What Your EMR Actually Measures, and Why It Follows You for Three Years

Chris Fredette
1 day ago
4 min read

A sub called me last year furious about his experience modification rate. He had one bad claim, a shoulder that turned into surgery, and his EMR went above 1.0. He wanted to know why one injury in four years was costing him bids.

The honest answer surprised him: it probably was not the shoulder. It was the three small claims nobody remembered.

What EMR is comparing you to

Your experience modification rate compares your actual workers compensation claim losses against what an insurer expects for a company your size in your classification. An EMR of 1.0 means you performed exactly as expected. Below 1.0 means better than expected, above means worse.

The part contractors miss is the phrase for your classification. You are not being compared to an office. A roofing contractor is measured against roofing expectations, which already assume a meaningful loss rate. That means a 1.0 in a high-hazard trade represents a genuinely different real-world safety performance than a 1.0 in finish carpentry, and both are average for what they are.

It also means you cannot benchmark your EMR against a contractor in a different trade and learn anything useful.

Frequency hurts more than severity

This is the mechanic that catches people, and it is the answer to the sub with the shoulder.

EMR calculations split each claim into a primary portion and an excess portion. The primary portion, the first slice of every claim, counts at full weight. The excess portion above that threshold gets heavily discounted. The reasoning is that insurers believe frequency predicts future losses better than severity does, because a company with many small claims has a systemic problem while a company with one catastrophic claim may have had one very bad day.

The practical result is counterintuitive. Four claims of $5,000 each will typically hurt your EMR more than one claim of $80,000, even though the second costs the insurer far more. Every one of those four claims contributes its full primary portion. The single large claim contributes one primary portion and then gets discounted on the rest.

So the sub was right that his shoulder was expensive and wrong about what moved his number. The strains and the stitches did that.

Why a bad year follows you for three

EMR uses a three-year experience period, and it excludes the most recent policy year because claims need time to develop. So the number a general contractor sees today reflects roughly years two, three, and four looking backward, not what happened last month.

Two things follow from that, and both matter for planning.

A bad year does not go away when you fix the problem. It stays in the calculation for three full years, which means a contractor who had a rough 2024 and genuinely overhauled their program will still be carrying that number into bids well into 2027.

The reverse is also true and gets forgotten. Improvements do not show up immediately either. If you tightened everything up six months ago, none of that is in your current EMR. Contractors get discouraged when a real turnaround does not move the number, and quit before it does.

What to do when your number is above 1.0

First, get your loss runs from your carrier and actually read them. Most contractors have never looked at the claim-by-claim detail behind the number. That is where you find out whether you have a frequency problem in one crew, one task, or one type of injury. Frequently the pattern is obvious once somebody looks.

Second, check the data for errors. Claims get miscoded, closed claims sometimes carry open reserves that inflate your losses, and reserves on open claims are estimates that can be revisited. An open claim sitting on a stale reserve estimate is money being counted against you that nobody actually paid. Your carrier can review it.

Third, write the explanation before anybody asks. When you submit a prequalification with an EMR above the threshold, attach a short narrative: what drove the claims, what changed, and what your current-year loss runs look like. General contractors are risk managers. A company that can explain its own number reads as lower risk than one that submits a bad number silently and hopes.

Fourth, understand that return-to-work programs move this number more than almost anything else. Claims that stay open and accumulate lost-time costs are what drives losses up. Getting an injured worker back on light duty, when medically appropriate, limits the claim's development.

Common Questions

Does one injury ruin my EMR?

Usually not by itself, especially if it is a single severe claim. Multiple smaller claims in the same period typically do more damage because of how the primary and excess split works.

How fast can I fix a bad EMR?

Three years, realistically, because that is the length of the experience period. Improvements you make now begin affecting the calculation as older years roll off. There is no fast version.

Can I challenge my EMR?

You can challenge the underlying data. Miscoded claims, incorrect payroll classifications, and stale reserves on open claims all get corrected, and any of them can move the number. Ask your carrier or your rating bureau for the detail behind the calculation.

Do GCs really enforce the 1.0 cutoff?

Many do as a first filter, but a substantial number will review above-threshold contractors case by case when there is a credible written explanation attached. Submitting without one is what closes the door.

Key Takeaways

EMR compares you to expected losses for your trade classification, not to zero and not to other trades.
Claim frequency hurts more than severity because the primary portion of every claim counts at full weight while the excess is discounted.
The three-year experience period excludes the most recent year, so both bad years and genuine improvements lag by years.
Read your actual loss runs. Miscoded claims and stale reserves on open claims inflate the number and can be corrected.
A high EMR with a credible written explanation and current loss runs beats a high EMR submitted silently.
Return-to-work programs limit claim development and move the number more than most other interventions.

If your EMR is above where you want it and you are not sure what is driving it, that is a solvable problem and it starts with reading the loss runs properly. Reach out to TriCore and we will work through it with you.