Extended definition
Adverse impact (also called disparate impact) is the legal recognition that discrimination can happen without discriminatory intent. A selection process that doesn’t explicitly consider race, gender, or other protected characteristics can still produce systematically different outcomes across those groups, and in many jurisdictions that constitutes a legal issue regardless of whether the process was designed to discriminate.
The concept emerged most prominently in US case law (notably Griggs v. Duke Power Co. in 1971) and was codified in the EEOC’s Uniform Guidelines in 1978.
Variants of the concept exist in many other jurisdictions — the UK has indirect discrimination under the Equality Act, the EU has similar frameworks under the Employment Equality Directive — though specific tests and remedies vary.
How adverse impact analysis works
A working adverse-impact analysis covers three steps:
- Identify potentially impacted groups — Demographic groups protected under applicable law — race, colour, religion, sex, national origin, age, disability, and others depending on jurisdiction.
- Measure selection rates by group — Calculate the percentage of applicants from each group who progress at each stage of the process — application to screen, screen to interview, interview to offer, offer to hire.
- Apply the relevant statistical test — In the US, the four-fifths rule is the most-cited screening test, alongside statistical significance tests for larger samples. Other jurisdictions use different tests. The result either flags the process for further investigation or doesn’t.
When adverse impact is flagged, the analysis moves to the validation question: can the selection process be justified as job-related and consistent with business necessity? Selection processes that survive validation may continue even with disparate outcomes; processes that don’t survive validation typically need redesign or replacement. The validation standard is meaningful — selection criteria that don’t actually predict job performance are particularly vulnerable when they produce disparate outcomes.
Common sources of adverse impact include cognitive ability tests not validated for the specific role, physical requirements that exceed actual job demands, educational credentials that aren’t truly necessary, and structured assessment criteria that correlate with protected characteristics in ways that don’t reflect job-relevant capability.
Why adverse impact matters
Adverse impact is the legal framework that makes hiring practices accountable for outcomes, not only intent. For employers, it means selection processes need to be defensible — based on job-relevant criteria, validated where they produce disparate outcomes, and documented enough to support that defence if challenged.
The framework drives much of the modern push toward structured assessment, validated selection criteria, and demographic outcome monitoring; these are the practices that survive adverse-impact analysis. For TA functions in jurisdictions with adverse-impact frameworks (most major economies have some version), the analytical capability to monitor for disparate outcomes is increasingly considered baseline operational capability rather than optional sophistication.
Common mistakes and misconceptions about adverse impact
- Confusing adverse impact with intentional discrimination — Adverse impact is unintentional disparate outcomes; intentional discrimination is something different (called disparate treatment in US law). The legal frameworks for each differ; the practical reality is that most modern hiring discrimination is adverse-impact rather than intentional.
- Treating adverse impact as US-only — The US has the most-developed adverse-impact framework but variants exist in many jurisdictions. UK indirect discrimination law, EU Employment Equality Directive, and similar frameworks elsewhere apply analogous concepts. International employers need jurisdiction-specific analysis.
- Assuming structured assessment eliminates adverse impact risk — Structured assessment significantly reduces risk but doesn’t eliminate it. Selection criteria that produce disparate outcomes can still create adverse impact even when applied through structured methods. The criteria themselves matter as much as the application.
- Failing to validate selection criteria — Job-related validation is the legal defence against adverse-impact claims. Selection criteria used without validation are particularly vulnerable when they produce disparate outcomes.
- Monitoring outcomes only at the headline level — Aggregate hire-rate monitoring misses stage-specific adverse impact. A process that has overall fair outcomes can still show adverse impact at specific stages where the cumulative effect cancels out.
Frequently asked questions
What is adverse impact?
Adverse impact is the legal concept that a hiring process — even one that's neutral on its face — may be discriminatory if it produces substantially different selection outcomes across protected demographic groups. It's the framework that addresses unintended discrimination. A selection process that doesn't explicitly consider race, gender, or other protected characteristics can still produce systematically different outcomes across those groups, and in many jurisdictions that constitutes a legal issue regardless of whether the process was designed to discriminate.
What's the difference between adverse impact and disparate treatment?
Adverse impact (disparate impact) is unintentional disparate outcomes from facially neutral processes. Disparate treatment is intentional discrimination — treating people differently based on protected characteristics. Both are legally actionable in many jurisdictions; the legal standards and remedies differ. Most modern hiring discrimination is adverse-impact rather than intentional.
What constitutes adverse impact?
Selection processes that produce substantially different outcomes across protected demographic groups. The specific threshold varies by jurisdiction — in the US, the four-fifths rule (selection rate for any group below 80% of the rate for the highest-selected group) is the most-cited screening test. Other jurisdictions use different tests.
How do you avoid adverse impact?
Through validated, job-related selection criteria, structured assessment methods, ongoing demographic outcome monitoring at each stage of the funnel, and willingness to redesign criteria that produce disparate outcomes without job-relevant justification. Bias mitigation practices generally also reduce adverse-impact risk because they reduce the latitude where bias can shape outcomes.
Does adverse impact apply outside the US?
Yes — variants of the concept exist in many jurisdictions. The UK has indirect discrimination under the Equality Act 2010. The EU has similar frameworks under the Employment Equality Directive. Canada, Australia, and others have analogous concepts. The specific tests and remedies vary; international employers need jurisdiction-specific analysis.