The WorkCompEdge Blog

by Jeff Adcock, FCAS, MAAA, Consulting Actuary, SIGMA Actuarial Consulting Group

Editor’s note: This is our final and most detailed article in a series on the 1/1/2010 changes in the California workers compensation experience plan rates and rules. It explains more about the key factors we summarized in our previous post, California 1/1/2010 Mod Calculation Change Advisory.

As you probably know from our previous blog entries, the Workers’ Compensation Insurance Rating Bureau of California (WCIRB) pure premium filing with effective date 1/1/2010 proposed a 22.8% increase in the pure premiums, or filed loss cost benchmarks. In early November, the insurance commissioner rejected the proposed rate increase and ordered no change in the pure premiums overall, although individual class codes did see shifts.

The filing also proposed significant changes to the California Workers Compensation Experience Rating Plan (ERP), primarily driven by recommendations of the Experience Rating Task Force. The insurance commissioner accepted the proposed changes to the ERP, with the caveat that the Expected Loss Rates (ELRs) originally filed should be modified by 1% to maintain the same experience rating off-balance factor of 1.030 that was underlying the 1/1/09 pure premiums.

The following summarize the significant changes to the ERP effective 1/1/2010:

1. Changes/enhancements to the methodology for calculating the ELRs.
2. Use of a single split point of $7,000 to segregate individual claims into primary and excess components, in lieu of the current formula for splitting claims.
3. Changes to the D-Ratios and Credibility Values (“B” and “W” values) to reflect the adoption of a primary/excess split point of $7,000 and to reflect recent frequency and severity trends.
4. Increase in the experience rating eligibility threshold from $15,700 to $16,300 to reflect wage inflation.

The WCIRB states in its executive summary of the filing that “based on the proposed methodology change, the impact should be relatively minor for the overwhelming majority of experience rated employers”. However, based on the several moving parts above and particularly the significant changes in the ELRs for many class codes, it is our opinion that the experience modification factor for more than just a small minority will change. That’s why we believe it is more important than ever to consider in advance the impact these changes will have on a specific client, particularly those that must maintain an experience modification factor below a particular number. Now let’s explore these changes, and the mod trends we’ve seen in our analysis, in more depth.

Changes/Enhancements to the Methodology for Calculating ELRs

The WCIRB recommended changes to the calculation of the ELRs based on

• the final report of the Experience Rating Task Force and
• hindsight studies that showed ELRs for some groups were consistently higher or lower than indicated by actual experience.

The most significant change was using adjustment factors based on and grouped in accordance with the North American Industrial Classification System (NAICS). The following table taken from the WCIRB January 1, 2010 Pure Premium Filing (page B:C-12), shows the impact by NAICS group related only to the change in calculating adjustment factors at the NAICS group level:

Again, the above estimated impacts reflect only the change in the methodology related to the adjustment factors. The ELRs effective 1/1/2010 also changed based on actual loss experience, of course. The individual class codes use between two and five years of policy year data (2002 through 2006) to calculate the indicated limited loss to payroll ratios which are the basis for the ELRs.  For individual class codes, policy year 2006 rolling into the data while an older year rolls off may also have a significant impact.

The table below shows the change from 1/1/2009 to 1/1/2010 in the ELR for the 20 largest class codes by payroll. So for example, for class code 8742, even though the new methodology led to an expected decrease in the ELR of 12.2%, the actual change in the ELR including the impacted of updated experience in the filing, was a decrease of 20.0%.

Change in the ELR for 20 largest class codes in California based on 2006 payroll:

PDF documents listing all class codes and their ELR and D changes, sorted three different ways, are available below

Sorted by Class Code Number

Sorted by Payroll (highest to lowest)

Sorted by Percentage Change in ELR between 1/1/2009 and 1/1/2010

Since most risks have several class codes, the impact on the individual risk will depend on the mix of payroll by class code.  It’s interesting to note that 116 out of 500 class codes, representing almost 50% of the total payroll in the state, have decreases of over 10% in the ELR from 1/1/2009 to 1/1/2010.  96 out of 500 class codes, representing less than 10% of the total payroll in the state, have increases of over 10% in the ELR from 1/1/2009 to 1/1/2010.

Impact of Change to a Primary Split Point of $7,000 per Claim

The ERP effective 1/1/2010 defines the primary loss amount as the actual loss amount capped at $7,000.  Contrast that with the prior method of calculating the primary loss using the formula (9,000 x Total Loss) / (Total Loss + 7,000). A single split limit was recommended to simplify the experience rating formula and to be consistent with the experience rating formula used in other states. While the split limit in most states is $5,000, the limit amount of $7,000 limit was selected to minimize the impact, in total, of the change in formula.  As noted on page B:C-35 of the WCIRB January 1, 2010 Pure Premium Rating Filing, based on the 2006 policy year projection, the D-Ratio for the $7,000 split is 0.203 and the D-Ratio based on the old formula is 0.199.  So the change in the formula had minimal impact on the overall D-Ratios.

Even though the intent of the plan is neutrality in terms of the split of losses between primary and excess, the impact on individual losses varies. The table below summarizes the impact on the actual primary losses included in the experience rating formula for several individual loss amounts.

Change in primary losses for various loss amounts, old vs. new split method

Losses less than $24,500 are included at higher amounts in the primary layer and therefore lead to an increase in the experience mod factor, all else being equal.  Losses greater than $24,500 are included at lower amounts in the primary layer and therefore lead to a decrease in the experience mod factor, all else being equal.

Impact of Changes in the Credibility Values (“B” and “W” Values)

The last significant change to the ERP effective 1/1/2010 is the impact of the updated credibility values – calculated as a function of the Ballast (“B”) and Weighting (“W”) values. The current “B” and “W” values were adopted effective 1/1/1997, so it has been a significant amount of time since those have been updated. In the 13 years in between, claim frequencies have decreased and claim severities have increased.

The decrease in claim frequencies and the increase in claim severities are reflected in the approved “B” and “W” values effective 1/1/2010.  The primary credibility, which is a function of the “B” and “W” values, increased by over 20% for risks with total expected losses (limited to $175,000 each) of between 11,820 and 40,147 during the experience period. Risks with total expected losses (limited to $175,000 each) of greater than 40,147 have primary credibility of 100%.  With the prior “B” and “W” values, a risk needed expected losses of over $1.4 million for full primary credibility.

The higher the primary credibility, the greater the impact of the actual loss experience of the risk. Higher primary credibility will raise the experience mod on a risk with primary losses greater than expected primary losses. Conversely, higher primary credibility will lower the experience mod on a risk with primary losses less than expected.

The excess credibility has decreased on risks with expected losses (limited to $175,000 each) greater than $708,762.

Bottom Line of Impact of Changes to ERP Effective 1/1/2010

1. Change in ELRs – ELRs decreased by 6.4% in total, due to change in methodology and actual experience.  116 out of 500 class codes, representing almost 50% of the total payroll in the state, have decreases of over 10% in the ELR from 1/1/2009 to 1/1/2010.  96 out of 500 class codes, representing less than 10% of the total payroll in the state, have increases of over 10% in the ELR from 1/1/2009 to 1/1/2010.
2. Change in primary loss “split amount” – The move to a fixed amount of $7,000 per claim leads to individual losses less than $24,500 being included at higher amounts in the primary layer and increases the experience mod factors, all else being equal.  Losses greater than $24,500 will be included at lower amounts in the primary layer and therefore decrease the experience mod factors, all else being equal.
3. Change in credibility values (“B” and “W” values) – Credibility factors related to the primary layer have mostly increased, and full primary credibility (i.e. 100% primary credibility) is now given to all risk with total expected losses (limited to $175,000 per claim) of $40,147. This increase in credibility will lower the experience mod for risks with primary loss experience less than expected.  The opposite will occur for risk with primary loss experience greater than expected.

Question: Why did the WCIRB file for a 22.8% increase in pure premiums, but ELRs decreased approximately 6.4%?  Aren’t they related, and shouldn’t the ELRs be increasing in line with proposed pure premiums?

WCIRB filed the 22.8% increase in pure premiums and the decrease in ELRs effective 1/1/2010 at the same time. It should be noted that the ELRs were later adjusted by 1.0% to take into account the impact of the experience rating off-balance.

There are probably a variety of factors at work including:

• the impact of reforms (the 2010 modification factors use 2006 through 2008 which are all post-reform)
• slight differences in indications based on policy year data used in setting the ELRs and the accident year data used in determining the overall pure premium increase
• the future impact of the Ogilvie and Almaraz/Guzman decisions

However, the main reason the indicated change in the ELRs and pure premiums are so different seems to due to the way the insurance commissioner has handled each in the past few rate decisions. While the insurance commissioner has not always approved the increase in the pure premiums proposed by the WCIRB, he has typically approved the changes (most increases) in the ELRs with only minor adjustments due to the impact of changes in the expected off-balance amount.

For example, as of 1/1/2009 the WCIRB filed for a 16.0% increase in pure premiums.  The insurance commissioner approved a 5.0% increase in pure premiums but accepted the ELRs with only a 1% modification based on the off-balance factor change.

Question: Are certain mods more likely to increase as a result of these rate and rule changes?

While there are clearly no hard and fast rules about how a particular mod will change under the new rules, we have observed these trends in our analysis:

• For insureds with a mod above or below 1.00, the new mod formula will increase the gap from 1.00.
• For insureds with a high frequency rate – that is, numerous losses under the 7,000 split point – the new mod formula will increase the mod.
• For insureds with high severity – that is, relatively few losses greater than the 7,000 split point – the new mod formula will show little change.
• The new mod formula will penalize the smaller insureds while having minimal impact on the larger insureds.

We’d love to hear from you. Let us know if your experience and observations with mods under the new California rates and rules coincide with our analysis.