Payroll Compliance Tax Planning

The 2027 ACA Affordability Threshold: What It Changes for Your Payroll Math

Published August 28, 2026 by Invisible LLC Team · 8 min read

First, the gate, so you can stop reading if this isn't your problem. Everything below applies to applicable large employers — businesses that averaged at least 50 full-time employees, including full-time equivalents, during the preceding calendar year. If you're under that line, none of this touches you, and you can go do something more useful with the next eight minutes.

Still here? Then one number just moved, and it quietly decides whether the plan you already offer still counts.

The short version

The IRS set the ACA affordability percentage at 10.22% for plan years beginning in 2027 — up from 9.96% for 2026 (Rev. Proc. 2026-26). Separately, the penalty amounts behind the employer mandate rose to $3,780 and $5,670 for 2027 (Rev. Proc. 2026-22).

A higher percentage means you can charge employees more for self-only coverage and still be affordable. That is genuinely good news — but only if you actually re-run the math. Employers who set a contribution once and left it alone are the ones who get surprised, in either direction.

What the percentage actually governs

The affordability test asks one narrow question: is the amount an employee has to pay for the cheapest self-only coverage you offer, that meets minimum value, more than the applicable percentage of their income?

Three things people routinely get wrong here:

It's self-only coverage, not family coverage. The test looks at what the employee alone would pay to cover themselves. What family coverage costs is not part of this calculation.

It's your lowest-cost plan that provides minimum value. If you offer three tiers, the test runs against the cheapest qualifying one. A rich plan priced above the threshold doesn't create exposure as long as an affordable, minimum-value option exists.

It's the employee's required contribution, not the total premium. Your share doesn't enter into it. Only what comes out of their check.

So when the percentage rises from 9.96% to 10.22%, the ceiling on what you can charge goes up. On a $50,000 salary, that's the difference between roughly $415 and roughly $426 a month — not dramatic, but real, and it compounds across a workforce.

The direction matters more than the size. A rising affordability percentage gives employers room; it does not create an obligation to use it. Plenty of employers will look at this and decide to hold contributions flat as a retention move. That's a legitimate call — just make it on purpose.

The penalties behind it went up too

Two separate assessments sit under the employer mandate, and conflating them is the most common mistake we see.

Section 4980H(a) — the "didn't offer" assessment. This triggers if 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. The calculation is brutal by design: it's based on your entire full-time headcount, minus up to 30 employees, multiplied by 1/12 of the adjusted amount, computed month by month. For 2027 that adjusted amount is $3,780.

Section 4980H(b) — the "offered, but it didn't work" assessment. This triggers if you did offer coverage to at least 95%, but at least one full-time employee got a premium tax credit because the coverage was unaffordable, didn't provide minimum value, or that employee wasn't offered coverage. Here the assessment is based solely on the number of employees who actually received the credit — not your whole headcount — at 1/12 of the adjusted amount each. For 2027 that's $5,670. This assessment is capped at what your (a) liability would have been (IRS).

The practical read: (a) is the catastrophic one and (b) is the expensive one. A 200-person employer who misses the 95% offer threshold is looking at an assessment computed on 170 employees. The same employer who offers coverage to everyone but prices it a few dollars over the line is exposed only for the employees who go to the exchange and qualify. Getting the offer right is the first-order problem. Getting the price right is the second.

One more thing worth stating plainly: these are computed monthly. A pricing mistake that exists for three months of the plan year is a three-month exposure, not an annual one. Catching it midyear is worth real money.

You don't know household income — so use a safe harbor

The statute measures affordability against household income. You have no way to know that, and the IRS knows you have no way to know that. So there are three safe harbors you may use instead (IRS):

Form W-2 safe harbor. Measured against the wages you report in Box 1 of the employee's W-2. Straightforward, and it uses a number you already produce — but Box 1 is net of pre-tax deferrals, so a heavy 401(k) contributor's Box 1 can be meaningfully lower than their gross pay. The catch is timing: you don't know final Box 1 wages until the year is over, which makes this a poor tool for setting contributions in advance.

Rate of pay safe harbor. Based on the employee's rate of pay at the beginning of the coverage period. For hourly employees, you may adjust if the rate of pay decreases during the year — but not if it increases. This is the one most employers should default to for planning, precisely because it's knowable in advance. You can set your contribution before open enrollment and know where you stand.

Federal poverty line safe harbor. Treats coverage as affordable if the required contribution doesn't exceed the applicable percentage of the federal poverty line for a single individual, divided by 12. This is the simplest and the most conservative: one contribution amount that works for every employee regardless of pay. It's also the most expensive, because you're pricing to your lowest-paid worker. Employers with a wide pay band often find it costs more than it's worth; employers with a compressed, lower-wage workforce often find it's the cleanest option available.

You can apply different safe harbors to different reasonable categories of employees, as long as you do it consistently within each category.

What to actually do before open enrollment

This is a payroll configuration exercise more than a benefits exercise, and it wants to happen before enrollment materials go out, not after.

Confirm ALE status for the coming year. It's based on the average over the preceding calendar year, counting full-time equivalents and aggregating across related entities under common control. Businesses that grew through 2026 can cross the line without anyone noticing.
Identify your lowest-cost, minimum-value, self-only plan. That specific plan's employee contribution is the only one the test cares about.
Pick your safe harbor per employee category and write down why. If you're setting contributions before the year starts, rate of pay is usually the workable choice.
Run the math at the bottom of each pay band, not the average. Affordability is tested employee by employee. Your median employee passing tells you nothing about your lowest-paid one.
Check the 95% offer threshold separately. Affordability is the (b) problem. The offer percentage is the (a) problem, and it's the bigger one. Look hard at variable-hour employees, new hires in their measurement period, and anyone whose classification changed midyear.
Set the deduction correctly in payroll on day one of the plan year. Nearly every affordability failure we've seen traces back to a payroll deduction that didn't match what the benefits team intended — not to a bad plan design.
Reconcile against your 1095-C coding. The affordability safe harbor you chose has to match what you report.

If you're doing this work, it's worth checking the rest of your employer-side payroll position at the same time. The paid family and medical leave credit became permanent and expanded, and if you offer leave above your state's floor you may already qualify without knowing it. And the overtime reporting rules changed for 2026 W-2s, which is a separate payroll-configuration deadline sitting on the same calendar.

Where this usually breaks

In our experience the failure is almost never a decision — it's a handoff.

The benefits broker models a contribution that clears the threshold. Somebody enters a slightly different number into payroll. The plan year starts. Twelve months later a 1095-C goes out coded as affordable, an employee who was actually charged over the line gets a premium tax credit on the exchange, and a letter arrives.

Nobody made a bad call. The number just didn't survive the trip from the spreadsheet to the deduction code. That's a bookkeeping and payroll-operations problem, and it's fixable with one reconciliation step at the start of the plan year.

Getting it right

We handle payroll and benefits administration for employers running mixed hourly and salaried workforces — including the part where the contribution that was modeled is actually the contribution that gets withheld. That's our payroll and benefits service, and when the question crosses into filings and assessments, our tax and compliance work picks it up.

If you're at or near the 50-employee line and you're not certain where you stand for 2027, request a quote and we'll walk through the math with you before open enrollment closes the window.

This article is general information, not legal or tax advice. Figures are the published 2027 indexed amounts; confirm your specific situation with your benefits counsel or advisor.