Workers Comp Mod Lag: Why Safety Wins Show Late
A cleaner jobsite in 2024 may not lower the mod until 2026 or 2027. The delay is built into NCCI's reporting calendar, not your carrier's mood.
Workers comp mod lag happens because NCCI uses older policy years, not the year you just finished. A policy enters the experience period only when its effective date is at least 21 months before the rating effective date (NCCI Item E-1411, July 2025), while first unit reports are valued at 18 months (NCCI, 2026).
The frustrating mod story is usually true: no claims in two years, better training, better supervision, and the Experience Modification Rate (EMR, also called the mod) still goes up.
That isn't proof the safety work failed. Workers comp mod lag is a calendar problem before it's a performance problem. The National Council on Compensation Insurance (NCCI) rules use older, valued policy years, so current safety gains reach the formula later than owners expect.
For a contractor that cleaned up operations in 2024, the first visible mod benefit may land in 2026. The full benefit can take until 2027, depending on renewal dates and which older claim year is still rolling through the experience period.
Why workers comp mod lag exists
NCCI's current Experience Rating Plan language says the rating organization includes payroll and losses when a policy effective date is at least 21 months before the rating effective date and not more than 57 months before it (NCCI Item E-1411, July 2025). In plain English, the most recent policy year usually sits out.
The reporting side creates another delay. NCCI's 2026 unit reporting guide says first unit reports are valued 18 months after policy effective date and due two months after valuation (NCCI Introduction to Unit Reporting, 2026). That gives carriers time to capture audited payroll and claim values, but it means the mod is not a live safety scoreboard.
So a January 1, 2026 mod commonly reflects policy years 2022, 2023, and 2024. It does not reflect 2025 yet. If 2024 was the first clean year after a bad 2023, the bad year may still have more weight in the story than the contractor wants to believe.
Why did my EMR go up after no claims?
Because a clean current year can be invisible while an older claim year is still active. The mod is rolling, not immediate.
Say a contractor had a rough 2023 and a clean 2024. The 2024 policy year is valued in 2025 and may enter the 2026 mod, but the 2023 year is still there too. If the 2023 claims developed upward at valuation, or if the oldest year dropping off was unusually clean, the 2026 mod can rise even though field performance improved.
That feels unfair. It is also exactly how the formula is designed. The mod compares actual losses with expected losses across the active experience period, so the year leaving the window matters almost as much as the year entering it.
In our reviews of Southeast contractor worksheets, this is where the owner and the underwriter often talk past each other. The owner is talking about the last 12 months. The underwriter is seeing a priced record built from older, valued policy years.
Primary loss is the frequency signal
The delay matters more when the older losses are primary losses. NCCI describes the primary portion as the amount of a ratable loss up to the split point, and the excess portion as the amount above it. Primary losses reflect frequency and receive greater weight in the formula (NCCI ABCs of Experience Rating, 2025).
That is why five small injuries can keep a mod elevated after the jobsite has improved. The safety improvement stops new frequency, but the old primary losses remain in the calculation until their policy year ages out.
Severity control matters too. A large claim can still distort a renewal, especially when reserves stay high at valuation. But severity is usually a different conversation with the carrier. It is about whether the claim value reported into the mod still matches reality, not whether the contractor had too many small incidents.
Underwriters do not have to wait for the mod
The mod moves slowly. Underwriting judgment can move sooner.
A contractor that improved safety in 2024 should not wait until 2027 to tell that story. Loss runs, claim-status notes, return-to-work outcomes, supervisor training records, and payroll growth can explain why the worksheet is stale as a forward-looking indicator. Those documents do not rewrite the NCCI calculation by themselves. They help an underwriter separate old formula data from current operating risk.
That distinction matters in a tight market. A 1.18 mod with a clear downward claim pattern prices differently than a 1.18 mod with fresh frequency. The number is the same. The underwriting story is not.
What an audit would check
An audit checks whether the lag is only timing or whether bad data is making the lag more expensive. It compares the active experience period, valuation dates, claim coding, reserve posture, and class exposure against the worksheet NCCI is using. It also distinguishes frequency problems in the primary layer from severity problems in the excess layer, because each points to a different premium impact. The point is not to make the formula instant; it is to make sure the delayed formula is carrying accurate facts.
A delayed mod is tolerable. A delayed mod built on stale or miscoded data is not. Send us your NCCI worksheet and we'll review it for free.