In the Matter of :
Workers’ Compensation Insurance Rate : OPINION AND DECISION
Application of the New York :
Compensation Insurance Rating Board :
On May 15, 2003, an application for a proposed 11.3% increase in manual rates
was submitted to the Insurance Department (Department) by the New York Compensation
Insurance Rating Board (NYCIRB). This change, combined with no change in the
previously approved provision for catastrophes, produces an overall rate increase of
11.0%.1 The catastrophe provision is separate and distinct from the surcharge for mandated
assessments used primarily to fund certain Workers’ Compensation losses. Although the
assessment surcharge is not formally a part of this application, and is not subject to
Department approval, the NYCIRB calculates changes in the assessments paid by Workers’
Compensation insurers. These assessments, according to NYCIRB, will increase by 1.2%.
Overall, the net effect is an increase of 12.3%. A public hearing was held on the rate filing
on July 3, 2003.
This opinion and decision sets forth a discussion of the NYCIRB 2003 filing. The
Department has reviewed the NYCIRB application for an increase in manual rate and hereby
concludes that the filing should be disapproved. The Department finds that the requested rate
change was unsubstantiated generally by the NYCIRB and that the following items, among
others, have not been satisfactorily considered in this filing:
• The filing does not consider the increased usage of Section 32 of the Workers Compensation
Law, which permits insurers to settle some claims on a sum certain basis at significant
ultimate savings. The Workers’ Compensation Board (WCB) reports that there were
A catastrophe load of 3% was approved in last year’s filing effective October 1, 2002.
approximately 1,000 such settlements per month in 2002, as opposed to only several hundred
per year in 1999. The use of Section 32 on these cases results in savings both in pure loss
costs and expenses related to these losses. Despite the marked increase in Section 32
settlements over the past two years, as reported by the Workers’ Compensation Board, the
filing fails to quantify or address any decrease in the expense allotment, or decrease in pure
loss costs attributable to these settlements.
• The 1996 reforms required each company to staff a special investigative unit to detect, report
and prosecute workers’ compensation fraud. The NYCIRB has only recently begun to
include a separate code for fraud cases. Although NYCIRB has conceded the beneficial
financial effect of recent fraud prevention and control measures, no fraud cost savings data
were included in the filing or considered in arriving at the submitted rate.
• At the hearing held on this filing, on July 3, 2003, the rising cost of pharmaceuticals was
cited as a primary driver of increased medical costs that supported the proposed rate increase.
Although requested by the Department, the NYCIRB was unable to supply any data that
specifically quantified the magnitude of the rise in drug costs and the relationship of these
costs to the rate request.
• NYCIRB unilaterally decided to eliminate the use of loss trends based on both 5 and 8 year
time horizons. Over the last several submissions, the frequency trend based on both these
time periods indicated a significant downturn. The use of these trends in the current filing
would have indicated a substantially lower rate indication. Their omission calls into question
the accuracy of the requested rate increase.
• Policies written on a large deductible basis continue to comprise a significant portion of the
premium volume. As in the past, the experience of these policies has been excluded from this
rate filing by NYCIRB; in each of the past filings the department has restored this experience
in its deliberations on rate decisions. If the experience of the large deductible policies had
been included, the overall indication from just this factor would have been materially lower.
• At the hearing held on July 3, 2003, the NYCIRB was specifically asked to provide more
detailed information on the calculation of both allocated and unallocated loss adjustment
expenses for New York. The Department was particularly concerned about how much of this
information reflects actual New York expenses as opposed to countrywide data. However,
the data supplied to the Department in response to this inquiry was unresponsive to
Department concerns. It remains unclear whether the allocated loss adjustment expense
figure, listed in the filing, was derived solely from the carriers’ New York experience. In any
event, NYCIRB did not provide sufficient analysis of the effect of numerous expense saving
mechanisms implemented by the Workers’ Compensation Board over the preceding years.
• The need for a rate increase given the industry-wide 4.5% return on investment cited in the
• Continued use of downward rate deviations is of great concern to the department. One
hundred nineteen carriers have received approval for such deviations which range from a low
of 5 percent to a high of 17.5 percent. Inasmuch as the department recognizes the slight
reduction in use of such downward derivations since the last filing, we also note the
continued incongruity of the industry’s request for a significant rate increase and the
continued use of downward deviations from the manual rate.
The NYCIRB serves as the private rate service organization for Workers’ Compensation
in New York State and is licensed pursuant to section 2313 of the Insurance Law. By statute, all
Workers’ Compensation insurers must report statistics to a licensed statistical agent.
In turn, NYCIRB compiles and evaluates the data and proposes rate changes, which
must have the Department’s prior approval before implementation. Pursuant to Section
2305(b) of the Insurance Law, a public hearing must be held if the requested rate change
is greater than 2%. In view of the fact that the NYCIRB filed for a rate increase of
greater than 2%, a public hearing was held on July 3, 2003.
The table below summarizes these rate changes from 1995 to the present:
Year Rate Level Change
1996 -14.9% and 3.2%*
A. Policy Year and Accident Year Experience
In this revision, the NYCIRB has submitted loss experience of Policy Year 2001 and Accident
Year 2002. Both Policy Year and Accident Year methods rely on the use of loss development
factors to project losses to ultimate settlement value. Such factors must be evaluated each year to
assure that they are reasonable projections of past development patterns. Application of these
loss development factors to Policy Year 2001 data indicates a rate level increase of 11.8% and a
rate level increase of 11.1% for Accident Year 2002 data. These rate level indications, however,
are subject to adjustment based upon appropriate trending, expense loading and other factors.
Moreover, as previously noted, significant loss and expense cost savings were not included in the
*3.2% decrease due to the impact of the 1996 Workers’ Compensation reform statute.
B. Underwriting Profit Provision
In 1991, in its first Opinion and Decision on Workers’ Compensation rates, the
Department established that the appropriate target underwriting profit for Workers’
Compensation would be set at 0%. This does not mean, however, that insurers make no profit at
all; rather, insurers’ profit margin is to be derived from investment return instead of underwriting
results. This target profit provision appears to have worked well in enabling insurers to earn a
reasonable return on capital, although actual underwriting results can and do vary from the 0%
target. In the recent past, New York has been able to maintain a competitive and healthy market,
with profitability in a reasonable range and the 0% target profit provision will be maintained.
The year 2001 does not appear to be as profitable, since it includes losses for the one-time 9/11
disaster. Nevertheless, NYCIRB noted at the public hearing that its latest internal study revealed
a 4.5% rate of return for the industry. The Department will, of course, monitor this issue closely.
The table below displays returns on net worth and rate history for each of the last eight
years for Workers’ Compensation business in New York. The source of these figures is the
National Association of Insurance Commissioners’ Report on Profitability by Line and by State.
Year Rate Requested Rate Approved Return on Net Worth
1994 1.6% 1.4% 5.4%
1995 -6.3% -8.4% 8.4%
1996 -14.9 -18.0 6.5%
1997 -2.8 -7.5 8.1%
1998 2.7 -3.2 11.4%
1999 12.6 0.0 8.6%
2000 -2.5 -2.5 11.0%
2001 0.04 0.0 -8.1% (Includes 9/11 losses)
The effect of the benchmark reforms enacted in 1996 continues to impact the data for the
years that support this application. These reforms include:
• Section 32 of the Workers’ Compensation Law which permits insurers to close some
claims on an expedited basis;
• Section 110 of the Workers’ Compensation law permitting insurers to close some
medical-only claims on an expedited basis;
• Limitation on certain “Dole v. Dow” claims;
• The increased use of managed care and preferred provider organizations;
• Anti-fraud initiatives required of companies, including the establishment of special
investigations units, and the sizeable contributions of government, e.g., the creation of the
Insurance Department Frauds Bureau’s Workers’ Compensation unit and the Workers’ –
Compensation Board’s inspector general unit, both of which have added substantially to
the efforts to detect, prosecute and prevent Workers’ Compensation insurance fraud.
C. Loss and Wage Trend Factor
In an effort to produce a more responsive trend factor, the NYCIRB included the
results of econometric modeling in its last two submissions. This year, because of the
Department’s previously articulated reservations about the accuracy, reliability and execution of
such methods and their less than stellar performance, the NYCIRB chose not to include it.
The wage/loss trend procedure, used in the last several rate revisions, is based on
actual values of the data being estimated. Loss data and wage data are analyzed separately. The
trend factors can be modified to exclude elements considered elsewhere in the filing, e.g. benefit
changes or statutory revisions.
The wage factor is designed to account for rising payrolls, which exceed the
corresponding increase in exposure to loss. The loss trend factor separately measures
changes in frequency and severity of medical and indemnity losses. As noted last year, this
often-revised trend methodology has not proven to be a consistent predictor of loss trend. One
shortcoming of this method is that the data supporting the indication is at least two years old.
The wage factor is based on published annual DOL statistics; in the past quarterly data
was provided verbally to the NYCIRB. The “published” aspect of the data, while not making a
material difference in the results, increases our confidence as to its verifiability. Unfortunately,
there was a delay by DOL in publishing the final 2001 wage figures so that they had to be
extrapolated from earlier data.
In this revision, there is a slight increase in the indemnity claim cost trend and the general
medical claim cost trend but both are tracking well below the rate of inflation and medical
inflation in particular. There is also continuation in the steady decline in the frequency trend that
has existed since 1992. The filing, however, fails to provide any vector analysis between claim
frequency and claim severity trends to illustrate the net impact or other correlative relevance
between the juxtaposed trends in frequency and severity. Additionally, the wage trend, which
has an inverse relationship to rate level, indicates a smaller rate need.
Contrary to prior filings, NYCIRB has not utilized an overall New York Trend factor for its
requested increase of 11.0%. Traditionally, the NYCRB filing has included loss trends based on
both 5 and 8 year time horizons. As with the large deductible factor issues, had a 5 or 8 year loss
trend been included in this filing the indicated rate would have been significantly lower than the
rate submitted for approval. It is plainly self-serving for NYCIRB to unilaterally deviate from the
time-tested practice of trending experience for determining rates and resorting to a methodology
of rate construction that results in exaggerated loss projections and, consequently, rate
indications. The Department must reject this departure from both the practice and actuarial
principle upon which previous rate determinations have been based.
D. Large Deductibles
In response to legislation effective July 1, 1990 the Department approved Large
Deductible Programs for Workers’ Compensation insurers in the State. Business written on a
large deductible basis has grown significantly in the time such programs have been available. For
example, in 2001, the equivalent premium cost for this coverage would have exceeded over $1.1
billion. This is in the face of a total decrease in policyholder cost level of 32.5% since 1995.
When these programs were first introduced, NYCIRB believed that this business might be
significantly better (in terms of underwriting results) than the business not written on this basis,
and therefore should not be included with the rest of the ratemaking data. This could result from
better claim control methods or the likelihood of superior risks to choose such a program. Data
from recent years seem to show that, as the volume of large deductible experience increased, its
indications began to map more closely nonlarge-deductible experience. This recent data supports
the Department position that large deductible experience should be included in the rate filing. As
previously noted, experience developed on large deductible policies was excluded from this
filing. The inclusion of this data, would have resulted in nearly a three per cent decrease in the
overall rate indication.
Effective April 1, 1994 the portion of the premium that paid for loss based assessments
was removed and replaced by an assessment fee to be paid by the insured at the same time as the
premium. With this revision the New York Assessment Fee that is applied to an insured’s
standard premium changed from 13.0% to 14.3%. While this element is technically not part of
the rate revision, it does affect the premiums paid by Compensation insureds and is therefore
taken into account by the Department.
Based on the review of the filing conducted by the Department and the public hearing held on
July 3, 2003, the Department hereby disapproves the rate filing submitted by NYCIRB on May
DATED: July 15, 2003 Gregory V. Serio
New York, New York Superintendent of Insurance