Shift & Shield

S corporation owners wear two hats: employee (your W-2 salary) and owner (your share of profits). Where you draw that line changes your tax picture in two directions at once. The famous direction is payroll tax — salary bears it, profit distributions don't. The direction most owners miss is the qualified business income deduction: above certain income levels, the 20% deduction on your business profits is limited by the wages your business pays — including yours. Set salary too low and you can shrink your own deduction; too high and you're paying payroll tax for nothing. Somewhere in between is a number that's both defensible and efficient, and it isn't an accident — it's a calculation.

Be clear about what this is not: a license to pay yourself almost nothing. The law requires "reasonable compensation" for the work you actually do, the IRS actively looks for S-corp owners underpaying themselves, and an unreasonably low salary is one of the more reliable ways to buy an examination. The strategy here is optimization inside the defensible range, with documentation for why your number is your number.

It fits S corporation owners with meaningful profits — especially once income reaches the levels where the wage-based limits on the 20% deduction start to bite, which is exactly when the salary decision starts doing double duty. It also fits owners who have never revisited a salary set years ago at formation.

It fits poorly if your business is in a field the law treats as a "specified service" and your income is past the phase-out — there the 20% deduction is gone regardless and the calculus changes. And it's not relevant to sole proprietors or most partnerships; this is an S-corp lever.

1. Start from reasonable, not from clever. The anchor is what your role would cost to hire: your duties, hours, experience, and what comparable positions pay. Gather real comparison points — industry pay data, job postings for your role, what you'd pay a replacement — so "we picked a number" becomes "here's why this is the number."
2. Then optimize within the range. With the defensible range established, model where in it your salary best serves the 20% deduction given your profit level. This is arithmetic, not judgment — but it's arithmetic with several moving thresholds, which is why the modeling matters.
3. Change it through payroll, before year-end. The salary is what actually ran through payroll by December 31. A January realization that last year's number was wrong is a next-year fix.
4. Document the decision. Keep the comparison data and a short memo on how the number was set. This file is what does the arguing if anyone ever asks.
5. Revisit annually. Profit moved, thresholds indexed, your role changed — the right number drifts. A year-end check against current profits keeps it honest.

A note on timing the year. Salary is one of the few levers you control right up until the final payroll run. In a bigger-than-usual profit year, the optimization is worth more; in a lighter year, the same salary may overshoot. That's why this chapter's calendar reminder lands in the fourth quarter, not April.

Your salary is an input to more than the QBI math. Retirement plan contributions are calculated from W-2 wages — set salary without looking at your plan and you can accidentally cap what you're allowed to put away. The state tax workaround (PTET) operates on the same profits this decision shapes. And the annual re-check pairs naturally with a year-end projection of any kind. If the retirement chapter is in your gameplan, read these two together — the right salary is one number serving both.

• The race to the bottom. A token salary on real profits is the classic S-corp audit flag, and losing that argument means back payroll taxes, penalties, and interest.
• Optimizing one variable. Minimizing payroll tax while quietly shrinking your 20% deduction — the mistake this chapter exists to prevent.
• Ignoring the retirement interaction. Cutting salary below what your contribution goals require.
• Set once, never revisited. The number that was right at $200K of profit is probably wrong at $800K.
• No file. A fine number with no support is a worse position than it deserves to be.

A retirement plan moves income across time: deduct it now at today's rate, invest it untaxed along the way, pay tax later at whatever rate applies when it comes out. For a business owner the interesting part is design. An owner-only business can run a solo 401(k) where you contribute twice — once as employee, once as employer — stacking to ceilings far above what W-2 employees see. A business with staff can still be designed thoughtfully (matching formulas, safe-harbor structures) so the owner's side of the plan is worth running.

Be clear about the trade: this is a deferral with a door that mostly locks. Money in a retirement plan is hard to reach before retirement age without penalty. The strategy's value depends on your rate now being meaningfully higher than your rate later — usually true for high earners, but it's an assumption, not a law of nature. And the deduction is worth the most in exactly the years you're taxed the most.

It fits business owners with real profits who aren't already maxing a well-designed plan — which, in practice, is most of them: the default plan a payroll company installs is rarely the one an owner would design on purpose. The owner-only business (or owner-plus-spouse) is the sweet spot, where the solo 401(k)'s double-contribution structure does the most work.

It fits poorly when cash is the constraint — locked-up dollars aren't available for the building you want to buy next year — or when this would be your lowest-income year in memory; deferring out of a light year saves tax at your cheapest rate.

1. Check who counts as an employee first. This is the gate. Eligible employees change everything about design — and the rules look across your businesses, so a second company's staff can count against the first. Get the headcount picture right before choosing anything.
2. Pick the design to match the business. Owner-only: a solo 401(k) usually beats the simpler alternatives because of the employee-side contribution. With staff: the design conversation is about what the employer side costs to make the owner side work.
3. Establish the plan by December 31. Documents signed, plan in existence. Some contribution types can be funded after year-end, but a plan that doesn't exist by then mostly can't be fixed retroactively.
4. Run employee-side deferrals through payroll before year-end. Your own employee contribution is a payroll event — it can't be conjured in March. This is where the salary decision and this chapter meet: your W-2 wage is the base your contributions are computed from.
5. Fund the employer side by your filing deadline, extension included — one of the few genuinely retroactive levers in the calendar, and worth knowing about in a bigger year than expected.

A note on timing the year. The employer-side contribution is decided after the year is over, with the actual results in front of you. That makes this one of the best tools for a year that came in hot — and one to throttle in a year that came in light.

Your salary decision is the direct input — contribution limits are computed from W-2 wages, so the two numbers are set together, not sequentially. For owners who want to go past the standard ceilings with after-tax dollars, a plan designed for it opens the mega-backdoor Roth door — a design feature to request up front, not a bolt-on. And in a light-income year the deferral logic flips toward its mirror image: Roth conversions, which belong in the same chapter of your thinking even though they're the opposite move.

• The invisible employees. Discovering after setup that staff in a related business had to be covered — the aggregation rules are the classic trap.
• December 31 as a surprise. No plan by year-end, no plan year.
• Deferrals outside payroll. An owner "contribution" that never ran through payroll isn't an employee deferral.
• The default plan. A provider's off-the-shelf design that quietly caps the owner far below what a deliberate design allows.
• Maxing the deferral in the wrong year. Deducting at your lowest rate in years defeats the entire trade.

The federal deduction for state and local taxes on your personal return is capped. Most states responded with a workaround: let the business elect to pay the state income tax on its own income directly — a pass-through entity tax, or PTET. The business deducts that payment in full as a business expense, uncapped, and the state gives you a credit or exclusion on your personal return so the income isn't taxed twice. Net effect: state tax your family was already going to pay becomes federally deductible again.

Be clear about the boundaries: this only works in states that offer it and for entities that qualify (generally S corporations and partnerships — not sole proprietors, and a single-member LLC reported on Schedule C counts as a sole proprietor for this purpose). The states' versions differ in rate, credit mechanics, deadlines, and who benefits. And it moves when cash leaves — the business pays the state on the state's schedule, which is often earlier than you would have paid personally. The savings are real; so is the calendar.

It fits owners of S corporations and partnerships with meaningful state income tax exposure — most obviously those living in or earning from high-tax states. Multistate owners often benefit most, and need the most care: each state's election is its own decision with its own rules.

It fits poorly if your states impose little or no personal income tax on the relevant income, if your entity type doesn't qualify, or — the trap for multistate filers — when one state's election creates a tax your home state won't credit. A workaround that double-taxes a slice of income is worse than the cap it avoided.

1. Map where the income is taxed before electing anywhere. List the states your business files in and where you pay personal tax. The question for each: does electing here produce a deduction and a credit I can actually use — or a credit my home state won't honor?
2. Check each state's mechanics. Election deadline, whether it binds all owners or only consenting ones, the payment schedule, and — critically in several states — whether payment must be made by December 31 for the deduction to land this year.
3. Make the election on time, in writing, per that state's procedure. Some states want it during the year you're electing for — New York is the sharp example: the election is due by March 15 of that year, not with the return the following spring. California runs a different clock: a prepayment due June 15 of the election year — pay short and the credit gets trimmed on the shortfall. Some states take the election with the return; some require new consent annually. In the hard-deadline states, missing the date means waiting a full year.
4. Fund the payments from the entity on schedule. The deduction belongs to the business, so the business writes the checks. Late or personal-side payments unwind the benefit.
5. Coordinate the personal-return side. The credit or exclusion has to be claimed correctly on your return, in every affected state. This is where a clean setup gets sloppy in April.

A note on timing the year. Because several states require payment inside the calendar year, this strategy has a December deadline dressed up as an April one. In a bigger-than-usual year, confirming the election and prepaying by year-end is often the single most valuable move on the list — and one of the easiest to miss.

PTET operates on the same profit base your salary decision shapes — the two are set against the same numbers and belong in the same year-end conversation. It also pairs with any strategy that changes where income lands: adding a state, an operating agreement change, or a shift in how owners take profits can each change the multistate answer. If you file in more than one state, treat this chapter and your year-end projection as one exercise.

• The missed election window. The most common failure is purely calendar: the state's date passes and the year is gone.
• Paying in January for a December deduction. In payment-deadline states, a week's slip moves the deduction a full year.
• The multistate credit mismatch. Electing in a state whose tax your resident state won't credit — the workaround becomes double taxation.
• Electing because it exists. In low-tax situations the admin outweighs the benefit; this is a strategy you re-justify annually, not a default.
• Forgetting the owners. In consent states, an owner who didn't sign on can put the whole election at risk.