What Is Disparate Impact?
Disparate impact (also called adverse impact) is when an employment practice that looks neutral, like a test, a degree requirement or an AI screener, selects people of a protected group at a substantially lower rate. Under Title VII it is unlawful unless the employer shows the practice is job-related and consistent with business necessity, and no less discriminatory alternative was available.
Who Is This For?
- Employers and HRYou’ll use it to know when a hiring step creates legal risk and how to check it.
How It Is Measured
The usual first check is the four-fifths (80%) rule from the Uniform Guidelines: if a group's selection rate is under four-fifths of the highest group's rate, that is generally regarded as evidence of adverse impact. Courts also use statistical significance tests, especially with large samples. Check yours with the four-fifths rule calculator.
The Three Steps in a Claim
- The plaintiff shows a specific practice causes a disparity.
- The employer shows the practice is job-related and consistent with business necessity, usually through a validation study.
- The plaintiff can still win by showing an equally valid, less discriminatory alternative the employer refused to adopt.
The doctrine comes from Griggs v. Duke Power (1971) and was written into Title VII in 1991.
AI Hiring Tools
Automated tools are judged the same way, and some laws now require audits: NYC Local Law 144, Colorado and the EU AI Act. An independent bias audit measures impact ratios on your own data.
Questions
What is the difference between disparate impact and disparate treatment?
Disparate treatment is intentional: treating someone differently because of a protected trait. Disparate impact needs no intent; the outcome alone is the issue.
Does passing the four-fifths rule mean no risk?
No. It's a rule of thumb. Small samples and significance tests can point the other way.
Run the Four-Fifths Calculator →
Free, no sign-up. Runs in your browser.
