Reform Retro!

The following is from Andrew with the Northwest Progressive Institute:

 

Here's RCW 51.16.035: The Department [of Labor and Industries] shall formulate and adopt rules governing the method of premium calculation and collection and providing for a rating system consistent with recognized principles of workers' compensation insurance which shall be designed to stimulate and encourage accident prevention and to facilitate collection. The department may annually, or at such other times as it deems necessary to achieve the objectives under this section, readjust rates in accordance with the rating system to become effective on such dates as the department may designate.

The above paragraph is from the chapter that establishes the retrospective rating system. The intention is clear - to provide incentives to make construction sites safer and reduce injuries.
The law needs to be clarified to say that middlemen like the BIAW cannot simply decide to keep part or all of their members' refunds. The BIAW is absolutely free to ask its members to voluntarily contribute to a political fund. But the BIAW should not be allowed to game the retrospective rating system for nefarious and vindictive ends.

 

Important Questions which have not yet been answered:

 

1.    How did the Department of Labor and Industries go from improving work place safety to giving away a billion dollars in tax payer money to fund right wing attack ads?

 

2.    A second question is a matter of timing. The BIAW set up their money laundering scam in 1993 and then in 1994, a computer error occurred in order to put millions of dollars into this scam. How did BIAW know about the computer coding error before it occurred? 

 

3.     A third concern is the statement from L & I in January 2009 that they discovered a computer coding error, but that it was only $100 to $150 million dollars in tax payer losses. The above charts make it clear that the losses resulting from this error exceed one billion dollars.

 

L & I has issued press releases about the computer coding error in which they dismiss it as being a “minor” problem which only caused $150 million in over payments to Retro programs since 1994 ($10 million dollars per year). L & I claims this caused retro subsidies to average 22 ½ percent when they should have only averaged 20%. (In other words, $110 million per year when they should have averaged $100 million per year). However, the problem may be 10 times greater than L & I has admitted. The coding error may have caused $1.5 billion in over-payments to Retro programs.

 

There is a very limited amount of information about the computer coding error. Vague statements were made by L & I staff at press conferences in Feb 2009. And the following statement was on page 2 of the Retro Proviso Study group meeting minutes for July 10, 2009: “Losses for non-retro employers included indemnity payments on open claims for the second quarter of the corresponding enrollment quarter twice.”

 

Thus, a single line of code was repeated which doubled the estimate of non-retro program indemnity payments on open claims for the second quarter only of every year for the past 15 years.

 

According to National Council on Compensation Insurance (NCCI), an open indemnity claim is a claim which has outstanding case reserves regardless of whether any payments have been made on those claims, as well as any reopened claims that remain open at the date in question.

 

What was not disclosed by L & I is the percent of open claims to total claims and the cost of the open claims that were double counted. These is some research which might support the conclusion that open claims at one year are about 10% of total claims. This would support the L & I estimate that retro refunds were 10% too high.

 

However, open claims at one year have an average value which is ten times greater than closed claims. In other words, open claims are about half of the total long term “developed” claim costs. This would support concluding that the cost of the double coding error is 5 to 10 times greater than the $10 million estimated by L & I. In other words, the double entry error may account for 100% of retro refunds during the past 15 years (or $1.5 billion dollars). Obviously, L & I should be required to fully disclose how they arrived at their estimate of $150 million in over-payments.

 

For example, if we assume that X = the quarter cost of Non Retro programs and that it also equals the quarter cost of Retro programs (because Retro and non-retro total premiums are nearly identical and the claims are also nearly identical), and that open claims are about one half of total (developed) claims, then the computer coding error =

 

Retro programs annual cost = X + X + X + X = 4 X

Non Retro annual cost = X + (1.5) X+X+X = 4.5X

The over-estimation of Non Retro annual cost is thus 4.5X/4X = 1.10 = Non Retro program costs = 10% too high.

 

Thus, the error in the PAF is 10% too low. Thus, the PAF should have been raised 10% and retro refunds should not have been merely reduced 2 ½ %, but should have been reduced by over 90% (bringing them to near ZERO).

 

 

Calculating revised Performance Adjustment Factor’s compensating for the computer coding error:

 

Note that the PAF is supposed to represent the ratio of savings in Retro programs compared to the claims of non-Retro programs i.e. PAF = Retro Loss Ratio / Non-Retro Loss Ratio. 

 

Year (3rd quarter only)	2002	2003	2004	2005	2006	2007
Reported PAF (Deflated due to coding error)	0.82	0.99	1.09	1.13	1.06	1.08
Adjusted PAF (using L & I  2.5 % correction)	0.85	1.03	1.12	1.15	1.09	1.11
Adjusted PAF (using 10% correction)	0.90	1.09	1.20	1.24	1.17	1.19

 

In other words, even using the L & I estimate over overpayments, since 2003, Retro programs have cost more than Non-Retro programs and thus there should not have been any retro refunds after 2002. Since 2004, retro programs have cost an average of 10% more than non-retro programs. If open claims represented half of total claims, then retro programs cost 20% more than non-retro programs.

 

PAF adjusted for double entry error using L & I estimate of 10% overpayment:

 

The following chart from page 15 of the August 2008 WCAC meeting minutes (http://www.lni.wa.gov/ClaimsIns/Files/Wcac/Minutes/2008/WcacMtgMin20080825.pdf), was intended to show that Retro Claims are 17.5% less than non-Retro claims:

 

Note that 20% subsidies were given even though L & I’s own calculations concluded that PAF’s were greater than one. Thus, the computer coding error can explain nearly all of the cost difference between Retro and Non Retro programs (I.e. all of the Retro Refunds) which has been $1.3 billion dollars during the past 15 years.

 

So who made the determination that the loss was only $150 million (when it was really $1.3 billion) and how they made this calculation?

 

1.    A fourth concern is who at L & I has been promoting the idea that Retro programs save money when there is no evidence to support the claim and plenty of evidence that Retro programs do not save money?

 

Bob Mallooly, Chair of the Workers Comp Advisory Committee (WCAC) stated at an August 2008 WCAC meeting in response to a question, that Retro programs were “20% more cost effective than non-retro programs. “

 

He failed to cite any research to back up this statement. But given the research confirming no cost benefit to Retro programs it is concerning that a high ranking DLI person would make such a claim.

 

At the same meeting, the following chart was shown:

 

When we take the computer coding error into account, Non-Retro programs have always been more cost effective than Retro programs. This should have been obvious to L & I, so why did the error take so long to uncover?

 

WAC 296-17-90402 requires that retro employers as a group and non-retro employers as a group fund the same portion of their total claim costs relative to their total premium charges.

In other words, each group is required to have the same final loss ratio.  (Loss ratios are simply total claims divided by total premiums… a loss ratio greater than one means that claims are greater than premiums). L & I claims that they insure that the loss ratios are equal by adjusting the Performance Adjustment Factor (PAF). However, this claim does not appear to be true because Retro Refunds were issued even when the PAF was greater than one.

 

It is important to note that no other State in America uses a Performance Adjustment Factor. Nor did the State of Washington before the Retro program was started in 1980.

Thus, whatever problems and inequities which exist in our system are due to the failure to correctly calculate the Performance Adjustment Factor.

 

The reason the PAF was not correctly calculated was because the computer coding error misled L & I to believe that non-retro annual claims costs were 10% higher than what they actually were.

 

This led Retro Refunds to be 800% higher than they should have been. In other words, the error was such and the process was such that a very small error in calculating non-retro costs resulted in  very big error in Retro refunds.

 

 

 

5. A fifth concern is the manner in which Retro Program financial information has been hidden on the L & I website. It is not even listed in the “Retro” section. It appears that there was a deliberate effort to hide the facts about Retro Claims and Refunds from the public. For example, this is the link to the Retro Financial information used in this report:

http://www.lni.wa.gov/ClaimsIns/Insurance/Reduce/Qualify/FinInfo/default.asp

Note that the Retro Financial Information does not come up when Retro or Retrospective is entered in the Department of Labor and Industries Website search engine. Nor is the Retro Financial information under the Retro page, but under the “Reduce Premiums” page.

Even if you happen to stumble onto the Reduce Premiums page, you still have to click onto the “Qualify” page before you can link to the Financial Information page.

 

It took clicking on nearly every link on the L & I website to finally track down the Retro Financial information. I would like the information in this report posted on the L & I Retro page including all the L & I Retro financial reports going all the way back to 1992 and I would like a link to this page clearly listed on the main L & I front page.

 

6. A sixth concern is the fact that even if you do finally get to the page posting the Retro Program financial data, it only posts data from 2002 to 2007. There is no historical data to show how the program has skyrocketed in cost over the past 15 years. I had to email the Department of Labor and Industries to get the historical data. Even then, they claimed to only have data going back to 1992. Why isn’t all the financial information posted on their website?

 

7. A seventh concern is the fact that the data they did post for 2007 was incorrect. Looking at the following chart, it is clear that the BIAW income data currently posted on the L & I website does not match up with the BIAW income data which BIAW provided to the PDC in August 2008.

 

L & I reports only $15.7 in Retro Refunds when BIAW reported $55 million in refunds. Were it not for the PDC 2008 complaint, we would not have known about the true extent of BIAW refunds. Why hasn’t L & I updated their table when they have had this information for more than a year? It appears to be a blatant attempt to mislead the public.

 

COMPARING BIAW and BIAW-MSC INCOME: PDC versus L& I reports (in $ Millions)

 

Year	PDC BIAW dues ($90 x 13,000)	PDC Retro Income - Enroll. Fees	PDC Retro Income	PDC est of total income	L& I Retro Income	L&I Retro Refund	Best Guess Retro Give-away
2006	1.2	2.3	3.9	7.4	3.9	39.0	$39 M
2007	1.2	2.5	5.5	9.3	1.6	15.7	$55 M
2008	1.2	2.6	4.4	8.2	NA	NA	$44 M

 

8. An eighth concern was L&I’ s failure to post 2008 Retro Refund information in a timely manner on their website. The above chart shows that BIAW had already received $44 million in Retro Refunds by August of 2008. Yet in August of 2009, there was no information about this on the L & I financial reports. Why is it taking one to two years for L & I to post information on their webpage? Again, it appears to be an attempt to hide important information from the public about the scale of tax payer dollars being diverted to Retro.

 

I believe there needs to be an independent investigation of the Department of Labor and Industries to look into all 8 concerns listed above and to determine how this entire Retro nightmare came about and who authorized giving away hundreds of millions of dollars of tax payer money each year for a program that has absolutely no public benefit.

 

L & I claims that the computer error was an innocent mistake. They also claim that it only cost Washington State tax payers $150 million during the past 15 years. Given that the BIAW alone benefited by over $200 million dollars during the past 6 years, and given that the Retro program errors cost our taxpayers over $130 million additional dollars just in 2006, it is certain that the actual cost of this error was over $1 Billion dollars during the past 15 years. In other words, the actual size of the error is at least 7 times greater than admitted by L & I.

 

Such financial negligence should make every tax payer in our State hopping mad and should be cause for replacing any member of the L & I staff whose job was to catch and correct this huge error during the past 15 years. 

 

The 1994 double entry in the L & I computer code, which artificially increased the estimated cost of non-retro programs by 10%, made retro programs appear to be doing much better than they were actually doing and thereby led L & I to set retro refunds at a rate far higher than they actually deserved.

 

Pure common sense would lead to the conclusion that Retro and Non-Retro programs would perform about the same. It is also highly suspicious that this coding error occurred in 1994, the very year that the Republican party (and the BIAW) took over our State legislature. And the year after BIAW re-organized itself. It is as if the leaders of BIAW knew about the coding error several months before it actually occurred.