Two ADEA tests decide whether your pension plan favors younger workers over older ones: an equal-or-greater accrued benefit test and a normal retirement age provision. Review both before you rely on the plan.
Key takeaways
- Test one: Compare an older participant's accrued benefit with any similarly situated younger individual who is or could be a participant, as of a given date.
- Test two: A normal retirement age described in ERISA and the tax code is not treated as failing solely for that reason. Check your plan document against both.
- Definitions: "Normal retirement age" and the plan-type terms follow ERISA; "compensation" follows the tax code. Use those definitions in every document.
- Allowed exception: A plan does not fail solely because the subsidized portion of an early retirement benefit is disregarded in accruals.
- Multi-amendment rule: The statute directs Treasury regulations against avoidance through two or more plan amendments. Recalculate accrued benefits before and after each amendment.
Why two tests, not one
You run a small firm. You offer a retirement plan. Your staff ranges from entry level to near retirement. Someone asks why a younger colleague seems to accrue faster. You need an answer that holds up.
The Age Discrimination in Employment Act gives you that answer through two tests built into its pension rules. They are separate tests. They catch different mistakes. You should know both before your next plan amendment goes out. For a broader view, see age discrimination rules for hiring and benefits.
The U.S. Equal Employment Opportunity Commission publishes the statute text. We will stay close to it. Anything the statute leaves open, we flag once and point you to your plan's counsel.
First test: Does your plan ever give an older worker less than a younger worker in the same position? If yes, you fail — unless a listed exception applies.
Second test: Does your plan's normal retirement age follow the ERISA and tax code definitions? If it does, the plan is not treated as failing solely for that reason.
How the ADEA defines its terms
Before the tests make sense, the vocabulary matters. The statute pulls its key terms from other federal law so the pension rules stay consistent across agencies.
The ADEA says, in its own words, that the terms "employee pension benefit plan", "defined benefit plan", "defined contribution plan", and "normal retirement age" carry the meanings ERISA gives them. It does not write its own separate definitions. Your plan document and ERISA's definitions should therefore say the same thing about normal retirement age. When they diverge, that divergence is a compliance question.
Compensation gets the same treatment. The statute points to the Internal Revenue Code's definition. So "compensation" in your plan's pension context means what the tax code says it means, not what a payroll spreadsheet column happens to be named.
This matters for recruiters and managers because job postings, offer letters, and benefit summaries often restate these terms loosely. A casual line like "you're vested at normal retirement" invites a question you may not answer cleanly. Tie every restatement back to your plan document and its governing statute.
Keep the definitions aligned across three documents: your plan document, your summary plan description, and your recruiting material.
Test one: the equal-or-greater accrued benefit test
This is the test you will run most often. It reads simply: your plan passes if, at any date, an older participant's accrued benefit equals or beats that of any similarly situated younger individual who is or could be a participant.
Three phrases carry the weight of that sentence.
Accrued benefit. What has the participant earned so far, measured under your plan's own terms as of a given date? The test does not ask what a benefit might become. It asks what exists on that date.
Similarly situated. The statute's phrase. The comparison is not every worker in the building. It is a worker in a comparable position with a comparable role. Your own job structure and eligibility rules shape who counts as similarly situated.
Is or could be a participant. This catches a trick. You cannot avoid the test by pointing at a younger worker who is not yet in the plan. If they could be in the plan, they still count.
Run the test on a calendar, not on a hunch. Pick dates. For each date, pull the accrued benefit for the older worker and for every comparable younger worker. If the older number is ever smaller, you have a failure on your hands. Then you check the exceptions.
The clean way to run this is a quarterly spreadsheet next to your normal plan review. Columns: date, older participant, accrued benefit, comparable younger participant, accrued benefit, pass or fail. You do not need software. You need the discipline to open the file and look.
The statute directs the Secretary of the Treasury to issue regulations preventing avoidance through the use of two or more plan amendments rather than a single amendment. If you think you can chip away at older workers' accruals in small steps over time, that provision is aimed at you.
Common failure patterns
Three patterns show up in small plans.
The first is an accrual formula that credits service at a lower rate after a certain age. Your older worker has more service. If the formula reduces the rate after a threshold, the older worker's accrued benefit can fall behind.
The second is an early retirement subsidy that is ignored in the accrual calculation. The statute allows this: a plan is not treated as failing the requirements solely because the subsidized portion of any early retirement benefit is disregarded in determining benefit accruals. So an ignored subsidy is a safe harbor. It is not a blanket permission to ignore everything else.
The third is a plan amendment that changes the definition of compensation used for accruals. If the new definition lowers accrual for long-service older workers faster than for newer younger workers, you have a problem. Check whether the amendment moved the comparison on a real date.
Each pattern has the same shape: a change that looks neutral in a summary but moves an accrued benefit down for the older side of the comparison. Recalculate the accrued benefit before and after any amendment. That single habit catches most issues early.
Test two: the normal retirement age test
The second test is narrower. It is not about accruals. It is about the plan's definition of normal retirement age.
The statute says a plan is not treated as failing the requirements solely because it provides a normal retirement age described in ERISA and the Internal Revenue Code. In plain words: if your plan's normal retirement age matches what those other two laws describe, the plan is not treated as failing for that reason alone.
This test exists because normal retirement age does real work in your plan. It anchors benefit commencement decisions. It shapes how your communications read. It frames which comparisons are fair.
If your plan's normal retirement age does not match what ERISA and the tax code describe, this provision gives the plan no protection, and that becomes a second question on top of the accrual one. Align the definition before you touch the accrual formula.
How you should check this: pull your plan document's normal retirement age clause. Pull ERISA's definition. Pull the tax code's definition. Compare all three in one sitting. When they match, this test clears and you return to Test one.
This test does not care about accrued benefit amounts. It cares about the definition. That separation matters for your document review. A clean normal retirement age clause does not save a plan that fails Test one. A failing normal retirement age clause does not excuse an accrual failure. Review both every cycle.
The side-by-side view below is the fastest way to see what each test asks and what it does not.
| Question | Test one: equal or greater accrued benefit | Test two: normal retirement age |
|---|---|---|
| What it measures | Accrued benefit amounts at a given date | Plan's definition of normal retirement age |
| Compared against | Any similarly situated younger individual who is or could be a participant | ERISA and Internal Revenue Code definitions |
| Trigger for a fail | Older participant's accrued benefit falls below a comparable younger worker | None alone; a normal retirement age described in ERISA and the tax code is not treated as failing |
| Main exceptions | Early retirement subsidy disregarded in accruals; plan permitted under a listed subsection | Not treated as failing solely when the definition is described in ERISA and the tax code |
| Who issues related rules | Secretary of Treasury, on multi-amendment avoidance | Definition inherited from ERISA and the tax code |
Built from the statute text linked above. This table summarizes the tests; it does not replace your plan document or counsel's review.
Working both tests into your plan calendar
Schedule two passes at least once a year. That cadence works if the review is real.
The first pass is the normal retirement age check. It is quick. Definitions either match or they do not. Do not skip it because it feels administrative. A mismatched definition is the kind of thing that sits unnoticed for years.
The second pass is the accrual comparison. Pull the numbers. Compare the older participant against every comparable younger worker on the review date. Record the pass or fail. If a fail appears, check the early retirement subsidy exception before you escalate.
After either pass, write down what you checked and what you found. A short memo dated the day of review protects you later. You will not remember the reasoning in eighteen months. The memo will. For a similar record format, see Hiring Decision Debrief: Notes to a Recorded Decision.
When you amend the plan, run the accrual comparison both before and after the amendment on the same date. If the amendment moves the number for the older side of the comparison, stop and involve counsel before the effective date.
Recruiters carry a specific piece of this work. Your offers and your benefits summaries describe the plan in plain language. Those descriptions can drift from the document. Review them whenever the plan changes. A summary that promises faster accrual after a threshold age creates exposure that the plan document may not support.
For day-to-day decisions, keep two habits. First, before you promise a benefit increase to a new hire, check it against the accrued benefit rule.
Where the statute leaves room
The statute sets the tests. It does not hand you a formula for every fact pattern. Two open questions matter for your desk.
First, "similarly situated" depends on your job structure. The statute does not define it line by line. Your plan's eligibility rules and your actual job classifications drive the comparison. Ask your plan counsel how your roles map before you run the numbers.
Read the allowance carefully. It names the subsidized portion of an early retirement benefit. It does not bless every reduction near retirement age. Ask plan counsel where your plan's subsidy falls.
Outside those two, the statute gives you enough to run the tests yourself. Do not wait for a perfect answer before you look at your numbers. An early look is cheaper than a late one.
Checklist before your next plan change
Use this list when a plan amendment, a new plan, or a recruiting promise is in front of you.
- Confirm definitions. Normal retirement age in the plan matches ERISA and the tax code. Compensation matches the tax code's definition.
- Pull both dates. The review date and the amendment date, each with accrued benefits for older and comparable younger participants.
- Check the exception. Disregarding the subsidized portion of an early retirement benefit in accruals is not treated as a failure by the statute when it is the sole reason for a fail. Note the reliance in writing.
- Watch the stacking rule. Judge a series of amendments by their combined effect, not one change at a time.
- Review recruiting language. Your offers and summaries should mirror the plan document, not drift from it.
- Date your memo. Record what you checked, what you found, and who reviewed it.
Run the checklist every time the plan text moves. It takes an afternoon. It prevents a year of quiet exposure.
Your next step today: pull your plan document's normal retirement age clause and compare it with the ERISA and tax code definitions.
FAQ
Does passing the normal retirement age test mean my plan is safe? No. That provision covers only the definition question. Your plan can match ERISA and the tax code on normal retirement age and still fail the accrued benefit test on the same day. Run both checks each review cycle.
What if my workforce has no clearly comparable younger worker? The test still speaks in terms of any similarly situated younger individual who is or could be a participant. That includes someone eligible but not yet enrolled. Map your eligibility rules first, then compare.
Can we reduce accruals gradually instead of in one amendment? The statute anticipates exactly that move, so plan around the aggregate effect, not the individual change.
We just hired a recruiter who rewrites benefits copy. Does that create risk? If the copy describes accruals or normal retirement age differently from the plan document, yes. Review every recruiting description after a plan change. The plan document governs; the summary should follow it.
Where do we go for the definitions the statute references? Start with ERISA's definitions and the tax code's definition of compensation, both cited in the statute text linked above. For your specific plan, your plan counsel reads your document against those definitions. That is the question to ask them.