aug. 2023, law.mit.edu/pub/promoting-ai-fairness-disparate-impact-theory/release/1. Accessed 16 Jan. 2026.

Tursunbayeva, Aizhan, et al. “The Ethics of People Analytics: Risks, Opportunities and Recommendations.” Personnel Review, vol. 51, no. 3, 2022, pp. 900–21, doi.org/10.1108/PR-12-2019-0680. Accessed 16 Jan. 2026.