ACA Employer Shared Responsibility Penalty Calculator

What is the difference between the 4980H(a) and 4980H(b) penalties?

One is a sledgehammer and one is a tack hammer, and which you get depends on a single test.

The 4980H(a) penalty applies when you fail to offer minimum essential coverage to at least 95% of your full-time employees and their dependents, and at least one full-time employee receives a premium tax credit through the Exchange. It's assessed on every full-time employee minus the first 30 — not just the one who went to the Exchange.

The 4980H(b) penalty applies when you clear the 95% test but the coverage you offered was unaffordable or lacked minimum value. It's assessed only on the specific employees who actually received a premium tax credit.

The calculator's defaults show the gap. With 220 full-time employees and 202 offered coverage, you're at 91.82% — under the threshold, so the A penalty applies to 190 employees at $278.33 a month: $634,600 a year. If you'd cleared 95%, the same 14 employees on subsidised coverage would produce a B penalty of $70,140.

Run the offer count up one employee at a time and the cliff is brutal. At 208 of 220 you're at 94.55% and owe $634,600. At 209 you're at exactly 95.00% and owe $70,140. One offer of coverage is worth $564,460.

That's the whole point of the sledgehammer nickname. Missing the offer test by a rounding error prices your entire workforce.

How do the three ACA affordability safe harbors work?

Affordability is legally measured against household income, which no employer can see. So the IRS provides three safe harbors using data you do have. You only need to pass one, and you may use different harbors for different employee classes as long as you apply them consistently and document the methodology.

W-2 safe harbor. The employee's required contribution for the lowest-cost self-only minimum value coverage must not exceed the affordability threshold times Box 1 wages. At the defaults, $41,000 × 9.96% ÷ 12 = $340.30 a month. Best for salaried staff with steady hours; awkward if someone takes unpaid leave and their Box 1 comes in lower than expected.

Rate of pay safe harbor. For hourly employees, take the lower of their rate on the first day of the coverage period or their lowest rate that month, multiply by 130 hours regardless of hours actually worked, then apply the threshold. At $17.50 an hour: $17.50 × 130 × 9.96% = $226.59. For salaried employees use monthly salary instead — but if that salary is reduced during the year, including through reduced hours, this harbor is unavailable.

Federal poverty line safe harbor. The threshold times the FPL for a single-person household, divided by 12. Using the $15,650 figure: $129.90 a month. It ignores employee earnings entirely, which makes it the simplest to administer and the only one that sets a single uniform contribution across the whole workforce.

At the defaults' $245 contribution you pass W-2 only. Drop to $129.89 and you'd pass all three. Rise to $360 and you pass none.

Why does the affordability threshold matter, and what is it right now?

It scales every safe harbor maximum, so it decides pass or fail at the margin — and this is one to check rather than trust, because the published sources genuinely disagree.

Most current guidance puts the 2026 threshold at 9.96%, citing Rev. Proc. 2025-25, and that figure reconciles cleanly with the widely quoted $129.90 FPL safe harbor maximum. But an April 2026 practitioner article states 9.86%, citing the prior year's revenue procedure — and contradicts its own headline in the body text. A major payroll vendor's glossary quotes 10.22%. These cannot all be right.

The spread is small per employee and decisive at the boundary. Run the three figures through the calculator and the FPL harbor moves from $128.59 to $129.90 to $133.29 — about $1.30 a month between the two most-cited values. Across 220 employees over a year that's roughly $3,400, which is trivial. What isn't trivial is an employee sitting at $130 a month who is affordable under one reading and not the other.

This is why the calculator takes the threshold, both penalty amounts, the 30-employee reduction, the offer percentage and the poverty line as inputs rather than constants. Every one of them is indexed or subject to change, and a hardcoded figure is a calculator that's quietly wrong the moment the IRS publishes a revenue procedure.

Pull your figures from the current revenue procedure for your plan year, enter them, and the arithmetic follows.

When is the B penalty capped, and why does it matter more for small employers?

The total 4980H(b) penalty for an employer is limited to what the 4980H(a) penalty would have been. You can never pay more for offering bad coverage than you would have paid for offering none.

At the defaults the cap is nowhere near binding — uncapped B is $70,140 against an A ceiling of $634,600. But the cap is a small-employer phenomenon, because the A ceiling shrinks fast when the 30-employee reduction eats a large share of your workforce.

Work the arithmetic.

The 40-employee case has an A ceiling based on just 10 employees, so uncapped B exposure runs nearly six times the ceiling and the cap saves $156,980.

The practical reading is counterintuitive. For an employer just over the ALE line with heavy subsidy take-up, offering unaffordable coverage and offering none can cost exactly the same. If you're in that position, the affordability fix has less financial value than it appears — the cap was already protecting you — and the real question becomes whether coverage that nobody can afford is worth the administrative cost of running at all.

The calculator flags whether the cap is binding so you know which conversation you're in.

Is it cheaper to pay the penalty or fix the offer rate?

Almost always cheaper to fix it, and usually not close — which is worth knowing because the penalty is non-deductible while the premium is a deductible business expense.

The calculator prices both sides. At the defaults you need 7 more employees offered coverage to clear 95%. At $620 a month of employer cost each, that's $52,080 a year. Clearing the test drops you from the A penalty to the B penalty, avoiding $564,460.

Net benefit: $512,380. That's $10.84 of penalty avoided for every dollar of premium spent.

Two honest qualifications.

1. This compares only the penalty against the incremental premium — it ignores the administrative cost of extending eligibility, and it assumes the newly covered employees don't change your affordability position, which they might if their wages are low enough to fail the harbors you're relying on.
2. The return collapses once you're already over 95%: the calculator will show zero employees needed and zero penalty avoided, because there's nothing left to fix on the offer side.

Also note what the calculator does not do. It works from monthly averages, while §4980H liability is determined month by month and the IRS assesses it that way in the Letter 226-J summary table. If your workforce fluctuates seasonally, run the months that differ separately rather than trusting a full-year average — an employer compliant for nine months and non-compliant for three has a very different bill from the annual figures.