Why does a restaurant that employs 70 servers, cooks, and hosts have such a hard time figuring out which ones it owes health insurance to? The ACA answer is that determining who is a full-time employee when hours change every week requires a method, not a headcount. The IRS built that method into the Section 4980H regulations: the look-back measurement period. Most applicable large employers with hourly workforces either do not know it exists or misapply it, which creates both penalty risk and employee relation problems that a broker can help prevent.
Key Takeaways
- ALEs must offer minimum essential coverage to employees who average 30 or more hours per week (or 130 per month) to avoid the Section 4980H(a) penalty.
- The look-back measurement method allows averaging hours over a 3- to 12-month standard measurement period for variable-hours employees, which the monthly measurement method does not allow.
- Employees who are full-time under the look-back method receive an offer during the entire stability period, regardless of actual hours worked in any individual month of that period.
- New variable-hours employees cannot be classified as full-time or part-time until after their initial measurement period completes; coverage must be offered by the first day of the first month after the initial measurement and administrative periods.
- Failure to offer coverage to even one full-time employee who then receives subsidized Marketplace coverage triggers the 4980H(b) penalty of $4,460 per year per affected employee in 2026.
Why the monthly measurement method fails variable-hours employers
The simpler of the two IRS-approved methods is the monthly measurement method: count actual hours each calendar month and determine whether the employee exceeded 130 hours in that month. If yes, they are full-time for that month and must be offered coverage. The simplicity breaks down for workers whose hours depend on seasonal demand, customer traffic, or employer scheduling decisions. A hotel front desk employee might log 160 hours in July and 80 hours in February. Under the monthly method, the employer owes a coverage offer for July but not February, and the coverage obligation appears and disappears month by month.
The operational problem is real. Offering and terminating coverage monthly is administratively unworkable, creates COBRA trigger issues, and results in employees losing coverage mid-year without a qualifying event to get back in. The look-back method solves this by smoothing hours over a longer observation window and then locking in the classification for a stability period, giving both the employer and the employee predictability.
The three-period structure of the look-back method
The look-back method runs on three consecutive periods: the measurement period, the administrative period, and the stability period. Employers choose the length of each within IRS bounds and must apply the same measurement period consistently to all employees in the same category.
| Period | Length | Purpose | Key note |
|---|---|---|---|
| Standard measurement period | 3 to 12 consecutive months | Measure hours for ongoing variable-hours employees | Usually calendar year or plan year aligned |
| Administrative period | Up to 90 days | Process data, generate offers, enroll employees | Cannot reduce stability period length |
| Standard stability period | At least 6 months; at least as long as measurement period | Coverage obligation window for employees found full-time | Employee is full-time for entire period regardless of current hours |
| Initial measurement period | 3 to 12 months from hire date | Observe new variable-hours hire's actual hours | No coverage offer required during this period |
| Initial stability period | At least 6 months; at least as long as initial measurement | Coverage obligation for new hires found full-time | Must offer by first day of month after admin period ends |
Illustrative structure based on IRS regulations under Section 4980H. Specific period choices must be documented in the employer's ACA compliance policy. Consult a benefits counsel for plan-specific guidance.
How the stability period protection works for employees
Once an employee is determined to be full-time based on the look-back measurement, the employer must offer coverage for the entire stability period even if the employee's hours drop below 30 per week during the stability period. This protection runs both ways: an employee who was part-time during the measurement period cannot be reclassified as full-time mid-stability-period just because their hours increased.
To illustrate: a retail employer uses a 12-month measurement period running October 1 through September 30 and a 12-month stability period running January 1 through December 31 of the following year. An employee averaged 32 hours per week during the October 2025 through September 2026 measurement period. That employee must be offered coverage starting January 1, 2027 and must receive that coverage through December 31, 2027, even if they drop to 20 hours per week starting in March 2027 due to a scheduling change.
The employer benefits as well: they do not owe a penalty if a previously part-time employee surges to 40 hours during the stability period without triggering a new obligation until the next stability period reflects that change. The lock-in works in both directions.
The new hire initial measurement period
New employees pose a specific challenge because there is no prior history to look at. For a new variable-hours hire, the employer starts an initial measurement period on the hire date or on the first day of the following month (employer's choice). During this initial measurement period, the employer is not required to offer coverage. At the end of the initial measurement period, the employer applies the same 30-hour-per-week threshold.
If the employee averaged 30 or more hours during the initial measurement period, the employer must offer coverage by the first day of the first calendar month after the combined initial measurement and administrative periods end. For a 6-month initial measurement period and a 30-day administrative period, a new hire whose start date is January 5 would complete their measurement period around July 5, complete the administrative period around August 5, and must receive a coverage offer starting September 1.
The compliance risk for new hires is timeline management. A large employer with high turnover may have hundreds of employees in initial measurement periods with different start dates, all requiring individualized calculations. Most ACA tracking software handles this with automation; a manual spreadsheet approach fails reliably at scale.
What brokers advising affected employers should cover
Brokers who work with restaurant groups, hotel chains, staffing agencies, or large retail operators are likely talking to applicable large employers who have not documented their measurement period elections or who have never confirmed that their HR and payroll systems are actually tracking hours against the right standard. The first conversation question is simple: "What measurement method are you using, and is it written down?"
When the answer is "I don't know" or "monthly," the next step is helping the employer evaluate whether the look-back method would reduce their 4980H exposure and administrative burden. The 4980H(b) penalty of $4,460 per full-time employee per year for unaffordable or low-value coverage is the more common trigger for hospitality and retail employers. Even if the look-back method is properly applied, the coverage offered still needs to pass the affordability and minimum value tests.
See also: the ACA minimum value standard and when employer plans fail the 60 percent threshold and the Section 4980H employer shared responsibility payment explained for the full penalty structure.
Look-back measurement method FAQ
Common questions from brokers advising applicable large employers with variable-hours workforces on ACA Section 4980H compliance.
What is the look-back measurement method under Section 4980H?
The look-back measurement method is an IRS-approved safe harbor that lets applicable large employers determine whether variable-hours employees are full-time by averaging their hours over a standard measurement period of 3 to 12 consecutive months. If the average reaches 30 hours per week (or 130 per month), the employee is treated as full-time for the subsequent stability period, which must be at least 6 months and at least as long as the measurement period. The employer has an administrative period of up to 90 days between the end of the measurement period and the start of the stability period to process the data and extend coverage offers. This method exists because variable-hours employees cannot be reliably classified month-to-month using the simpler monthly measurement method.
What is the difference between a standard measurement period and an initial measurement period?
The standard measurement period is the recurring look-back window that applies to ongoing employees, typically set on a calendar or plan-year basis. The initial measurement period is a separate window that applies to new variable-hours employees starting with their hire date. The initial measurement period gives the employer time to observe the new hire's actual hours before classifying them as full-time or part-time. During the initial measurement period, the employee is not yet in a stability period and must be treated consistently with their classification. The initial stability period that follows must be at least 6 months for employees who turn out to be full-time, or it can match the standard stability period calendar for the rest of the workforce.
Do we need to offer coverage during the look-back measurement period itself?
No. An employer using the look-back method does not owe a coverage offer to a variable-hours employee during the measurement period itself. The offer is required at the start of the stability period that follows. However, if the employer uses a waiting period (up to 90 days under the ACA), that waiting period cannot extend the start of coverage beyond the first day of the first full calendar month after the measurement and administrative periods end. A common mistake is treating the measurement period as equivalent to a waiting period and then adding an additional 90-day wait on top, which can result in coverage starting later than the law permits.
How does the look-back method interact with the 4980H(a) vs 4980H(b) penalty?
Section 4980H(a) is the broader penalty: it applies when an ALE fails to offer minimum essential coverage to at least 95 percent of its full-time employees. The annual penalty in 2026 is approximately $2,900 per full-time employee minus the first 30. Section 4980H(b) is narrower: it applies when an offer is made to 95 percent of full-time employees, but at least one full-time employee receives a subsidized Marketplace plan. The 4980H(b) penalty is $4,460 per year per affected employee in 2026. An employer that correctly uses the look-back method and makes offers on schedule can avoid 4980H(a) entirely. It still faces 4980H(b) risk if the coverage offered is unaffordable or does not meet minimum value, because unaffordable coverage still lets the employee claim Marketplace APTC.
What counts as a variable-hours employee under the IRS rules?
A variable-hours employee is one whose hours the employer cannot determine at hire whether the employee is reasonably expected to average 30 or more hours per week. This is a facts-and-circumstances test, not a job title. Employees hired for a fixed-hours role at 40 hours per week are not variable-hours employees even if actual hours fluctuate occasionally. Employees in hospitality, retail, healthcare staffing, and gig-adjacent roles where scheduling is demand-driven are the most common examples. An employer who classifies a worker as variable-hours for look-back purposes needs documentation supporting that the initial hours were genuinely uncertain at hire. Retroactively reclassifying a known full-time hire as variable-hours to delay coverage is not a safe harbor use.
