Glossary
Adverse impact
Adverse impact is when a hiring practice that looks neutral selects people from a legally protected group, such as a particular race, sex or age group, at a much lower rate than others.
Adverse impact, also called disparate impact, can happen even when nobody means to discriminate. A requirement or test that treats everyone the same on paper can still rule out one group far more often than another. In many places, the law says that difference has to be justified by what the job needs.
This entry explains an idea. It isn’t legal advice. Talk to an employment lawyer about the rules where you hire.
The four-fifths rule
In the US, the Uniform Guidelines on Employee Selection Procedures, used by the EEOC (the federal agency that enforces job discrimination laws) and other agencies, give a rule of thumb. Work out each group’s selection rate: the share of its applicants who pass a step. Then compare each rate with the rate of the group that passes most often. If a group’s rate is less than four-fifths (80%) of the highest rate, that is generally treated as evidence of adverse impact that’s worth looking into.
A worked example
Say a hiring step has these results:
| Group | Applicants | Passed | Selection rate |
|---|---|---|---|
| Group A | 100 | 60 | 60% |
| Group B | 50 | 20 | 40% |
Group B’s rate divided by Group A’s is 40 ÷ 60, or about 0.67. That’s below 0.8, so under the rule of thumb this step deserves a closer look.
Limits of the rule
- Small numbers mislead. With 10 applicants in a group, one person changes the ratio a lot. Statistical tests are often used alongside it.
- It’s a warning sign, not a verdict. Passing the four-fifths check doesn’t prove a step is fair, and failing it doesn’t prove it’s illegal.
- Whether it relates to the job matters. A step with adverse impact may be defensible if it measures something the job truly requires, and no less discriminatory option works as well.
Where it creeps in
- Degree requirements for jobs that don’t need a degree.
- Résumé keyword filters based on the people you hired in the past.
- Scoring accent, fluency or grammar when the job doesn’t depend on them.
- Short time limits that put people with slow internet or disabilities at a disadvantage.
- Scheduling that only works for people with flexible days.
Checking your own process
Where it’s legal, invite applicants to share details such as sex or ethnicity voluntarily, and keep them separate from hiring decisions. Look at pass rates at each step, especially steps a computer decides automatically. Some laws, such as New York City’s rules on automated employment decision tools, require bias audits (independent checks for unfair results) that report these ratios for certain tools. A pass mark is one of the first places to check.