A step-by-step guide to understanding your Experience Modification Rate, identifying which injuries drive it, and the interventions that move the needle fastest.
Your Experience Modification Rate is the most consequential number in your entire workers’ comp program, and most of the people it affects cannot explain how it is calculated. That is a problem, because you cannot manage a number you do not understand. The EMR is not a grade you receive at the end of the year. It is a multiplier applied directly to your premium, and increasingly, it is a gate that decides whether you are even allowed to bid on work.
The key to reading it is the part most people skip: the three-year rolling window. Once you understand which years are in the calculation, which losses get weighted most heavily, and which lever the formula quietly rewards, the EMR stops being a mystery and becomes something you can actually steer. Here is how to read yours, figure out which injuries are driving it, and act on the things that move it fastest.
What the EMR actually is
At its core, the EMR is a ratio: your actual workers’ comp losses divided by the losses a business of your size, in your industry, in your state would be expected to incur. The result is benchmarked to 1.0.
- An EMR of 1.0 means your loss experience is exactly average for your class.
- Below 1.0 is a credit. You are safer than average, and you pay less than the baseline premium. An EMR of 0.85 indicates roughly 15% below-average risk.
- Above 1.0 is a debit. You are riskier than average, and you pay a surcharge. An EMR of 1.25 means you are paying about 25% more than the baseline for the same coverage.
That multiplier applies to your manual premium, so the EMR is not a secondary metric. It is one of the largest single factors in what you pay. And because it is a published, portable number, it serves as a prequalification screen. Most major construction clients require an EMR below 1.0 to bid, and many tier-one contractors and energy operators set the bar at 0.85 or lower. For many companies, the EMR is not just a cost lever. It is a revenue gate.
The three-year rolling window
Here is the piece that unlocks how to read it. Your EMR is not calculated on last year’s claims. It uses a three-year window of loss history, and it deliberately excludes the most recent policy year.
A 2026 EMR is built from your 2022, 2023, and 2024 policy years. The year 2025 is left out because those claims are still developing and their costs are not yet stable. Each year at renewal, the window rolls forward: the oldest year drops off, and a newly matured year is added.
This rolling structure has two consequences that matter enormously for how you manage it.
First, a bad year follows you. A cluster of claims in 2024 will sit inside your EMR calculation through roughly three rating cycles before it finally ages out. You do not get to put it behind you next year.
Second, improvements take time to surface. Because the most recent year is excluded and the window is three years deep, the safety gains you make today generally take one to two years to start lowering your published EMR. That is not a reason to wait. It is the opposite: the sooner you act, the sooner the clock starts, and everything you fix now is working its way toward your number while the old losses age out.
How the calculation works, in plain English
Two inputs drive the ratio.
Expected losses are based on your payroll and classification codes. Each class code carries an expected loss rate per $100 of payroll, and the rating bureau uses those to calculate what an average employer with your payroll, in your industry and state, would be expected to lose. This is the benchmark.
Actual losses are the incurred cost of your claims during the three-year window. The word incurred is important: it includes both what has been paid and the reserves, the insurer’s estimate of what an open claim will ultimately cost. That means an open claim with a large reserve hits your EMR at its full estimated value long before a dime of it is actually paid out. Reserves are real EMR dollars.
Then comes the mechanic that surprises almost everyone.
Why frequency hurts more than severity
The formula does not treat all loss dollars equally. Every claim is split into two parts at a dollar threshold called the split point:
- Primary losses are the portion of a claim below the split point. They are weighted at full value because the frequency of claims is the most stable, predictable signal of future risk.
- Excess losses are the portion above the split point. They are heavily discounted because severity is harder to predict, and the formula is designed not to ruin an employer over a single catastrophic event.
One current detail to know: as of a major NCCI plan change that took effect November 1, 2023, and is still rolling through state filings in 2026, the split point is now state-specific rather than a single national figure. It varies by local claim severity, ranging from roughly $9,500 in some states to around $38,000 in others, with most NCCI states falling between $15,000 and $25,000. A representative figure used in NCCI’s own examples is about $18,500. You can get your state’s exact number from your broker or the rating bureau. If you operate in a state with its own independent bureau, such as California, the process is similar but not identical.
Now watch what that split does. Imagine your state’s split point is $18,500. A single $25,000 claim produces $18,500 of fully weighted primary loss and only $6,500 of discounted excess loss. But three separate $8,000 claims produce $24,000 of fully weighted primary loss, every dollar of it primary, with no discount at all. The three smaller claims total less money, $24,000 versus $25,000, yet they hammer your EMR far harder, because the formula reads frequency as the better predictor of future losses.
This is the single most important thing to understand about your EMR: many small claims are more dangerous to your number than one large one. A culture that lets minor injuries pile up will carry a worse mod than one that occasionally has a serious injury but keeps frequency low.
The biggest lever hiding in the formula
There is a powerful incentive built into the plan that most employers underuse: the medical-only reduction.
In most states, through an Experience Rating Adjustment, the losses from a medical-only claim, one with no lost-time or indemnity component, are reduced by 70% in the EMR calculation. Only 30% of a medical-only claim’s primary and excess losses are counted.
Sit with what that means. The difference between a claim that stays medical-only and one that tips into lost-time is not just the indemnity dollars. It is that the entire claim gets a 70% haircut in your EMR if it stays medical-only, and counts at full weight if it does not. Keeping an injury from becoming a lost-time claim is one of the most valuable things you can do to your mod, and it is decided in the hours and days right after the injury.
Which injuries are driving your mod
The deck promised a way to identify what is actually moving your number, and the tool is your NCCI experience rating worksheet. Request it from your broker or the rating bureau and read it, because it shows you exactly which claims are contributing the most primary loss.
When you review it, look for three things:
- The biggest primary-loss contributors. These are your highest-impact individual claims, usually lost-time claims, and usually the ones where early intervention could have changed the trajectory.
- Open claims with large reserves. Because reserves count at full estimated value, an over-reserved open claim can be inflating your mod right now, and resolving or correcting it pays off quickly.
- Frequency clusters. Repeating injury types, the same strain in the same department, are a sign of a prevention gap that is generating multiple fully weighted primary losses.
While you are in the worksheet, audit it for errors. Misclassified payroll, claims attributed to the wrong company, incorrect claim amounts, and closed claims still showing open reserves are all common, and all of them inflate your EMR. Errors are also the fastest thing to fix.
The interventions that move the needle fastest
Ranked roughly by how quickly they show up in your mod:
- Dispute worksheet errors. If the rating bureau is using the wrong data, a correction can lower your mod almost immediately. This is the single fastest lever, and it costs nothing but attention.
- Manage reserves and close open claims before the unit’s statistical reporting date. Claims are valued at a snapshot taken roughly six months after the policy period. A claim closed or properly re-reserved before that snapshot carries a lower value into your next mod. Stale, over-reserved open claims are quiet money.
- Keep claims medical-only. Immediate, appropriate clinical care after an injury prevents a large share of cases from ever becoming lost-time claims, thereby earning the 70% reduction and avoiding a fully weighted indemnity primary loss. This shows up in your very next mod calculation.
- Return injured workers to modified duty quickly. Shortening or avoiding lost-time limits indemnity, holding reserves down, and supporting keeping claims in the lighter category.
- Reduce frequency through prevention. Fewer injuries mean fewer primary losses, which is the highest-weighted input in the entire formula. This is the slowest to mature but the most durable, and within the three-year window, it compounds.
Notice that the first four all operate on the claims you already have, and most of them can affect your next published mod rather than the one after. The EMR’s lag is real, but it is not an excuse to wait, because several of the strongest levers act fast.
Where occupational health changes the math
Look back at that list, and a pattern emerges: almost every fast lever depends on what happens in the first hours after an injury and how the claim is managed from there. That is precisely where an occupational health partner changes your EMR math.
HealthcareLive’s Remote Injury Care connects an injured worker to an occupational medicine clinician in under 10 minutes, which keeps a case medical-only and earns that 70% reduction instead of letting it escalate into a fully weighted lost-time claim. On-Site Programs place a clinician in the building to achieve the same effect. Stretch & Flex attacks frequency at the source, with a documented reduction in musculoskeletal recordables, translating directly into fewer primary losses. And modified-duty support plus virtual MSK care shortens the indemnity and the reserves on the claims that do happen. Each of these maps onto a specific term in the formula that sets your EMR. The number is not a coincidence. It is the sum of choices you make at the point of injury, and those choices are manageable.
The bottom line
Your EMR runs on a three-year rolling window, weights frequency more heavily than severity, counts open-claim reserves at full value, and quietly rewards keeping claims medical-only with a 70% reduction. Once you can see those mechanics, the path to a lower mod is clear: read your worksheet, fix the errors, manage your reserves, and get fast clinical care in front of injuries so they never become the fully weighted lost-time claims that drive the number up. The lag means the work you do now takes a little time to show, which is exactly why now is the time to start. If you want to connect the clinical side of this to the numbers on your worksheet, HealthcareLive can help.
Frequently asked questions
What is a good EMR? An EMR of 1.0 is exactly average. Below 1.0 is better than average and earns a premium credit; above 1.0 is worse and carries a surcharge. Many strong employers run in the 0.70 to 0.90 range. Beyond cost, the threshold often matters for winning work, since many clients require an EMR below 1.0, and some require 0.85 or lower, to bid.
How is the EMR calculated? It is the ratio of your actual workers’ comp losses to the losses expected for a business of your size, industry, and state, measured over a three-year window that excludes the most recent policy year. The formula splits each claim into fully weighted primary losses and discounted excess losses, and it counts both paid amounts and reserves on open claims.
Why do several small claims hurt my EMR more than one big claim? Because of the primary and excess split. Each claim generates a primary loss up to the split point, weighted at full value, while amounts above the split point are heavily discounted. Multiple small claims create multiple fully weighted primary losses, so they can damage your mod more than a single large claim of equal or greater total dollars.
Does a medical-only claim affect my EMR? Yes, but much less. In most states, medical-only claims are reduced by 70% in the calculation, so only 30% of the loss is counted. Keeping an injury from becoming a lost-time claim is one of the most effective ways to protect your mod.
What is the fastest way to lower my EMR? Auditing your experience rating worksheet for errors is usually the fastest, since a correction can lower your mod almost immediately. Managing reserves and closing open claims before the unit statistical reporting date, and keeping new claims medical-only through immediate care, also affects your next mod rather than a distant one.
How long do claims affect my EMR? Because the rating uses a three-year window, a claim typically influences your EMR for about three rating cycles before it ages out. Improvements you make today generally begin to appear in your published EMR within one to two years.
Sources and methodology
This article reflects current NCCI experience rating methodology and 2026 guidance, drawn from NCCI materials including its ABCs of Experience Rating and Experience Rating Plan methodology updates, and from 2026 explainers and worksheets published by brokers and rating specialists, including Akker, PolicyBenchmark, MetricGate, SmartQHSE, and others. Key points include the three-year rolling experience window excluding the most recent policy year, the actual-to-expected loss ratio benchmarked to 1.0, the inclusion of reserves in incurred losses, the primary and excess split and its frequency emphasis, the state-specific split point introduced in NCCI’s plan change effective November 1, 2023 (ranging roughly from $9,500 to $38,000 by state, most between $15,000 and $25,000), and the Experience Rating Adjustment that reduces medical-only losses by 70% in most states. States with independent rating bureaus use similar but not identical methods, and split points and rating values are updated periodically, so confirm your state’s current figures with your broker or rating bureau.
Service descriptions and outcome figures attributed to HealthcareLive, including Remote Injury Care, On-Site Programs, Stretch & Flex, modified-duty support, and virtual MSK care, reflect HealthcareLive’s own program design and network experience.
