Below 1.00 is good — it cuts your workers' comp premium and keeps you eligible to bid. Here is exactly what the EMR measures, how it is calculated, why 0.90 is the unofficial benchmark, and ten proven ways to lower yours.
If you bid construction, manufacturing or industrial work, your Experience Modification Rate (EMR) — also written EMod, X-Mod or "the mod" — is one number that quietly decides two things: how much you pay for workers' compensation insurance, and whether you are even allowed onto the jobsite. A good EMR earns you a premium discount and clears prequalification gates. A bad one does the opposite. This guide explains what the number means, how the rating bureaus calculate it, and what counts as "good."
The Experience Modification Rate is a multiplier applied to your workers' compensation premium that reflects your company's actual injury and claims history compared with other businesses of the same size and type. It is calculated by a rating bureau — in most states the National Council on Compensation Insurance (NCCI), and in independent bureau states such as California, Pennsylvania, New York, New Jersey, Michigan and others by the state's own bureau.
The mechanics are simple even if the math behind them is not. The insurer takes the "manual premium" your class codes and payroll would normally generate, then multiplies it by your EMR:
| EMR | Meaning | Effect on premium |
|---|---|---|
| 1.00 | Industry average loss experience | No change — you pay the manual premium |
| Below 1.00 (e.g. 0.85) | Better than average — a credit | Discount (0.85 = 15% off the manual premium) |
| Above 1.00 (e.g. 1.25) | Worse than average — a surcharge | Increase (1.25 = 25% more than the manual premium) |
So an EMR of 1.00 is the dividing line. Everyone in your industry is benchmarked against it. The whole point of the system is to reward employers whose safety record is better than their peers' and to charge more to those whose record is worse — turning safety performance into a direct, repeating line item on the books.
The short answer: anything under 1.00 is good, because you are paying less than the average competitor for the same coverage. But "good" has tiers, and the practical bar is often set by the people hiring you, not the insurer:
That last point is the one people underestimate. On commercial, industrial and public projects, many general contractors require subcontractors to carry an EMR of 1.0 or lower simply to prequalify or bid, and some large or hazardous projects set the threshold at 0.90 or below. An EMR over 1.0 can knock you out of the running before your price is ever read. In that sense a good EMR is not just cheaper insurance — it is a license to compete. (See our contractor prequalification tool if you are assembling a bid package.)
At its core the EMR is a ratio:
EMR = Actual Losses ÷ Expected Losses
"Expected losses" are what a company of your size, in your class codes, would normally incur given its payroll. "Actual losses" are what you actually reported. If your real losses come in below what's expected for a business like yours, the ratio drops below 1.00 and you earn a credit; if they come in above, you get a surcharge. A few details make this far more nuanced than a plain ratio.
The EMR is not based on last year. It uses the three most recently completed policy years, excluding the current and immediately preceding period. So your 2026 mod is built from roughly 2022–2024 data. This lag has two consequences: a claim does not hit your mod immediately, and once it does it stays in the calculation for about three years before rolling off. Sustained improvement therefore takes a few rating cycles to fully show up — and so does damage from a bad year.
This is the part most people miss, and it is the key to managing your mod. NCCI does not treat every dollar of a claim equally. Each claim is divided at a split point into:
Following NCCI's most significant experience-rating overhaul in a decade — implemented from November 1, 2023 and still rolling through state filings into 2026 — the single nationwide split point was replaced with state-specific split points that track local claim severity. As of 2026 these range from roughly $9,500 in Oregon to about $38,000 in Louisiana, with most NCCI states landing between $15,000 and $25,000.
The practical takeaway is blunt: frequency hurts more than severity. Three $5,000 strain claims will usually push your EMR higher than one $60,000 claim, because each small claim sits almost entirely in the heavily weighted primary layer, while most of the large claim lands in the discounted excess layer. The full methodology is published in NCCI's ABCs of Experience Rating and its Experience Rating Plan Manual.
| Company | Expected losses (3 yr) | Actual losses (3 yr) | Resulting EMR |
|---|---|---|---|
| Acme A (clean record) | $200,000 | $150,000 | ~0.85 (credit) |
| Acme B (industry average) | $200,000 | $200,000 | ~1.00 |
| Acme C (frequent claims) | $200,000 | $260,000 | ~1.25 (surcharge) |
This is simplified — the real worksheet applies primary/excess weighting, a stabilizing value and a ballast factor — but it shows the lever: lower your actual losses relative to expected, and your mod falls. On a $300,000 manual premium, the gap between Acme A (0.85) and Acme C (1.25) is about $120,000 a year for identical coverage.
Because of the rolling three-year window, an injury's effect on your mod is delayed and durable. A serious claim in 2025 will not change your 2025 mod at all; it enters your experience around 2027 and remains for roughly three rating periods before dropping off near 2030. That means:
Lowering an EMR is a multi-year discipline, not a one-time fix. These are the levers that actually move the number:
Notice how many of these are claims-management moves rather than prevention. Both matter: prevention lowers actual losses over the long run, while claims management keeps individual claims small and out of the heavily weighted primary layer.
Want to see what a given EMR does to your premium, or model the savings from dropping your mod from 1.10 to 0.90? Skip the spreadsheet:
What is a good EMR?
Anything under 1.00 is good — it means your loss experience beats the average for your industry and you earn a premium credit. Exactly 1.00 is average. Below 0.90 is considered very good and is a common informal benchmark for well-run safety programs.
What EMR do general contractors require to bid?
Many GCs and owners require an EMR of 1.0 or lower to prequalify, and some large projects require 0.90 or below. An EMR above 1.0 can disqualify you from work even when your price is competitive.
How is the EMR calculated?
The rating bureau divides your actual losses by your expected losses over a three-year experience window. The formula weights frequent small claims (primary losses below a state split point) more heavily than large excess losses — so several small injuries hurt your mod more than one big claim.
How long does an injury affect your EMR?
About three years. A claim enters the calculation roughly a year after the policy period and stays for three consecutive policy years before rolling off.
Can you lower your EMR?
Yes — through return-to-work programs, frequency reduction, active claims management, payroll/class-code audits and fast reporting. Because the formula penalizes frequency, preventing many small claims has the biggest effect over time.
This guide is general information, not insurance or legal advice. EMR methodology varies by rating bureau and state; always confirm your figures against your official experience-rating worksheet from NCCI or your state bureau and consult your broker for advice specific to your business.