The standard framing on SECURE 2.0 §603 in general personal-finance coverage runs like this: starting in 2026, if you earned more than $145,000 (now $150,000 indexed) in wages last year, your catch-up contribution has to go in Roth instead of pre-tax. It's typically presented as a uniform rule that applies to "high earners," with a one-paragraph nod to the eventual loss of the current-year pre-tax deduction on the catch-up dollars. That framing is technically accurate, mostly useful, and quietly incomplete for the hybrid earner who sponsors a solo 401(k).
The incompleteness is the editorial point of this piece. The §603 trigger attaches to FICA wages from the plan sponsor. For an S-corp operator, the W-2 the S-corp pays is FICA wages. For a sole-prop operator, the Schedule C net earnings flowing into the solo 401(k) sponsor are not FICA wages in the §603 sense — they are self-employment earnings, taxed under SECA rather than FICA, and the statute that defines the §603 trigger explicitly excludes them. That asymmetry means the same operator running the same business at the same income level faces a different §603 outcome depending on which entity sponsors the plan.
The frame for this piece: hybrid earners weighing solo 401(k) coordination with a W-2 401k plan have been doing two-layer math for years — FICA arbitrage on the reasonable-comp side, and the §199A QBI posture on the deduction side. SECURE 2.0 §603 adds a third layer. The third layer is small in any single year and meaningful across a fifteen-year compounding window, and it interacts with the first two in ways the standard write-ups don't surface. The piece below maps the mechanics, runs the worked math at two realistic income points, and lands on what the differential changes about the entity-choice decision for an operator age 50 or older with a solo 401(k) in the picture.
The frame: a third layer of entity-choice consequence
Most hybrid-earner solo 401(k) coverage treats entity choice as a two-variable problem. The first variable is the FICA arbitrage available through the S-corp election framework for W-2 earners — the ability to split owner compensation between W-2 wages (subject to payroll tax) and S-corp distributions (not subject to payroll tax), with the savings governed by what defends as reasonable compensation. The second variable is the §199A qualified business income deduction, which phases out for specified service trades or businesses above defined income thresholds and behaves differently depending on whether the business pays W-2 wages and how much. Hybrid earners running profitable side businesses have been triangulating these two layers since the 2017 tax reform.
What changed on January 1, 2026: the Roth catch-up rule under IRC §414(v)(7), added by section 603 of the SECURE 2.0 Act, became operative. For plan participants whose prior-year FICA wages from the plan sponsor exceeded $150,000 (the 2026 indexed threshold per IRS Notice 2025-67), any catch-up contribution to the employer's plan must be designated Roth. The current-year deduction on those catch-up dollars goes away. The dollars themselves still grow tax-free in the Roth bucket and come out tax-free in retirement, but the front-end tax preference disappears.
The piece that the general coverage hasn't picked up: the trigger isn't "high earner." The trigger is FICA wages from the plan sponsor. Self-employment earnings — the kind reported on Schedule SE under §1402 and taxed via SECA rather than via FICA — are not FICA wages in the §603 sense and don't count toward the threshold. A solo 401(k) sponsored by a sole proprietorship runs on those self-employment earnings. A solo 401(k) sponsored by an S-corp runs on the W-2 the S-corp pays the owner-employee. Same operator, same income, different §603 outcome.
That asymmetry creates a planning lever. The lever sits inside an already-multidimensional entity-choice decision, and it doesn't usually flip the decision on its own. But it changes the breakeven where it would have been close, and it accumulates across years of catch-up-eligible service in a way that matters when the comparison runs in dollars rather than rates.
What §603 actually does
SECURE 2.0 section 603 amended IRC §414(v) by adding paragraph (7), which conditions a participant's ability to make pre-tax catch-up contributions on a wage-based threshold. The mechanics, parsed carefully:
Effective date. Plan years beginning after December 31, 2025. The original statute set an effective date of plan years after 2023; an administrative transition period announced in IRS Notice 2023-62 deferred enforcement through 2025. Treasury issued proposed regulations interpreting §414(v)(7) at REG-100669-24 (Jan. 13, 2025); the statutory rule is operative for 2026 plan years and taxpayers may rely on the proposed regulations in good faith pending finalization.
Applicable plans. §414(v)(7) applies to "applicable employer plans" — 401(k) plans, 403(b) plans, and governmental 457(b) plans. SEP IRAs and SIMPLE IRAs aren't applicable employer plans under §414(v) for this purpose; their catch-up architecture sits in different statutory machinery and isn't reshaped by §603. The solo 401(k) is a 401(k); it is an applicable employer plan; §414(v)(7) reaches it.
The wage test. The trigger is whether the participant's wages "as defined in section 3121(a)" from "the employer sponsoring the plan" for the preceding calendar year exceeded the indexed threshold. §3121(a) is the FICA wage definition — the same wage figure that drives Social Security and Medicare tax. For 2026, the threshold is $150,000 (per Notice 2025-67). The threshold is indexed annually and applies to the prior calendar year — meaning the 2026 plan-year determination looks at calendar-year 2025 wages.
The consequence. If the wage test is met, the participant's catch-up contribution must be designated Roth — that is, made on an after-tax basis with tax-free growth and qualified distributions. If the wage test is not met, the participant retains the choice between pre-tax and Roth catch-up. The standard catch-up limits for 2026 (per the IRS retirement-topics page) are $8,000 for participants 50 and older and $11,250 for the SECURE 2.0 "super catch-up" tier covering participants age 60 through 63.
The first three mechanics are uniform. The fourth — the consequence — has a defensive feature worth flagging: if the plan doesn't offer a Roth catch-up option at all, and the participant is wage-test-positive, the plan can either disallow catch-up contributions entirely from that participant or amend to add Roth. Most solo 401(k) documents from the major specialty third-party administrators have already been amended to add Roth catch-up; the operator running a solo 401(k) on a custodian-prototype document should verify the plan document explicitly. The catch-up-Roth-or-nothing posture only bites if the document doesn't accommodate Roth at all.
The FICA wages test, read narrowly
The statutory language in §414(v)(7) is precise about what counts. The condition runs to wages "as defined in section 3121(a)" — not to compensation generally, not to gross income, not to earned income, not to net earnings from self-employment. That precision matters because Title 26 contains multiple wage and compensation definitions that mostly converge but diverge at the edges, and §603 was drafted onto the FICA definition specifically.
§3121(a) defines "wages" as remuneration paid for "employment" — the latter term defined later in the same section as services performed by an employee for an employer. Self-employed individuals don't have an employer in the §3121 sense; they have a trade or business. The remuneration they take from that trade or business isn't §3121 wages. The corresponding payroll-tax regime for self-employment is the Self-Employment Contributions Act, structured in §1402, which defines "net earnings from self-employment" and applies SECA rather than FICA to those earnings.
The split matters for §603 because the trigger anchor — §3121(a) wages — is the FICA side of that bifurcation. Net earnings from self-employment are on the SECA side and don't enter the §603 wage test. An operator whose only relationship to the solo 401(k)'s sponsor is "I am the sole proprietor and the proprietorship is the plan sponsor" doesn't have §3121 wages from that sponsor. The wage test reads to zero. The catch-up Roth mandate does not apply.
An S-corp owner-employee runs on the FICA side. The S-corp pays the owner W-2 wages, withholds and deposits FICA, and reports those wages on Form W-2 in Box 3 as Social Security wages and Box 5 as Medicare wages — both populated under §3121(a). When the §414(v)(7) test asks whether the participant's prior-year §3121(a) wages from the plan sponsor exceeded $150,000, the S-corp owner-employee answers from those W-2 boxes. There is no SECA carve-out available.
One narrow nuance worth surfacing: a hybrid earner who is both a W-2 employee of an unrelated employer and the operator of a sole proprietorship that sponsors a solo 401(k) is tested only on the wages from "the employer sponsoring the plan." The unrelated W-2 doesn't enter the wage test for the solo 401(k). It enters the wage test for that unrelated employer's 401(k), if there is one and if the participant makes catch-up contributions there. Each plan runs its own §414(v)(7) determination against its own sponsor's wages. The math doesn't stack across unrelated employers, which is part of why the solo 401(k) entity choice can carry its own catch-up outcome separate from whatever is happening on the W-2 401(k) side.
The S-corp vs. sole-prop differential
Put the two strands together and the mechanism is straightforward. Two operators, both age 56, both running profitable side businesses, both eligible to make catch-up contributions to their respective solo 401(k) plans for the 2026 plan year. The difference is the entity that sponsors the plan.
Operator A runs through an S-corp. The S-corp pays Operator A $180,000 of W-2 wages for 2025, which is what showed up on Box 3 of the W-2. Box 3 is FICA wages under §3121(a). $180,000 exceeds the $150,000 threshold for 2026 plan-year application. §414(v)(7) is satisfied. Any catch-up contribution Operator A makes to the S-corp's solo 401(k) for 2026 must be designated Roth — $8,000 of after-tax catch-up contribution, no current-year deduction on those dollars, tax-free growth and tax-free qualified withdrawals on the back end.
Operator B runs through a sole proprietorship. The proprietorship's Schedule C net profit for 2025 was $300,000. After the deductible half of SE tax and the contribution itself, Operator B's net earnings from self-employment available to support a §414(v) catch-up are well above any threshold the statute uses for any other purpose. But §414(v)(7) doesn't look at SE earnings; it looks at §3121(a) wages from the plan sponsor. The sole proprietorship doesn't pay Operator B §3121 wages — there is no employer-employee relationship in the FICA sense. The wage test reads to zero. §414(v)(7) is not satisfied. Operator B's $8,000 catch-up contribution to the sole-prop's solo 401(k) can be made pre-tax — current-year deduction on those dollars, tax-deferred growth, ordinary-income tax on withdrawal.
Both operators end up with $8,000 in their retirement account on the catch-up line. The difference is where the tax friction sits. Operator A paid tax on the $8,000 in 2026 (at marginal rate, less whatever the tax-deferred alternative would have cost). Operator B got a deduction on the $8,000 in 2026 (at marginal rate) and will pay tax on the withdrawal in retirement (at marginal rate then). Whether that tradeoff favors current pre-tax or future Roth depends on the bracket comparison — a calculation that has its own literature and that this piece doesn't try to resolve. What this piece does claim: §603 takes the choice away from one operator and leaves it with the other, based solely on entity structure.
The differential also stacks across the catch-up-eligible window. An operator from 50 to 63 has 14 years of standard or super-catch-up eligibility. At 2026 limits — $8,000 for ages 50-59 and 64+, $11,250 for the 60-63 super-catch-up tier under the SECURE 2.0 enhancement — the cumulative catch-up capacity over the full window runs in the neighborhood of $125,000 to $135,000 of contribution dollars before indexing. The S-corp operator above the threshold pays the Roth-versus-pre-tax differential on all of it. The sole-prop operator retains the choice on all of it.
Worked example: S-corp operator at $180,000 W-2
Setting up the math for the S-corp operator at the $180,000 W-2 wage level: 2026 plan year, age 56, federal marginal bracket of 32 percent on the catch-up dollars (the 2026 32-percent bracket for a married-filing-jointly taxpayer covers roughly $400,000 to $510,000 of taxable income — the operator's combined household, including W-2 from an unrelated employer and S-corp pass-through, sits in that band for purposes of this example). State marginal of 5 percent. The catch-up amount: $8,000 (standard, age-50-plus). The relevant comparison is what $8,000 of catch-up contribution actually costs the operator on a present-value basis, given that the Roth designation is now mandatory.
| Component | Pre-tax catch-up (hypothetical, not available) | Roth catch-up (mandated under §603) |
|---|---|---|
| Catch-up contribution amount | $8,000 pre-tax | $8,000 after-tax |
| Current-year federal tax effect (32%) | −$2,560 (deduction) | $0 |
| Current-year state tax effect (5%) | −$400 (deduction) | $0 |
| Out-of-pocket cost in year of contribution | $5,040 | $8,000 |
| Gross account balance after contribution | $8,000 | $8,000 |
| Tax character of future withdrawal | Ordinary income (taxed at then-marginal rate) | Tax-free if qualified |
| Net after-tax cost — current year | $5,040 | $8,000 |
Read the table as a current-year cost comparison: the S-corp operator pays $2,960 more in current-year out-of-pocket cost to fund the same $8,000 catch-up contribution because the deduction is no longer available. That's the §603 cost on a single-year basis, before any growth or back-end-withdrawal modeling.
The full lifetime comparison depends on the retirement bracket assumption. If the operator's retirement bracket is materially lower than the contribution-year bracket, the foregone pre-tax option was the better choice and the §603 mandate carries a real cost. If the retirement bracket is at or above the contribution-year bracket, Roth was the right choice anyway and the §603 mandate doesn't change the answer — it just removed the option to choose wrong. Most operators at the $180,000-W-2 level expect a lower retirement bracket than their working bracket, which is why the loss-of-pre-tax framing is the dominant cost view in the practitioner discussion. The point of the example isn't to settle the Roth-versus-pre-tax debate; it's to surface that §603 takes the lever away.
Stack the cost across the catch-up window — fourteen years of catch-up-eligible service from age 50 to 63, with the super-catch-up tier at $11,250 from 60 to 63 — and the cumulative deduction lost runs to roughly $40,000 to $50,000 of foregone current-year federal-plus-state tax benefit, depending on bracket trajectory and the indexing of both the catch-up limit and the wage threshold. That's the order of magnitude. Whether it's worth caring about depends on the operator's broader retirement-bracket modeling, which is a question with multiple right answers and no single right framework.
Worked example: sole-prop operator at $300,000 SE earnings
The same operator, hypothetically running the same business through a Schedule C sole proprietorship instead of an S-corp, would face a different §603 outcome at any income level — including at SE earnings well above the W-2 wage threshold an S-corp owner-employee would defend. Assume 2025 Schedule C net profit of $300,000, age 56, and the same 32 percent federal plus 5 percent state marginal stack used in the prior example.
| Component | Pre-tax catch-up (available) | Roth catch-up (elective) |
|---|---|---|
| Catch-up contribution amount | $8,000 pre-tax | $8,000 after-tax |
| Current-year federal tax effect (32%) | −$2,560 (deduction) | $0 |
| Current-year state tax effect (5%) | −$400 (deduction) | $0 |
| Out-of-pocket cost in year of contribution | $5,040 | $8,000 |
| Gross account balance after contribution | $8,000 | $8,000 |
| Tax character of future withdrawal | Ordinary income (taxed at then-marginal rate) | Tax-free if qualified |
| Net after-tax cost — current year, operator's chosen path | $5,040 if pre-tax elected; $8,000 if Roth elected — the operator keeps the choice | |
The mechanical takeaway from the sole-prop table is that the optionality remains. The sole-prop operator at $300,000 of SE earnings is, on a SECA-tax basis, paying meaningfully more current-year payroll tax than the S-corp operator at $180,000 of W-2 wages would. That is the well-known FICA arbitrage that drives much of the S-corp election decision in the first place, and it isn't changed by §603. What §603 does change is the catch-up-treatment layer sitting on top of the FICA arbitrage — and on that layer, the sole-prop operator retains a flexibility the S-corp operator does not.
Worth noting on the sole-prop side: the absence of §414(v)(7) constraint does not mean the sole-prop operator should default to pre-tax catch-up. Roth catch-up can still be the right choice — particularly for an operator whose retirement-bracket modeling suggests the future rate will be at or above the current rate, or for an operator already heavy in pre-tax retirement balances and looking to build the tax-free bucket for withdrawal-sequencing flexibility. The point is that the choice exists on the sole-prop side and is foreclosed on the S-corp side once the wage threshold is crossed.
When the S-corp election still wins, and when it tips
The S-corp election decision for a profitable side business is multi-variable, and §603 doesn't dominate the math. The dominant variables in most realistic operator situations remain the FICA arbitrage and the §199A QBI posture. The Roth catch-up exposure is a third layer that adjusts the decision at the margin.
Working through how the layers interact:
The FICA arbitrage layer. An S-corp owner-employee paying themselves $180,000 of W-2 wages on $300,000 of net business profit splits the compensation between the FICA-loaded W-2 leg and the FICA-free distribution leg. The arbitrage saves the operator the 2.9 percent Medicare portion (uncapped, since the wage already exceeds the Social Security wage base) plus, if the operator's combined Medicare wages clear the additional Medicare threshold, the 0.9 percent additional Medicare surtax on the distribution portion. Across the $120,000 of distributions, the annual savings runs in the $3,500 to $4,500 range. Across a fifteen-year operating window, that compounds into real money even on a flat basis. The sole proprietorship pays SECA on the full Schedule C net earnings, which means it pays Social Security (capped) and Medicare (uncapped) on dollars the S-corp would have shielded.
The §199A QBI layer. For 2026, the QBI deduction begins phasing out at $201,775 of taxable income for unmarried filers and $403,500 for married-filing-jointly. The phaseout interacts with W-2 wages paid and unadjusted basis of qualified property — both wage and basis measures favor the S-corp structure for operators above the threshold, because S-corp wages count toward the W-2-wage limitation while sole-prop SE earnings do not. The interaction is fact-specific and depends on whether the trade or business is an SSTB (specified service trade or business, where the deduction phases out entirely above the upper threshold) or non-SSTB (where the W-2-wage and basis tests apply but no SSTB cliff). For non-SSTB hybrid earners above the threshold, the S-corp wage component can materially increase the QBI deduction relative to a sole-prop structure.
The §603 Roth catch-up layer. Above the $150,000 prior-year wage threshold, S-corp owner-employees lose the pre-tax catch-up option. Below the threshold — which means an S-corp electing to pay the owner-employee less than $150,000 — the option remains. Sole proprietorships are outside the test regardless of income.
What that adds up to in the entity-choice math:
For an operator above the §199A threshold running a non-SSTB business, the S-corp election typically still wins on the combined first-and-second-layer math — the FICA arbitrage and the QBI-enhancement combine to dollar-amounts that outweigh the catch-up Roth premium. The §603 layer is a real cost but a small one relative to the dominant variables.
For an operator below the §199A threshold, the QBI math is largely neutral on entity choice — the deduction is available either way without the W-2-wage limitation biting. The decision collapses back to the FICA arbitrage vs. the §603 layer plus the operational cost of running an S-corp (payroll administration, separate return, the formality of reasonable-comp documentation). At lower income levels — say, $80,000 to $130,000 of business net profit — the FICA arbitrage is smaller in absolute terms, and the §603 catch-up exposure becomes a larger fraction of the comparison. The breakeven where the S-corp stops winning is plausibly higher than the conventional rule-of-thumb suggests once §603 is in the model, particularly for operators 50 and older who are actively maxing catch-up contributions.
For an operator at or just above the $150,000 threshold, there is a reasonable-comp question worth thinking through. The reasonable-comp determination has its own audit posture and legal framework, and the §603 trigger is not a permission slip to underpay W-2 wages below what the operator could otherwise defend. But for an operator whose defendable range straddles the $150,000 line, the §603 consequence is a legitimate input — alongside the QBI mechanics and the FICA arbitrage — into where in that defendable range to land. The threshold itself is indexed annually, so the room beneath it grows incrementally each year.
What the piece does try to land: the §603 layer is real, it asymmetrically attaches to S-corp owner-employees, and it deserves a slot in the entity-choice model that most general coverage hasn't built in yet. The variables remain too interactive, the operator situations too varied, and the bracket assumptions too speculative to compress into a single decision rule.
What this piece does not do
Three clarifications about scope, so the piece is read for what it argues and not for what it doesn't.
First, this article does not try to resolve the broader Roth-versus-pre-tax debate for retirement contributions generally. The question of whether to designate the standard $24,500 (2026) employee deferral as Roth or pre-tax is governed by bracket-comparison modeling that has its own substantial literature and that varies sharply by operator. This piece is about a narrow $8,000-to-$11,250 slice of the contribution architecture — the catch-up dollars specifically — and about an asymmetric statutory rule that attaches to that slice based on entity structure. The base-deferral Roth-or-pre-tax question is a different piece.
Second, this article does not address SIMPLE IRA or SEP IRA catch-up architecture. SIMPLE plans have their own catch-up under §414(v) but sit in a different §603 implementation track; SEP IRAs do not have catch-up contributions in the §414(v) sense. The piece focuses on the solo 401(k) — the dominant high-income hybrid-earner retirement vehicle and the one where the entity-choice differential is sharpest.
Third, the analysis assumes the operator's plan document permits Roth catch-up contributions. Most specialty third-party-administrator solo 401(k) documents have been amended to add Roth catch-up; some custodian-prototype documents have not. Verification of the plan document on this point is a pre-condition to relying on any catch-up architecture going forward — a single email to the plan administrator confirms the document language.