Promotion Test Shows Unequal Results: What Can an Employer Change?

Review the primary official source.

Related case analysis: Ricci: Fear of Disparate Impact Alone Did Not Justify Discarding Promotion Exams.

A promotion test can produce uneven outcomes. Before an employer certifies or discards scores, it needs a documented assessment of job relation, validation, available alternatives, and Title VII obligations.

The question and the records that answer it

Preserve the job analysis, exam design, vendor materials, scoring rules, notice to candidates, score distributions, validation evidence, and decision minutes. Identify whether a less discriminatory alternative would serve the same legitimate needs and when that alternative became known.

Scope of Title VII disparate-impact standards

Ricci v. DeStefano held that fear of disparate-impact liability alone did not justify race-based rejection of completed exams. The Court required a strong basis in evidence of exposure to disparate-impact liability. That rule does not prohibit prospective design reviews or correction of a genuinely invalid selection device.

Separate the score pattern from the explanation for it. A large disparity can trigger further inquiry, but the employer still needs evidence about job-relatedness, business necessity and available alternatives before relying on anticipated disparate-impact liability.

The next steps and deadline check

Keep a timeline of the test announcement, administration, release of scores and proposed change. Mark when each validation concern or alternative first reached the decisionmakers. That sequence helps distinguish a documented problem from a reason supplied only after the dispute began.

Document counsel’s analysis before action, identify the governing selection procedure, and avoid retroactive rationales. Applicants should preserve scores, notices, and stated reasons for any changed certification decision.

Validate Test Design and Job Relation Before Exam Administration

The strongest time to review a selection device is before candidates rely on it. Compare each test component with essential duties, how questions were developed, what training candidates received, and how scores will be weighted. Document validation methods and reasonable alternatives. A distribution of scores may signal a question worth investigating but does not itself prove the device unlawful. Similarly, a vendor’s assurance that a test is fair is not a replacement for the employer’s evidence of job relation.

Establish a Strong Basis in Evidence Before Action on Completed Scores

If an employer considers discarding results, record the actual evidence of potential disparate-impact liability available at that moment. Ricci required a strong basis in evidence before New Haven could intentionally use race to set aside completed firefighter promotion exams. A desire to avoid controversy or litigation was insufficient. On the evidence before it, the Court found that New Haven lacked that strong basis and directed summary judgment for the plaintiffs on the Title VII claim. Ricci, pages 592–593. The rule does not prevent a defensible prospective redesign for a later competition. Treat timing, candidate expectations, and less discriminatory alternatives as distinct issues, not one generalized fairness judgment.

Keep a reviewable file for both sides

Employers should preserve notices, exam versions, scoring, expert validation work, meeting minutes, legal analysis, and alternative proposals. Applicants should retain their own score notice, preparation materials, any change notice, and communications explaining certification decisions. Do not share private test content contrary to lawful restrictions. The record should permit an independent reviewer to see what the employer knew when it acted rather than only a polished account prepared after litigation began.

Related reading: Companion primary-source case analysis; Employment litigation practice; Employment claim deadlines; Retaliation and wage-claim guide.

Keep the exam design file, job analysis, scoring instructions, pass rates, and communications made before the results were known. If leaders consider changing a selection method, record the evidence supporting its job relationship and any workable alternative. A post-results reaction driven only by the racial pattern presents a different record from a documented validity problem identified through a careful review.

What this means: practical action checklist

• Preserve the exam version, job analysis, announced scoring rules and validation work created before scores were known.
• Keep score distributions and the actual evidence supporting a proposed change; a disparity alone does not establish liability.
• Document any equally valid, less discriminatory alternative and when decisionmakers learned about it.
• Retain candidate notices and decision minutes, while protecting confidential exam material and separate filing deadlines.