Job Relatedness

Job relatedness is the legal standard under Title VII and the EEOC Uniform Guidelines that requires employer selection procedures to be job-related and consistent with business necessity.

Summarise this post with:

Why job relatedness exists: the Griggs decision

Before 1971, US employers commonly used selection criteria – high school diplomas, written intelligence tests – that were not demonstrably related to the work being performed but that systematically excluded Black candidates from production and supervisory roles. In Griggs v. Duke Power Co. (401 U.S. 424), the Supreme Court held that Title VII prohibits not only intentional discrimination but also facially neutral employment practices that have disparate impact on protected groups, unless the practice is shown to be job-related and consistent with business necessity. The decision created the modern framework: adverse impact triggers a job-relatedness defence.

In 1978, four federal agencies – EEOC, OFCCP, Department of Justice, and Civil Service Commission – jointly adopted the Uniform Guidelines on Employee Selection Procedures (UGESP), codifying how employers can demonstrate job relatedness through formal validation. The Guidelines are not law per se, but they are the framework EEOC and OFCCP use in enforcement actions, and courts give them substantial weight.

The two-step framework: adverse impact, then job relatedness

1. Adverse impact analysis. For each selection procedure, calculate selection rates by protected class. The Uniform Guidelines use the ‘four-fifths rule’ (also called the 80% rule) as a rule of thumb: if the selection rate for any group is less than 80% of the selection rate for the highest-scoring group, this is generally regarded as evidence of adverse impact.

1. Job-relatedness defence. Where adverse impact is found, the employer must demonstrate that the procedure is job-related and consistent with business necessity, typically through formal validation under one of three recognised strategies.

The four-fifths (80%) rule: worked example

100 male applicants and 100 female applicants take a selection test. 50 men pass (selection rate 50%), and 30 women pass (selection rate 30%).

• Lower group rate / Higher group rate = 30% / 50% = 0.60, or 60%.
• 60% is below 80%, indicating adverse impact under the four-fifths rule.
• The employer must now demonstrate that the test is job-related and consistent with business necessity.

Important nuance: the four-fifths rule is a rule of thumb, not a legal threshold. Courts and enforcement agencies also consider statistical significance (Fisher’s exact test, z-tests), sample sizes, and practical significance. Use both.

The three validation strategies

The EEOC Uniform Guidelines recognise three strategies for demonstrating job relatedness. Any one of the three (or a combination) is sufficient when properly conducted.

Strategy	What it demonstrates	Evidence required	Best suited for
Criterion-related validity	Test scores predict job performance	Empirical correlation between scores and performance criteria (concurrent or predictive)	Any selection procedure with adequate sample size
Content validity	Test content represents actual job behaviours	Job analysis + subject-matter-expert linkage of test items to job tasks	Job-knowledge tests, work samples, skill demonstrations
Construct validity	Test measures the underlying psychological characteristic relevant to job performance	Convergent + discriminant evidence; factor analysis; theory-based hypothesis testing	Personality, cognitive ability, abstract trait assessments

Best-practice validation often combines strategies. A cognitive ability test in selection might be defended on construct validity (decades of evidence on GMA) plus criterion-related validity (local data showing it predicts performance in the specific role). Layered evidence is more defensible than single-strategy evidence.

How to make a selection procedure job-related

1. Start with a job analysis. Document the work behaviours, knowledge, skills, abilities, and other characteristics (KSAOs) actually required for the role. Job classification frameworks help structure this. A defensible job analysis observes the work, interviews incumbents and supervisors, and reviews relevant documentation. Without this, no validation strategy succeeds.

1. Map selection criteria to job analysis. Every assessment, interview question, and qualification should link to a documented KSAO. The link is the job-relatedness argument.
2. Choose the validation strategy that fits. Skills assessments suit concrete job tasks (content validity); abstract traits suit construct validity. Abstract traits: construct validity. Where sample size allows and outcomes are measurable: criterion-related validity.
3. Validate before, not after, adverse impact emerges. Pre-validation is dramatically more defensible than post-hoc validation conducted in response to a charge. Many EEOC settlements happen because employers had no validation evidence on hand.
4. Document everything. The Uniform Guidelines specify documentation requirements (Section 15) including validation studies, adverse-impact data, and selection procedure specifications. Documentation IS the defence.
5. Investigate alternative procedures. Even with a valid procedure, the Guidelines expect the employer to consider whether an equally valid alternative exists with less adverse impact. If one does, the employer should use it.
6. Re-validate periodically. Roles change, candidate populations change, validation evidence ages. Periodic re-validation (every 3-5 years for stable roles; sooner for rapidly evolving ones) maintains defensibility.

Common red flags

• Generic minimum qualifications without justification. A ‘bachelor’s degree required’ qualification with no documented link to the actual work is the most-litigated job-relatedness failure. If the work doesn’t require a degree, neither should the posting.
• Off-the-shelf assessments with no validation in the buyer’s context. Vendor-provided validity evidence is helpful but rarely sufficient under the Guidelines’ transportability requirements. Local validation, or transportability evidence specifically applicable to your job and population, is needed.
• Unstructured interviews with adverse impact. Unstructured interviews are difficult to validate because they vary by interviewer and candidate. If they produce adverse impact, defensibility is weak.
• Cut scores set without justification. Selection thresholds (e.g. ‘must score 75%’) require justification linked to job performance. Arbitrary cut scores fail the business-necessity prong of the standard.
• Performance appraisals used in RIFs without validation. Per case law and SHRM guidance, performance appraisals are ‘selection procedures’ under the Uniform Guidelines when used for reductions in force, promotions, or terminations. They are heavily litigated and often poorly validated.

Job relatedness and AI in hiring

AI-based selection tools – resume scorers, video interview analysers, predictive matching – fall squarely within the Uniform Guidelines’ definition of ‘selection procedure’ and must be defended on the same job-relatedness standard. EEOC, OFCCP, and several state attorneys general have issued guidance making this explicit. NYC Local Law 144 requires bias audits for automated employment decision tools; the EU AI Act classifies most hiring AI as ‘high-risk’ subject to documented risk and bias assessments. Vendors who cannot produce validation evidence for their AI tools are not safe partners.

Frequently asked questions