Qualified 401(k) administration is hard because several independent limit-and-test systems read the same census data differently: a deferral can sit under the 402(g) cap while the same participant's annual additions exceed the 415(c) limit, and a plan can pass coverage while failing ADP. This guide teaches you to label every figure with the system it belongs to, then work the arithmetic by hand. Do that before memorizing any table: compute one ADP test and one annual additions worksheet yourself, because applied case work rewards knowing which basket a number belongs to.
What 401(a) qualification requires before the 401(k) feature matters
Qualification under Section 401(a) imposes plan-wide conditions — exclusive benefit, a definite written program, participation, coverage, vesting, and nondiscrimination. A 401(k) arrangement is a CODA layered on that base, adding elective deferral rules.
Treat 401(a) as the foundation. The exclusive-benefit rule and the definite-program requirement govern how the document is written; participation and coverage rules decide who may benefit; vesting and nondiscrimination rules govern how much each group receives. These conditions apply to any qualified plan, including profit-sharing and money-purchase designs that have no deferral feature at all. Coverage under 410(b) tests who is included, while nondiscrimination in amount under 401(a)(4) tests how benefits are distributed between groups.
A CODA is a specific add-on: participants elect to receive cash or defer it into the plan, usually alongside an employer match and optional after-tax contributions. Distinguishing the layers matters in scenarios because a deferral problem is never fixed by a coverage fix — they are separate systems. When you read a case, first confirm the plan satisfies the 401(a) baseline, then analyze the deferral mechanics as their own question. A profit-sharing plan with employer money behaves differently from one with elections.
Why a deferral under the 402(g) cap can still break the 415(c) limit
Section 402(g) caps elective deferrals only. Section 415(c) caps annual additions — employer contributions, after-tax employee contributions, and allocated forfeitures. Elective deferrals do not count toward 415(c), so passing one limit proves nothing about the other.
Name the baskets precisely. Elective pre-tax and Roth deferrals live under 402(g). Annual additions under 415(c) include employer nonelective and matching contributions, employee after-tax contributions, and forfeitures allocated during the year. Both limits are dollar amounts indexed periodically, so verify the current figures before applying them to a real plan. The structural point is that they measure different money and run on different clocks — calendar-year deferrals versus plan-year additions.
Worked scenario, with illustrative caps only: assume a 20,000 deferral cap and a 60,000 additions cap. A participant earning 120,000 defers 18,000, contributes 55,000 after tax, and receives a 4% match (4,800) plus 2,000 of allocated forfeitures. The deferral is compliant, but annual additions equal 61,800 — an excess of 1,800, correctable by returning excess after-tax contributions with income. The common error is stopping at the 402(g) check; the better habit is aggregating every account inflow before signing off.
Hand-calculating the ADP test to expose the plus-two percentage-point band
The actual deferral percentage test compares average deferral rates of highly compensated employees with non-HCEs. When the NHCE rate falls between 2% and 7.99%, HCEs may average the NHCE rate plus two points.
Compute it in three steps: classify participants as HCEs or NHCEs using ownership and an indexed compensation threshold from the census; average each group's deferral rates (real tests are compensation-weighted); then apply the band table — below 2% allows twice the NHCE rate, 2% through 7.99% allows the rate plus two, and 8% or more allows 1.25 times the NHCE rate. ACP works identically but measures match and after-tax contribution percentages instead.
Practical exercise: build a small sheet with NHCE deferral rates of 3%, 5%, and 4%, and HCE rates of 8% and 10%. Expected observations: the NHCE average is 4%, the HCE average is 9%, the limit is 6%, so the plan is three points short. Notice the cap moves point-for-point with the NHCE average inside the middle band, so a two-point QNEC lifts the ceiling two points. Self-check rubric: you can name the correct band, and you can convert each percentage-point gap into dollars from actual compensation.
Pricing a QNEC against HCE distributions when the ADP test fails
A failed ADP test is usually corrected by funding a qualified nonelective contribution to NHCEs, raising the cap, or distributing excess amounts to HCEs. Price both routes before choosing, and document the comparison.
Worked scenario with illustrative figures: NHCEs earn 135,000 in total and average 4%, so they have already deferred 5,400. HCEs earn 380,000 and average 9%, against a 6% limit. Route one: lift the NHCE average to 7%, which requires total NHCE deferrals of 9,450, so a QNEC of 4,050 on top of the 5,400 already deferred. Route two: cut HCE deferrals from 34,200 (9% of 380,000) to 22,800 — 11,400 plus allocable income returned to the executives. The mistake is defaulting to distributions without pricing either route.
Here the QNEC costs less in dollars, stays in the plan, and benefits the workforce, but the ranking is not universal — it depends on group sizes, deferral levels, and whether the document permits the method. Correction rules also prescribe how QNECs must be allocated and require distributions to carry earnings, so follow the plan's adopted correction methodology rather than assuming a shortcut. Professionally, record the two calculations in a short memo; that comparison is the decision trail a reviewer would ask to see.
Safe harbor designs: what testing disappears and which vesting schedule is overridden
Safe harbor plans trade a mandatory employer contribution for exemption from ADP/ACP testing. Basic and enhanced matching contributions must vest immediately, overriding any slower schedule, while the QACA variant adds default enrollment and permits limited vesting.
Three designs dominate: the basic match, the enhanced match, and the 3% nonelective alternative. The vesting override is a recurring trap in scenarios — employer safe harbor match under the basic and enhanced formulas is 100% vested at all times, so a participant terminating mid-schedule takes it all, and elective deferrals are always fully vested under any design. QACA matching may instead be subject to a cliff of up to two years, and QACA brings automatic enrollment defaults with it.
The table below compares the two families on the decisions that recur in case questions. Read each row as a design choice, not a ranking: the safe harbor route buys testing certainty at the price of a nonforfeitable employer cost every year, while the traditional route preserves flexibility at the price of annual testing and possible correction funding. Top-heavy treatment follows the contribution mix, so verify what the plan actually contributes before assuming an exemption applies.
| Design decision | Traditional 401(k) | Safe harbor 401(k) |
|---|---|---|
| ADP/ACP testing | Required annually; correction funding or distributions if it fails | Exempt when safe harbor requirements are met |
| Employer contribution | Discretionary match or nonelective | Mandatory basic or enhanced match, or a 3% nonelective |
| Vesting of employer match | Per the plan's schedule | Immediately vested (basic/enhanced); QACA match may use a two-year cliff |
| Top-heavy minimums | Apply if the plan is top-heavy | Generally exempt when only safe harbor contributions are made |
| Mid-year design changes | More flexible | Restricted to limited amendment situations |
Coverage and top-heavy: two tests that read the same census differently
Coverage under 410(b) asks whether enough non-HCEs benefit, benchmarked at 70% or more. Top-heavy asks whether key employees hold more than 60% of account assets. Different inputs, different fixes.
Run coverage first: divide benefiting nonexcludable NHCEs by benefiting nonexcludable HCEs; 70% or more passes the ratio percentage test. Employees may be excluded only under the document's stated age and service conditions, so flag anyone near the boundary. The average benefits test exists as a fallback but carries additional classification and benefit-ratio requirements, so check the ratio test before reaching for it. Coverage failures are design problems — widen eligibility or entry dates rather than adjusting dollars.
Top-heavy uses a different lens: key employees are more-than-5% owners, more-than-1% owners earning above an indexed compensation threshold, and officers above that same threshold, measured as of the prior year's determination date. If key balances exceed 60% of total balances, non-key participants need an employer contribution of at least the lesser of 3% or the highest key employee's rate. Mini-check: an officer is key only if compensation clears the threshold — an ownership flag in the census alone is not enough.
A five-week worksheet-first sequence with concrete readiness checks
Sequence the weeks around artifacts you produce: a limits worksheet, a hand-calculated ADP/ACP set, a safe harbor comparison, a correction memo, and integrated cases. Confirm credential logistics directly with the issuer.
Week one: the 401(a) baseline plus the 402(g)/415(c) worksheet. Week two: hand-calculate ADP and ACP on three census variants. Week three: safe harbor designs, the vesting override, and top-heavy identification. Week four: correction pricing memos for a failed test. Week five: mixed cases where you label every figure with its system before computing. Compress to days if needed — the artifacts, not the calendar, are the progress markers. One administrative note: eligibility, format, and fees are set by the issuer, so confirm them at asppa.org.
Treat the checks below as learning milestones, not passing predictions — they tell you whether the label-the-basket habit has stuck. If any check still needs notes, return to that section's worksheet instead of rereading passively, because each check maps to one system and one worked example here. Re-run the two scenarios after a week away; if the 4,050 QNEC and the 1,800 excess additions still reproduce without a prompt, the arithmetic is yours.
- State the 402(g)/415(c) difference in one sentence and reproduce the 1,800 excess-additions scenario.
- Compute an ADP test from raw percentages and price both the 4,050 QNEC and the 11,400 distribution.
- Explain why a basic-formula safe harbor match vests immediately while a QACA match may use a two-year cliff.
- Identify key employees from a census line using ownership and the indexed compensation threshold, then compute a top-heavy ratio.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.