An organization's hiring process shows a selection rate for one demographic group at 85% of the selection rate for the highest-selected group, comfortably above the commonly cited four-fifths (80%) threshold, and concludes the process is therefore free of adverse impact concerns and legally safe. This conclusion overstates what the four-fifths rule actually establishes — the rule functions as a practical screening threshold that regulators use to help identify situations warranting closer scrutiny, not as a definitive legal test that, once satisfied, guarantees a selection process is free of any adverse impact concern.
What the four-fifths rule actually measures
The rule compares the selection rate for each demographic group against the selection rate for the group with the highest selection rate, flagging a potential concern when any group's selection rate falls below four-fifths (80%) of that highest rate — a straightforward, easily calculable ratio that originated in federal guidelines specifically as a practical rule of thumb to help identify situations that warrant more detailed statistical analysis, not as an independently sufficient legal standard on its own.
Why passing the threshold doesn't guarantee legal safety
A selection process with a selection rate ratio above 80% can still face a legally valid adverse impact challenge if a more rigorous statistical analysis, examining the actual sample sizes and the statistical significance of the observed difference, finds a meaningful and statistically significant disparity that the simple four-fifths ratio alone doesn't fully capture — particularly relevant with smaller sample sizes, where the four-fifths ratio can be a less reliable indicator of genuine statistical significance than it is with larger samples.
Why failing the threshold doesn't automatically mean the process is illegal
A selection process with a ratio below 80% can still be legally defensible if the employer can demonstrate the selection criteria are genuinely job-related and consistent with business necessity, and that no equally effective, less discriminatory alternative was reasonably available — exactly the same job-relatedness and business-necessity standard discussed in the broader adverse impact legal framework, which the four-fifths rule doesn't resolve on its own in either direction. A failed four-fifths screen is a flag prompting further analysis and justification, not an automatic legal conclusion.
Why the rule's simplicity is both its main appeal and its main limitation
The four-fifths rule's straightforward calculation makes it an accessible, practical first screening step that doesn't require sophisticated statistical expertise to apply, which is precisely why it remains widely used and cited — and that same simplicity means it deliberately doesn't incorporate sample size, statistical significance, or the deeper job-relatedness analysis that a complete adverse impact assessment actually requires, which is exactly why treating it as a complete, standalone legal test misunderstands its intended, more limited screening function.
What this means for interpreting adverse impact analysis correctly
- Treat the four-fifths rule as a practical screening threshold prompting further analysis, not as a definitive legal safe harbor in either direction
- Conduct a more rigorous statistical significance analysis alongside the four-fifths ratio, particularly for smaller sample sizes where the simple ratio can be less reliable
- Recognize that failing the four-fifths threshold still allows for a legally defensible process if job-relatedness and business necessity are properly demonstrated
- Be specifically cautious of any organization presenting four-fifths compliance alone as complete legal justification for a selection process
The four-fifths rule earns its widespread use specifically as a practical, accessible screening tool — the mistake lies in treating a screening threshold as though it were a complete legal verdict, when it was always designed to prompt further, more rigorous analysis rather than replace it.