Request a demo

Why Did the Ex-Mod Go Up? Five Questions Lead to the Culprit

Thursday, January 19, 2012
Written By
Kory Wells

I have an affinity for detective and crime shows, and this year I’ve added The Mentalist to my list of favorites. The show features character Patrick Jane, who formerly posed as a psychic and now consults with the California Bureau of Investigation. In each episode, Jane reliably uses his superior powers of observation to have the “whodunnit” figured out from the beginning, even though we viewers have to watch the story unfold and wonder, along with the rest of the CBI force, if  the complicating factors and clues are leading us to the right culprit or not.

Why Did the Ex-ModWith a company’s experience rating mod factor, what we want to discover is not WHOdunnit, but WHATdunnit: that is, what caused the mod to increase or decrease from one year to the next? We can be almost as slick and successful as Patrick Jane if we know the right things to observe and the right questions to ask.

Of course companies are more likely to be concerned by a mod increase, but they may also celebrate a decrease (and the broker, risk control expert, and/or safety consultant who helped them achieve such a decrease). Understanding the specific reasons that a company’s mod fluctuates from year to year is important because, as the adage goes, in understanding the past we also understand the future.

Here are 5 questions you can consider and answer, using mostly the summary numbers from 2 consecutive years of a company’s mod worksheets from the applicable bureau, to uncover the cause(s) for a mod increase or decrease from one year to the next. For the purposes of this discussion, remember that the mod is, simply put, a ratio of actual to expected losses that occur over a (typically) three-year experience period: for example, 2012 mods are based on loss and payroll data from 2008, 2009 and 2010.

1. How did total expected losses for the experience period change?

Expected losses are the product of payroll (divided by 100) times the expected loss rate for each job classification in a company. Total expected losses may change, therefore, because payroll has changed and/or because the expected loss rates have changed. For the big picture, here’s what important: if expected losses have decreased, then actual losses have hopefully decreased proportionately – otherwise the mod is going to go up. This leads to questions 2 and 3:

2. How did total payroll for the experience period change?

Slight fluctuations in total payroll aren’t likely to be too significant, unless the payroll codes (that is, the job classifications) being used for the company change. But a notable increase or decrease in total payroll should inspire you to ask other important questions. For example, in a company where layoffs have caused payroll to drop significantly, have claims also decreased proportionately? Or, in a rapidly growing (or in this economy, recovering) company that’s seen a mod increase, could a lack of proper hiring or initial safety training be leading to an upsurge in claims? For example, this Business Insurance article says “recent hiring of new, temporary and seasonal workers may be among factors behind a 2% to 5% increase in claims over the past year.”

Note: Total payroll is not shown on the summary of some bureau’s worksheets, but you can sum it up from the policy period totals, or the company likely has this total handy from another source. 

3. How did the average expected loss rate change?

You can also go deeper with question 1 by considering the overall change in the average expected loss rate that applied to the mods you’re comparing. For a fairly simple mod you may be happy with “eyeballing” the expected loss rates on one worksheet as compared to the next year’s. However, it’s not unusual to see one rate go up a bit while another one goes down from year to year. So you can calculate the average expected loss rate by dividing total expected losses by total payroll times 100.  Payroll and actual losses could stay the exact same from one year to the next, but if the average expected loss rate decreases,  that means expected losses also decrease, causing the mod to increase. This is a scenario I often have to help clients understand, and one that’s beyond the company’s control.

 (Shameless self-promotion: the mod comparison reports in ModMaster help you easily see the total payroll and average expected loss rate, and answer all of the questions posed here.)

4. How did actual losses for the experience period change?

An increase in actual losses is, of course, the most obvious reason for a mod increase. However, we occasionally hear from bewildered clients who have seen a mod increase despite a decrease in losses, they report. Sometimes those clients are forgetting that total losses cover the three years of the experience period, so a one-year improvement in claims isn’t all that’s needed to improve the mod. Sometimes there aren’t any new claims since the prior experience period, but a change in reserves on an open claim makes the actual losses go up. A  state’s implementation of ERA can also affect losses, although most states have been settled into ERA for some time. In some cases, actual losses are down, but expected losses are down more, so the mod still goes up. After you look at total losses, you also need to consider:

5. How did the mix of primary and excess actual and expected losses change?

Both actual and expected losses are split into primary and excess components, and primary losses affect the mod more than excess losses do. Although this is over-simplifying the formula, if you see that actual primary losses have increased  from one experience period to the next, even if total losses stayed about the same, this could be the reason a mod has increased. This scenario will be especially common in 2013 when the primary-excess split point increases from $5,000 to $10,000 for many states.

In detective shows like The Mentalist, there’s usually a twist or two, but in the end everything almost always comes down to one culprit. Understanding why a company’s mod changes from one year to the next is admittedly more complex, because all five of these criteria are likely to have some change.  In my experience, though, you can use these questions to often isolate one or two main factors that have contributed most to a change in the mod. (There can be other primary, though less common, reasons for a mod change, such as application of the maximum mod formula, or a change in ballast or weighting factors.)

Is it in your best practices to analyze what factors caused the mod to change from one year to the next?

– Kory Wells, WorkCompEdge Blog Editor

© 2012 Zywave, Inc.  All rights reserved. For reprint permission, contact the blog editor.







6 responses to “Why Did the Ex-Mod Go Up? Five Questions Lead to the Culprit”

    • Hi, Ron,

      The short answer is that ELR and D-Ratios are developed for each state by actuaries from the appropriate bureau and are, as you probably know, updated annually in most states. I am not familiar with any public documentation that explains the process of developing the rates, although certainly they are based on actual loss experience. Your question does present a good opportunity for more of an explanation of how they function:

      Every payroll classification, or class code, has an expected loss rate (ELR) and a discount ratio (D-Ratio) which reflects the loss history associated with that particular kind of work. Specifically, the ELR indicates the losses expected per $100 of payroll, for purposes of the experience modification calculation, for that specific type of job classification. The D-Ratio reflects the split point between expected primary and excess losses. (This is why the D-Ratio is being adjusted as part of the new split point approved in many states for 2013.)

      Expected losses are determined by multiplying total payroll (per $100) for each class code by the ELR for that class code.

      Expected primary losses are determined by multiplying expected losses for a class code by the D-Ratio for that class code.

      Expected excess losses are the difference between expected losses and expected primary losses for each class code.

      Let’s look at an example:

      In Tennessee, a long-haul driver (code 7228) has an ELR of 3.61 and a D-Ratio of 0.11 effective 3/1/2012.

      If a company has $1,000,000 in code 7728 payroll, then its expected losses are 3.61 * 1,000,000/100 = 36,100

      Its expected primary losses are expected losses * the D-Ratio: 36,100 * 0.11 = 3,971

      Its expected excess losses are expected losses – expected primary losses: 36,100 – 3,971 = 32,129

      In 2013, when D-Ratios are expected to increase approximately 50% to account for the new $10,000 split point, the D-Ratio will likely be approximately .17, and this example would show

      Expected primary losses: 36,100 * 0.17 = 6,137

      Expected excess losses :36,100 – 6,137 = 29,963

      One important aside to the discussion of ELRs: Sometimes folks confuse loss costs and ELRs. Loss costs are the expected ULTIMATE cost per $100 of payroll for a specific class code. The ELR represents the expected incurred losses per $100 of payroll as of the evaluation date of the (typically) three years going into the experience mod calculation. Loss costs are not used in the mod calculation.

      If any of our other readers know more details about the development of the rates, Ron and I both would be glad to hear from you!

      Ron, thanks for reading and being in touch!

      Kind regards,


  1. Are premium rates tied to the ELR? Curious how premium rates can go up for a classification code while the ELR goes down for the same classification code in a given year

  2. Great, emphasize great, article. Question for you- after the insured’s policy was issued, the rating board revised the Ex-Mod factor down (from 1.63 to 1.06). The insured then received an endorsement from the insurer amending the ex-mod as indicated but also adjusting a “Loss Rating Factor” upwards from 0.95 to 1.46. The result was a new higher premium during the policy period. The policy doesn’t provide for such increases. Am I crazy or is there something amiss here? Please, your thoughts.

Leave a Comment

Your email address will not be published.