What Is Adverse Impact in Hiring?
Published
Updated
Adverse impact in hiring concerns a selection practice that disproportionately excludes people from a protected group. In the United States, the EEOC’s Uniform Guidelines explain the concept through differences in selection rates affecting race, sex or ethnic groups. EEOC guidance, question 10.
A recruiting process can look neutral on paper while still producing materially different outcomes for different groups. That risk increases when teams add assessments, knockout rules, ranking models, or rushed interview shortcuts without checking the data.
What to Distinguish
Adverse-impact analysis examines selection outcomes, including those produced by apparently neutral procedures. A difference in rates is a signal to investigate with appropriate expertise; a dashboard alone does not establish the cause or settle a legal conclusion. Legal tests and permitted data use depend on jurisdiction.
A Practical Example
Illustrative example: A company uses an assessment to screen sales candidates. After reviewing pass-through rates by group, the team finds that one section is removing qualified candidates who later perform well in interviews. The fix is not to ignore standards. The fix is to test whether the section is job-related, explainable, and necessary.
What to Measure
Use explicit definitions for your reporting. The following are practical measurement choices, not universal benchmarks or claimed product results.
| Measure | Definition | Interpretation |
|---|---|---|
| Stage selection rate | People in a defined group who advance ÷ people in that group considered at the stage × 100. | Use comparable cohorts and show counts alongside percentages; small samples need particular care. |
| Completion rate by group | People completing a selection step ÷ people invited from the same group × 100. | Use demographic data only where lawful and appropriate, with restricted access and clear definitions. |
How to Apply It
Review each selection stage, the criteria applied and any access barriers. Involve the people responsible for employment law, privacy and assessment validity when deciding what data to collect and what a disparity means.
Common Mistakes
- Treating an overall hire rate as proof that every stage is fair.
- Using vendor scores without understanding what they measure.
- Forgetting that interviews and manual screens are also selection procedures.
- Waiting for a complaint before reviewing selection data.
Where Skill Society Fits
SkillSociety helps teams keep the screening criteria visible, keep human review in the loop, and compare candidate evidence against role requirements instead of relying on vague shortcuts.
Book a demo to discuss your screening workflow, evidence requirements and human review points.
Related Reading
Continue with algorithmic bias in hiring.
Sources
- EEOC: Uniform Guidelines questions and answers — U.S. guidance; concept and selection-rate interpretation, reviewed 16 September 2026.
- EEOC: Hiring practices with a negative effect on certain applicants — U.S. employer guidance, reviewed 16 September 2026.
- EEOC Uniform Guidelines Q&A - official guidance on adverse impact and validation of selection procedures.
- EEOC: What is the EEOC's role in AI? - official U.S. guidance that employment discrimination laws apply to AI and other hiring technologies.
- NIST AI Risk Management Framework - risk management and trustworthiness considerations for AI systems.
