A company sees a 1.18 experience mod on the workers compensation worksheet and treats it like weather. Bad year. Higher number. Nothing to do but wait for renewal and ask the market for a better price.
That misses what the number is doing. The mod is built from the company's own historical payroll and loss experience. It compares that history with what the rating plan expected for businesses of similar size and class, then applies the result to manual premium.
On $1 million of manual premium, the difference between a 1.00 and a 1.18 mod is $180,000 before other rating factors. For a contractor, the number can also affect whether a project owner or general contractor will let the company bid. The insurance calculation can become a margin problem and a revenue problem at the same time.
What the number actually measures.
NCCI describes the purpose of experience rating plainly: use an employer's historical loss experience to predict future loss experience. The calculation uses payroll, classifications, expected losses, and actual losses. It gives primary losses more weight than excess losses because a pattern of claims generally says more about what may happen next than one large outlier.
That makes the mod useful, but delayed. The current number is looking backward through completed policy periods and previously reported claim valuations. A claim can affect future calculations long after the accident date. By the time the debit reaches the invoice, the operating problem that caused it may be years old.
This is why the mod gets mistaken for a fixed penalty. Management sees the result after the most obvious opportunity to prevent the underlying claims has passed.
What cannot be changed.
You cannot shop away a legitimate mod by changing carriers. You cannot erase a claim that happened. You should not pressure an adjuster to understate a reserve or close a file that still needs work. And unless the worksheet contains an error or later receives corrected data, the current factor is usually the current factor.
That is the honest part of the conversation. The number is not a broker promise. It is an output of reported experience under the applicable rating plan.
Where the next number is still moving.
Claim frequency. Repeated smaller claims can weigh heavily because the formula is trying to detect a pattern. Safety work, hiring, training, supervision, and fixing the process behind the repeat loss matter before anyone talks about insurance.
Open claims. Paid amounts and reserves flow into the loss data used for experience rating. That makes return-to-work planning, regular claim reviews, accurate reserves, and a named owner for follow-through operating work, not renewal cleanup.
The data. The worksheet is only as good as the payroll, classifications, ownership history, and claims attached to it. A claim that is not the company's, an ownership change that was never reported, or payroll sitting in the wrong class can distort the result. The fix is evidence and corrected reporting, not a clever explanation.
Timing. Loss data is valued and reported on a schedule. Waiting until the renewal meeting can mean the next calculation has already taken its picture. The review has to happen early enough for legitimate claim and data work to reach the rating process.
The renewal conversation comes too late.
A useful mod review starts with the worksheet and the loss runs:
- Which claims are driving the result?
- Is the problem frequency, severity, or both?
- Which open files need attention?
- Which policy period will age out next?
- Is the payroll and classification data right?
- What operating issue keeps repeating?
Then play the decision forward. Model what happens if nothing changes, and what changes if the repeating loss stops, the open claims are managed, and the data is correct. No promises. Just the math, the timing, and a plan with owners.
The mod is a lagging indicator. It tells you where the business has been, but it also gives management a schedule for what shows up next. Pull the worksheet and the loss runs before the next renewal meeting. By then, some of the most useful work may already be late.