Picture the list before anyone has seen it. A 60-person distribution business — a composite, not a real company — has to cut 12 roles in a reduction in force. The criteria are written down, every person in scope has been scored, and twelve names sit at the bottom of a spreadsheet. The last question before anyone is told is the one that decides whether the list survives scrutiny: does this selection land harder on one group than another?
The standard first test for that question is the four-fifths rule — also called the 80% rule: a group whose selection rate is under 80% of the highest group’s rate is flagged for a closer look. It takes about five minutes to calculate and it is easy to misread, in both directions: people treat a pass as a clean bill of health and a fail as proof of discrimination, and it is neither.
The ratio is only ever as defensible as the selection behind it — the criteria you set before you looked at names, the comparator pools, and the scores. That part is the Reduction in Force (RIF) Layoff Selection Workbook: criteria and weights first, each pool ranked against its own target, and the file you take to counsel. The four-fifths check is run separately, by HR or counsel, on the finished list: it tests the file, it does not replace it.
What is the four-fifths rule?
The four-fifths rule is a rule of thumb, from the federal Uniform Guidelines on Employee Selection Procedures, that compares selection rates between groups: a group selected at less than 80% of the rate of the most-selected group is generally treated as evidence of adverse impact.
The regulation’s own wording, in 29 CFR 1607.4(D) (opens in new tab), is:
“A selection rate for any race, sex, or ethnic group which is less than four-fifths (4/5) (or eighty percent) of the rate for the group with the highest rate will generally be regarded by the Federal enforcement agencies as evidence of adverse impact, while a greater than four-fifths rate will generally not be regarded by Federal enforcement agencies as evidence of adverse impact.”
Three terms carry the whole idea:
- Selection rate — the Guidelines define it as “the proportion of applicants or candidates who are hired, promoted, or otherwise selected” (29 CFR 1607.16 (opens in new tab)).
- Impact ratio — one group’s selection rate divided by the highest group’s selection rate.
- Adverse impact — a substantially different rate of selection that works to the disadvantage of members of a race, sex, or ethnic group (29 CFR 1607.16 (opens in new tab)). The four-fifths rule is the agencies’ shorthand for “substantially different.”
It applies to layoffs, not only to hiring. The Guidelines cover selection procedures used for “any employment decision,” and the list of decisions in 29 CFR 1607.2(B) (opens in new tab) names “retention” alongside hiring and promotion.
The agencies were explicit about its status. In their 1979 Questions and Answers on the Uniform Guidelines (opens in new tab), they call it a rule of thumb that “is not intended as a legal definition, but is a practical means of keeping the attention of the enforcement agencies on serious discrepancies in rates of hiring, promotion and other selection decisions.”
How to calculate the four-fifths rule on a layoff selection
The calculation is five steps, and every one of them is a decision before it is arithmetic.
- Define the pool and the groups. Decide who was in scope for the decision — the comparator pool, meaning the set of people competing for the same remaining roles — and which groups you are comparing within it.
- Compute each group’s selection (retention) rate. For each group, divide the number of people who keep their job by the number of people in that group who were in scope.
- Find the highest rate. Identify the group with the highest retention rate. Every other group is measured against it.
- Divide and compare to 0.80. Divide each group’s rate by the highest rate. A result under 0.80 is the four-fifths flag.
- Repeat for each pool and in both directions. Run the same check inside every comparator pool, and again on termination rates. On termination rates the lower rate is the favorable one, so divide the lowest termination rate by each group’s rate.
In one line: impact ratio = (group’s retention rate) ÷ (highest group’s retention rate), and anything under 0.80 flags.
Step 2 says “keep their job” deliberately, and step 5 exists because of the two traps in the example below: which outcome you count changes the answer, and so does the level you count it at.
A worked example: one layoff, two comparator pools
Here is the composite business from the opening, with illustrative numbers. It has 60 people in scope across two comparator pools, and it is cutting 12 roles. The comparison is between people aged 40 and over and people under 40, because age is a comparison worth running on any layoff (the section on age below explains why).
The whole-company view
| Group | In scope | Laid off | Retained | Retention rate |
|---|---|---|---|---|
| Under 40 | 35 | 5 | 30 | 85.7% |
| 40 and over | 25 | 7 | 18 | 72.0% |
The impact ratio divides the 40-and-over rate by the under-40 rate: 72.0 ÷ 85.7 = 0.84. At 0.84 the selection clears the four-fifths line. Most people would stop here. The next two views show why that would be a mistake.
The same layoff, pool by pool
The 12 cuts were not made across one company-wide list. They were made inside two comparator pools, and the pools are where the decisions actually happened.
| Pool and group | In scope | Laid off | Retained | Retention rate | Impact ratio |
|---|---|---|---|---|---|
| Office, under 40 | 12 | 1 | 11 | 91.7% | Reference (highest rate) |
| Office, 40 and over | 8 | 4 | 4 | 50.0% | 0.55 |
| Warehouse, under 40 | 23 | 4 | 19 | 82.6% | Reference (highest rate) |
| Warehouse, 40 and over | 17 | 3 | 14 | 82.4% | 1.00 |
The warehouse selection is almost perfectly even. The office selection cut half of its 40-and-over staff and one of its twelve younger staff. The whole-company ratio of 0.84 combined a pool with no gap and a pool with a large one — aggregating the pools hid the flag.
That is the first lesson of the example: run the ratio at the level where the decision was made. If each pool was ranked separately, each pool gets its own check, and the company total is a summary, not a test.
The direction trap: retention rate or termination rate?
Now run the same whole-company numbers the other way round, on the rate at which each group was laid off:
- Under 40: 5 of 35 laid off, a 14.3% termination rate.
- 40 and over: 7 of 25 laid off, a 28.0% termination rate.
The 40-and-over group is being laid off at nearly twice the rate of the younger group. Divide the lower termination rate by the higher one and you get 14.3 ÷ 28.0 = 0.51 — far under 0.80 — on exactly the same list that scored 0.84 on retention. With more than two groups, divide the lowest termination rate by each other group’s rate.
Neither number is wrong. They are answers to two different questions, and they do not mirror each other: the gap looks small when both groups mostly keep their jobs, and large when you look only at the minority who lose them. The Guidelines define selection rate on the favorable outcome — people “hired, promoted, or otherwise selected” — and in a layoff the favorable outcome is keeping the job, which is why steps 2 to 4 count retention. But a plaintiff’s expert is free to present the termination-rate version, and “laid off at twice the rate” is the sentence a reader remembers.
So compute both, and ask counsel which one they expect to rely on before you rely on either.
What the four-fifths rule does not tell you
The rule answers one narrow question — are the selection rates far enough apart to deserve a closer look? — and it has four well-documented blind spots.
It is not a verdict on discrimination
The agencies’ Questions and Answers (opens in new tab) say it plainly: the rule “is not intended to resolve the ultimate question of unlawful discrimination.” A ratio of 0.95 does not make a biased decision lawful, and a ratio of 0.70 does not make a sound one unlawful. It decides where attention goes.
Small numbers make it unreliable
The office pool in the example has eight people aged 40 and over. Moving one person changes that group’s retention rate by 12.5 points. The regulation anticipates exactly this: larger gaps “may not constitute adverse impact where the differences are based on small numbers and are not statistically significant” (29 CFR 1607.4(D) (opens in new tab)). The Questions and Answers (opens in new tab) give an example of an employer that selects three men from 20 and one woman from 10 — a ratio of 66.7% that fails the rule, on numbers too small to support a finding.
Treat a flag on a small pool as a question to answer, not a result. The answer is the documented reasoning behind each selection in that pool — which is exactly the part a ratio cannot supply.
Large numbers can flag smaller gaps
The same paragraph runs the other way: smaller differences “may nevertheless constitute adverse impact, where they are significant in both statistical and practical terms.” On a large layoff, a ratio of 0.85 is not automatically safe. That is a statistics question for counsel or a labor economist, not something to settle in a spreadsheet.
Age sits outside the Guidelines — and is still the comparison that matters most
The Uniform Guidelines do not cover age. 29 CFR 1607.2(D) (opens in new tab) states that they “do not apply to responsibilities under the Age Discrimination in Employment Act of 1967.” So the four-fifths rule is not, formally, the age test.
It gets run on age anyway, for two reasons:
- Age disparate-impact claims exist. In Smith v. City of Jackson (2005) (opens in new tab), the Supreme Court held that the ADEA authorizes disparate-impact claims, though with a narrower scope than under Title VII — including a defense for decisions based on “reasonable factors other than age.”
- The age numbers may be handed to the people you lay off. Under the Older Workers Benefit Protection Act provisions at 29 U.S.C. 626(f)(1)(H) (opens in new tab), if you ask employees to waive age claims as part of a termination program offered to a group, the waiver is not considered knowing and voluntary unless you tell them in writing “the job titles and ages of all individuals eligible or selected for the program, and the ages of all individuals in the same job classification or organizational unit who are not eligible or selected for the program.”
Read that second point twice. In a group layoff with release agreements, the ages you would use to run the four-fifths check on age may already be in the hands of everyone asked to sign a release. Assume someone will run the ratio on your list — and run it first.
When a group flags: what to check before anyone is told
A flag means look closer, and the useful place to look is the selection, not the ratio. Work through it in this order:
- Check the pool definitions. Were people who do genuinely interchangeable work put in the same pool, and people who don’t kept apart? A pool drawn around one team can concentrate a group by accident.
- Check each criterion for a proxy. A criterion can be neutral on its face and still track a protected characteristic — salary level, “future potential,” “adaptability,” recent-technology skills and length of service can all move with age. Ask of each one: is it genuinely job-related, and was it measured the same way for everyone?
- Check where the scores came from. Performance ratings are a common layoff criterion and a frequent source of a hidden gap, because they carry every manager’s inconsistency into the ranking. Ratings that were calibrated across managers — the work that goes into a 9 box talent grid or a performance calibration — are more consistent across a pool than one manager’s unreviewed score.
- Check the written reason for each selection in the flagged pool. Every person selected should have a reason tied to the criteria, written down before notification, not reconstructed after a complaint.
- Take it to counsel with the file. The criteria memo, the pool definitions, the full scoring for everyone in scope and the comparison you ran. A qualified employment lawyer can tell you whether the flag needs a statistical test, a different pool, or nothing more.
Do not fix the ratio by swapping names. Choosing who stays by their group, to move a percentage, is the thing the check exists to catch. If the criteria are wrong, fix the criteria and re-run everyone; if they are right, document why.
Keep protected characteristics out of the scoring sheet
The scoring and the impact check are two separate jobs, and they belong in two separate places.
The scoring sheet should only ever hold job-related criteria — skills, performance, certifications, critical knowledge (the bus factor is a legitimate reason to keep someone). Age, sex, race and the other protected characteristics have no place in it, because a column that is visible while you score is a column that can influence the score.
The four-fifths comparison is then run on the finished list, by HR or counsel, joining the selections to the demographic data held elsewhere.
That is how the RIF layoff selection workbook is built, deliberately. It holds no protected characteristics at all. Its Impact Review checks who the selection lands on by department, site and length of service against the overall rate — your own composition check, with groups under five people marked unreadable rather than flagged — and it is explicit that this is not an adverse-impact analysis. What it gives you is the file the analysis runs on: the dated criteria and weights, the comparator pools with their reduction targets, the full scoring, the written reason for every selection, and a dashboard that will not say “ready to notify” until the evidence is on file and the counsel review is marked done — a check on your process and your file, not on your legal position.
Common questions about the four-fifths rule
What is the four-fifths rule in simple terms?
It is a rule of thumb that says a group’s selection rate below 80% of the most-favored group’s rate is generally treated as evidence of adverse impact.
Does passing the four-fifths rule mean a layoff is lawful?
No. Passing it means the selection rates are not far enough apart to trip the rule of thumb. The underlying decisions can still be discriminatory, and a difference that is significant in both statistical and practical terms can count even above 0.80.
Does the four-fifths rule apply to age?
The Uniform Guidelines that define it do not cover the Age Discrimination in Employment Act, but age disparate-impact claims exist under that Act, so age is a comparison worth running on any layoff list.
Should I run it for every group?
Run it for every group your counsel names, in every comparator pool, on both retention and termination rates. The agencies’ Questions and Answers (opens in new tab) compare men with women, and each race or ethnic group with the highest-rate group, rather than every sub-combination — but which characteristics are protected where you operate is a question for counsel, because many states and cities protect more than federal law does.
The ratio is five minutes; the file is the defense
The four-fifths rule is the fastest honest question you can ask of a layoff list, and the example above shows why it has to be asked carefully: at the pool level, on both retention and termination, with the small-numbers caveat in mind, and for age even though the Guidelines leave age out.
But the ratio only ever points. What answers a flag is a selection you can explain — criteria set before names, pools that make sense, scores that were measured the same way for everyone, and a written reason for every person on the list. The Reduction in Force (RIF) Layoff Selection Workbook was built to produce exactly that file, in a spreadsheet you own rather than a consulting engagement you rent — so that when you take it to counsel, the conversation is “here is everything, what have we missed?”
