⚖️ Four-Fifths (80%) Rule

Four-Fifths (80%) Rule#

An adverse-impact guideline flagging a group selection rate below 80% of the top group’s.

Important

✨ AI-generated content. This page was written with the assistance of an AI language model and is provided as a learning aid. Despite careful review, it may still contain mistakes, omissions, or out-of-date information. Whether you are new to the topic, a team lead, or a senior practitioner, treat it as a starting point rather than an authoritative reference: read it critically and independently verify anything you act on (code, commands, figures, and factual claims) against official documentation and primary sources before relying on it.

What it is#

The four-fifths rule — also called the 80% rule — is the primary rule-of-thumb the U.S. EEOC uses to flag adverse (disparate) impact in a selection process. It comes from the Uniform Guidelines on Employee Selection Procedures (UGESP, 1978) and operates under Title VII of the Civil Rights Act. The rule: the selection rate for any protected group (by race, sex or ethnicity) that is less than four-fifths (80%) of the rate of the highest-selected group is generally treated as evidence of adverse impact.

How the ratio is computed#

Four steps. Compute each group’s selection rate — the number selected divided by the number of applicants. Identify the group with the highest rate; that becomes the baseline. Divide every other group’s rate by that baseline to get an impact ratio. If any ratio falls below 0.80, the procedure is flagged for that group.

A worked example#

An employer hires for a warehouse role: of 100 white applicants, 40 are hired (a 40% rate); of 80 Black applicants, 24 are hired (a 30% rate). The impact ratio is 30 / 40 = 0.75. Because 75% is below 80%, adverse impact is indicated against Black applicants and the procedure would be flagged for review. Note the test compares rates, not head-counts — a smaller group can still clear the bar if its rate is high enough.

A threshold, not a verdict#

Falling below 80% does not prove discrimination — it triggers investigation. Because the raw ratio ignores whether the gap is beyond chance, practitioners pair it with significance tests (Fisher’s exact, chi-square, or a standard-deviation analysis). Once adverse impact is indicated, the employer must show the procedure is job-related and consistent with business necessity, offer a less-discriminatory alternative, or modify it. Courts treat the rule as guidance, not a definitive legal standard.

Why it matters for ML#

The rule governs automated selection exactly as it governs human decisions: an AI resume screener or scoring model that selects one protected group at under 80% of the top group’s rate carries indicated adverse impact — even with no discriminatory intent. This is why the 80% ratio is the legal touchstone behind fairness criteria like demographic parity, which compares positive-outcome rates across groups.

Theme: Fairness & Calibration · All terminology

Hint

Mind map — connected ideas

Demographic Parity (Statistical Parity) · Equal Opportunity (Fairness) · Equalized Odds (Fairness) · Predictive Parity (Calibration) · Discriminatory Power · Statistical Significance

Hint

More in Fairness & Calibration

Demographic Parity (Statistical Parity) · Equal Opportunity (Fairness) · Equalized Odds (Fairness) · Fairness Guardrails · Fairness parity · Predictive Parity (Calibration) · Selection Rate