Skip Maine state header navigation
Skip All Navigation
|Home | Contact Us | Careers | Calendar|
Maine.gov > PFR Home > Insurance Regulation > All Hearing Decisions > INS 11-1000 : Hearing Decision
STATE OF MAINE
II. PROCEDURAL HISTORY
On January 28, 2011, Anthem filed a request to increase rates for its individual HealthChoice, HealthChoice HDHP, HMO Standard, HMO Basic, and Lumenos Consumer Directed Health Plan products and to maintain its current rates for its HealthChoice Standard and HealthChoice Basic products. The Bureau of Insurance designated the matter as Docket No. INS-11-1000.
On February 7, 2011, Anthem submitted revised rate sheets correcting an error in the original filing, which applied the wrong age factors to the HealthChoice HDHP product. As a result, the average requested increase was 9.7% rather than the 9.6% stated in the original filing.
On February 8, 2011, the Superintendent issued a Preliminary Notice of Filing and Related Procedural Matters. The Preliminary Notice advised of Anthem’s rate filing; outlined the purpose and legal standards for the proceeding; explained that a public hearing would be held in early April, 2011; set evening public comment sessions for March 14, 2011, in Orono and March 22, 2011, in Portland; set an intervention deadline; established the scheduling of weekly meetings with Bureau staff to facilitate the information gathering (discovery) process and provide parties an opportunity to clarify information responses through open dialogue; and explained about discovery and proceeding procedures. The Superintendent further advised that she had hired the consulting firm Compass Health Analytics, Inc. to assist the Superintendent’s hearing panel in the proceeding.
On February 18, 2011, and March 4, 2011, weekly Bureau staff meetings were held with Anthem, the Maine Attorney General, and Consumers for Affordable Health Care (“CAHC”) participating. No party requested any additional weekly meetings thereafter. On March 8, 2011, Anthem filed a summary of the March 4th meeting, which is part of the record of the proceeding.
On March 1, 2011, the Superintendent issued a Notice of Hearing, setting April 5, 2011, and, if necessary, April 6, 2011, as the dates for the public hearing. The Superintendent further identified the Attorney General and CAHC as parties to the proceeding.
On March 2, 2011, Anthem submitted a revised filing and supporting exhibits. The revised exhibits corrected the error noted in its February 7 submission and reflected healthcare management costs, which had not been included in the initial filing due to an error. The proposed rates were not changed from the corrected rates submitted on February 7.
On March 2, 2011, Anthem requested, due to witness unavailability, that the public hearing be re-scheduled. The Superintendent, upon agreement of all parties, re-scheduled the public hearing to April 12, 2011, and, if necessary, April 13, 2011. The Superintendent further scheduled an evening public comment session in Gardiner for April 11, 2011. Members of the public also were advised of their opportunity to attend the hearing and to provide comments during the hearing if unable to attend the scheduled evening public comment sessions.
On March 16, 2011, the Superintendent issued a Procedural Order establishing additional procedures for the conduct of the proceeding, including deadlines for serving discovery requests and for submission of pre-filed testimony and exhibits. Moreover, given that notice to policyholders of the first Orono evening public comment session was delayed, a second evening public comment session was scheduled for Orono and held on April 5, 2011.
Beginning in February, 2011, the Attorney General and CAHC engaged in discovery on Anthem’s rate filing. The Attorney General and CAHC each served Anthem with five separate discovery requests. Anthem filed responses, objected to certain discovery, and supplemented many responses (all as reflected in the record of the proceeding). Some of Anthem’s responses were provided after rulings by the Superintendent on motions to compel production.
Specifically, the Superintendent issued discovery rulings by Orders issued on March 15, 2011, March 22, 2011, March 23, 2011, and March 31, 2011. Beginning in March, 2011, the Superintendent issued two pre-hearing discovery requests on Anthem, to which Anthem filed responses.
On March 22, 2011, the Superintendent issued a Protective Order granting Anthem’s request for confidential treatment (in varying degrees) to certain provider contracting information. Citing 1 M.R.S.A. § 402(3)(B), and 24-A M.R.S.A. § 2736(2) (exempting certain insurer-third-party contract information from public disclosure).
On March 28, 2011, the Attorney General requested that the Superintendent enlarge the discovery and pre-filing deadlines, and continue the evidentiary public hearing. Anthem opposed a continuance of the hearing. By Order issued March 30, 2011, the Superintendent granted that part of the motion which requested an enlargement of discovery and pre-filing deadlines, but denied a continuance of the hearing. In denying the continuance, the Superintendent explained:
On March 14th in Orono, March 22nd in Portland, April 5th in Orono, and April 11th in Gardiner, the Superintendent held evening public comment sessions providing members of the public an opportunity to make either sworn or unsworn statements for her consideration. Sworn testimony was received from nearly 45 members of the public.1
On March 31, 2011, Anthem submitted a supplemental filing with revised exhibits to reflect experience paid through February, 2011, and including other additional updates. While the supplemental filing stated Anthem’s position that a larger rate increase would be justified (an average of 10.2%), Anthem did not request a modification to the rates proposed in the initial filing.
On April 7, 2011, Anthem, the Attorney General, and CAHC filed prefiled testimony and exhibits.
The public hearing was held on April 12 and 13, 2011, and was conducted nearly exclusively in public session.2 Members of the public had an opportunity to make either sworn or unsworn statements for consideration by the Superintendent. Members of the public also submitted written comments outside the public hearing which the Superintendent designated a part of the record of this proceeding. The Superintendent has read each of the written comments provided. To the extent that they comment on facts that are in the record, they shall be considered for their persuasive value in the same manner as legal arguments and other comments submitted by the parties. However, the Maine Administrative Procedure Act bars the Superintendent from relying on unsworn submissions as evidence when making her substantive decision. 5 M.R.S.A. § 9057.
At hearing, Anthem presented testimonial evidence from Jennie Casaday, Actuarial Director supporting individual product pricing; Andrew Wei, Actuarial Director with the Advanced Analytics and Innovation department; and William Whitmore, Regional Vice President of Underwriting. As part of its direct case at hearing, Anthem made certain changes to its filing resulting in a revised average rate increase request of 9.2%. CAHC presented testimonial evidence from Lawrence Kirsch, managing partner of IMR Health Economics;3 and Barbara Niehus, a consulting actuary with Niehus Actuarial Services, Inc. The Attorney General presented testimonial evidence from Beth Fritchen, Actuary and Principal with Oliver Wyman Actuarial Consulting, Inc. The Superintendent admitted into evidence Anthem Exhibits 1 through 11;4 CAHC Exhibits BN 1 through 9, and Appendices BN A through P, Exhibit LK 1, and Appendices LK A and B; and Attorney General Exhibit 1, with attached Exhibits A through C. The Superintendent also admitted into evidence responses to discovery filed throughout the proceeding, and took official notice of certain matters as reflected in the hearing transcript.5 After the parties rested their cases at hearing, the Superintendent adjourned the hearing for the submission of responses to certain questions posed at the hearing, followed by written closing argument.
Post-hearing responses to hearing panel inquiries were filed on April 14, 2011, by the Attorney General; April 15, 2011, by CAHC; and April 19, 2011, by Anthem.
On April 22, 2011, each party filed its written closing argument.
On April 29, 2011, the Superintendent issued a follow-up inquiry to Anthem’s post-hearing responses (filed April 19, 2011), to which Anthem responded on May 2, 2011.
On May 9, 2011, the Superintendent issued a further post-hearing inquiry to Anthem, to which Anthem responded on May 10, 2011. Also on May 9, 2011, the Superintendent offered for admission into the record (either through official notice or as Hearing Officer Exhibits) Anthem’s 2007 through 2010 Management Discussion and Analysis (“MD&A”) regulatory filings made with the Bureau of Insurance.6 Anthem objected based on relevancy, and the AG and CAHC had no objection to the Superintendent’s offer. The Superintendent hereby overrules Anthem’s objection and admits the 2007 through 2010 MD&As into the record.
On May 10, 2011, the Superintendent issued a follow-up inquiry to Anthem’s second post-hearing responses (filed May 10, 2011), to which Anthem responded on May 11, 2011.
Anthem has provided direct written notice by mail to every affected policyholder advising of the proposed rate increases.
III. LEGAL STANDARD
Anthem is required by 24-A M.R.S.A. § 2736(1) to file proposed policy rates for its individual health insurance products with the Superintendent. The Superintendent may approve the filed rates only if they are not excessive, inadequate, or unfairly discriminatory. 24-A M.R.S.A. § 2736(2). Pursuant to 24-A M.R.S.A. § 2736-C(5), the rates must be likely to yield a loss ratio of at least 65% as determined in accordance with accepted actuarial principles and practices. That is, expected claims payments must be at least 65% of premium. Anthem as proponent of the filed rates bears the burden of proving by a preponderance of the evidence that the proposed rates meet statutory requirements.
The Superintendent finds that the proposed rates filed by Anthem in this proceeding are not inadequate. However, the Superintendent does find that the proposed rates as submitted by Anthem are excessive and unfairly discriminatory in contravention of section 2736 for the reasons discussed more particularly below.
1. Removal of Large Claims.
A common and appropriate step in developing premium rates for health insurance is to smooth the impact of large claims on claim levels and claim trend. Often, an apparent increase in claim trend is actually the result of an unusually large concentration of large claims in a relatively compressed time frame, and conversely, an unusually low number of large claims could result in an understated claim trend. The smoothing of claim experience is accomplished by subtracting out large claims in excess of a specified threshold and adding back in a pooling charge. The pooling charge is designed to yield the same amount over a period of several years, but without the month-to-month or year-to-year peaks and valleys that are apparent from the actual large claim experience.
In last year’s filing, Anthem derived its rate calculation from a process that did not include pooling of large claims, but supported it with an alternate calculation that pooled all claims in excess of $100,000 for any one claimant in a 12-month period. In this year’s filing, Anthem is pooling all paid claims in excess of $50,000 in any month.
In the original filing, Anthem subtracted $3,269,203 for large claims in excess of $50,000 in a month. This was determined as the large claim amount for the base claim period of October 2009 – September 2010, with runout through November, 2010. In its revised filing submitted on March 31, Anthem updated this amount to $3,458,182 by considering additional runout through February, 2011. In her prefiled testimony, Ms. Fritchen asserted that an additional completion factor adjustment of 1.045 was appropriate for a 12-month period with five months of runout, based on a more complete analysis of runout over a longer runout period. At the hearing, Ms. Casaday indicated agreement with Ms. Fritchen’s approach and agreed to incorporate the 4.5% adjustment into the Anthem proposed rate increase (April 12 hearing, p. 17, lines 19-25).
The CAHC expert, Ms. Niehus, did not address this issue in her pre-filed testimony, but endorsed the 4.5% adjustment in her testimony (April 13 hearing, p. 33, line 25, through p. 34, line 4).
The Superintendent has reservations about the 4.5% factor determined by Ms. Fritchen and adopted by Ms. Casaday, because the factor was developed as the average of four calculated values which are based on 12-month periods with a high degree of overlap. In fact, all four 12-month periods include the experience for the nine-month period from August 2009 – April 2010. A better approach might have been to test multiple independent periods over a larger time frame and to consider a longer runout period. Data provided in discovery indicates that more comprehensive sampling would be feasible. However, there is no evidence in the record of this proceeding to support an alternate value, so the 4.5% adjustment proposed by Ms. Fritchen and accepted by Anthem is approved.
2. Base Trend.
Anthem based the claims projection used in its proposed rates on an allowed-charge trend of 6.3%. “Allowed charges” are the total cost of covered services before considering deductibles and other cost-sharing, as distinguished from paid claims, which are the actual benefits that were paid. The use of allowed charges results in a trend calculation that reflects the expected changes in underlying costs and utilization of health care services. Adjustments are then made for deductible leveraging and deductible mix in order to derive an annualized trend rate that can be applied to paid claims for the baseline experience period to project benefits paid during the rating period.
Anthem’s first step in its trend calculation was to determine “spot trend” by evaluating rolling 12 month moving weighted averages of allowed charges per member per month (PMPM) for HealthChoice claims only. From this analysis, Anthem determined that, in its judgment, allowed-charge trends followed a quasi-cyclical pattern over time and the 30-month period from October 2007 – March 2010 was the appropriate period for measuring trend because it encompassed two full trend cycles.
The regression results are shown in Exhibit VI.A of the rate filing. Beginning with allowed charges, Anthem subtracted out all claim dollars in each month for any member with claims in excess of $50,000 in that month, consistent with the large claim pooling approach discussed above. Anthem then adjusted for seasonality and workdays and fitted a linear regression curve to the observations for the 30-month period from October 2007 – March 2010. The resulting allowed-charge trend in the initial rate filing was 6.5%. In the March 31 revision, which included three additional months of claim runout, the regression indicated a trend of 6.3%.
Ms. Fritchen was generally supportive of Anthem’s trend analysis in her pre-filed testimony, while Ms. Niehus did not comment. No significant disagreement on the base trend level emerged during the hearing.
The Superintendent’s consultants also reviewed the trend analysis and determined the following:
When these observations are considered, the Superintendent is satisfied that the allowed-charge trend of 6.3% proposed by Anthem is supported by the preponderance of the evidence in this proceeding.
In future filings, Anthem should consider more approaches to pooling and select one that produces a material smoothing of the observed results, as that is the purpose of pooling. For example, using the sequence of monthly PMPMs for the HealthChoice product that was the basis of Anthem’s analysis, the coefficient of variation for the PMPM with all claims is 0.105, while the coefficient of variation for the sequence which removes all claims for those individuals with claim amounts in excess of $50,000 in a month used by Anthem is 0.104. Though it had no appreciable effect on smoothing, the use of the sequence that removed claims in excess of $50,000 PMPM produced a predicted claim trend rate in the consultant’s modeling that was about a percentage point higher than one using all claims. This increase in the estimated trend, while it would have been objectionable in isolation, was offset by the decrease resulting from Anthem’s choice of products to include in its analysis. The inclusion of all products, other things equal, would have increased the projected trend by between one and one and a half percentage points.
There are other issues worth noting that do no appear to have had an effect on the projected trend in this case but are not advisable for future rate filings. First, the use of what was termed a “spot trend,” which appeared to mean using visual inspection of data to determine the period from which to base an analysis is not a sound approach to estimation. Second, a central virtue of multiple regression analysis is the ability to include multiple covariates in the regression to explain and predict the dependent variable (in this case, PMPM claims). It would be preferable to include seasonal variables and work-day variables in the regression equation rather than adjusting the data for these factors and then performing the regression. Including these variables in the regression allows the relationships between the variables to be inferred directly from the historical data and then used in the projection, rather than forcing an assumed relationship on the data via adjustment prior to regression.
Overall, the approach taken by Anthem does not appear to have misstated the estimated trend, and the trend Anthem calculated is therefore approved, but the reasonableness of the result may be due more to coincidence than to suitability of the methods chosen. The Superintendent’s approval in this proceeding should not be construed as an endorsement of Anthem’s approach.
The Superintendent notes that various exhibits developed by Anthem to support this rate filing reflect varying approaches to large claim pooling and also varying definitions of which individual products are included. Some exhibits are HealthChoice only, others are HealthChoice + Lumenos while others are HealthChoice, Lumenos, HDHP, and HMO. Large claim pooling is based on paid claims in some exhibits and allowed charges in others. In some exhibits, pooling is accomplished by removing only claim amounts in excess of $50,000 per member in the month, while in others all claims for members exceeding $50,000 are removed.
Anthem may have valid technical reasons for using this variety of approaches, but if so, Anthem should make clear what is being varied and provide a comparison of results and justification of the deviation from a consistent approach when judgment suggests such a deviation is necessary. One essential aspect of the Superintendent’s review of this filing is to confirm that the various exhibits are consistent with each other and also consistent with the raw claim and enrollment data as well as revised data that may emerge during the review process. Prior experience with Anthem filings has confirmed the need to do this testing. The inconsistent approaches in the various exhibits in this filing greatly increased the resources and elapsed time required to complete the review, which is counter to Anthem’s desire to have the increase reviewed in a timely fashion.
3. Deductible Leveraging Factor.
Anthem applied an adjustment of 1.29 to the allowed-charge trend to reflect the impact of deductible leveraging on paid claims. It is a well accepted principle of health insurance that claim trend for claims in excess of a fixed deductible amount will be higher than claim trend for all allowed claims or charges.
For this filing, Anthem calculated this factor based on the distribution of premiums by deductible level. In the past, Anthem calculated this factor based on the distribution of claim amounts by deductible. When asked to explain the changed approach (Superintendent discovery question 1-28a), Anthem responded: “Anthem used premiums because the amounts are more stable than the claims cost associated with each plan design.” Anthem further responded that the factor would be 1.24 if based on claims.
Ms. Fritchen, in her prefiled testimony, stated that she had arrived at a similar result using proprietary data. Ms. Niehus did not comment on this assumption.
The hearing panel requested additional information on this topic (April 12 hearing, pp. 151-153), but Anthem’s April 19 answer was not responsive to the question. Anthem instead provided an alternate calculation using a proprietary model based on the Milliman Health Cost Guidelines showing a deductible leveraging factor of 1.30. The Superintendent followed up with another request to: 1) provide the originally requested information; and 2) provide more information on the alternate calculation, the basis of which did not seem to line up with HealthChoice experience. Anthem’s response on May 2 was (in part): “As Anthem reviewed this request post-hearing, we determined that the analysis was not meaningful because the merit of using a premium or claims based weighting was not really the issue, but, rather, the appropriate deductible leveraging factor.” Whether to use premium or claims was precisely the issue Anthem was asked to address. Anthem did not answer the original question, but simply reiterated that its approach was correct. Anthem described its alternate calculation as “modeled” values based on a commercial population, but provided no evidence that the population’s claim distribution had been adjusted to conform to the HealthChoice population. Therefore, the Superintendent concludes that the alternate calculation is not reliable.
The Superintendent finds that Anthem has not satisfied its burden of proof that the deductible leveraging factor can reasonably be determined by using a premium distribution rather than a claim distribution. The Superintendent, therefore, would approve a claim-based factor of 1.24.
4. Deductible Mix Factor.
Anthem proposed a Deductible Mix factor of 3.1%. This factor is characterized by Anthem as “a normalization factor of sorts to recognize changes in the allowed cost trends over time due to benefit shifts to higher deductible plans.” (Response to Post Hearing Information Request PH4.)
This factor was calculated in Exhibit VI.C of the rate filing. It is calculated as the annualized rate of change of a weighted average of claim costs from the baseline experience period to the rating period. The weighted average computed reflected the change in deductible mix over time.
It was difficult to confirm the reasonableness of the results from Anthem’s exhibit. The methodology of the exhibit was very different from the approach taken last year to quantify this factor. Anthem did not adequately disclose how some of the data in the exhibit was derived. In future filings, Anthem should strive to make this exhibit more transparent by disclosing the derivation of the data used in this exhibit and by explaining how the method used correctly captures the intended adjustment.
Despite some inadequacy in the data, there was no evidence offered that cast doubt on the need for this adjustment or supported an alternate calculation. Anthem’s proposed value of 3.1% for this adjustment is therefore approved.
5. Pooling Charge.
In the original filing, Anthem proposed a pooling charge of 8.5% to recoup the large claims in excess of $50,000 in a month that were removed from the baseline incurred claims in an earlier step. Anthem determined this charge in Exhibit V.B of the rate filing. Anthem’s approach was to calculate the pooling charge for successive rolling 12-month periods, as the percentage of claims in excess of $50,000 PMPM relative to the claims up to $50,000 PMPM. Anthem considered the average pooling point for each of the following periods:
Anthem discarded the high and low values observed above and calculated 8.5% as the unweighted average values for the remaining six periods.
Ms. Fritchen accepted Anthem’s proposed pooling charge. Ms. Niehus, in her pre-filed testimony, presented Attachment BN-1, which summarized pooled claims, unpooled claims, and the indicated pooling charge by month for the 35-month period from October 2007 – August 2010. Ms. Niehus then excluded the claims data for the three months with the highest indicated pooling charge and the three months with the lowest indicated pooling charge. The pooling percentage for the remaining 29 months was 7.4%, calculated as a weighted average for the 29 months; i.e. by dividing the 29 month total pooled claims by the 29 month total unpooled claims.
At the hearing, Anthem witness Mr. Wei presented Anthem Supplemental Exhibit 7, which he described as a correction of Ms. Niehus’s pooling charge exhibit in her pre-filed testimony. According to Mr. Wei, this exhibit demonstrated that a range of reasonable pooling charges from 8.4% to 8.9% could be derived from Ms. Niehus’s exhibit. Mr. Wei explained that the averages in his Exhibit 7 were unweighted average monthly pooling charges for the most recent 20-, 32-, 44-, and 56-month periods, with all periods ending August, 2010.
Anthem’s original proposal was based on the average of six 12-month periods which overlapped considerably. By using six overlapping 12-month periods, Anthem’s calculation is assigning the following weights to the 30 months spanning October 2007 – March 2010:
Mr. Wei’s correction of Ms. Niehus’s Attachment BN-1 in Anthem’s Supplemental Exhibit 7 was flawed by the use of unweighted averages (which suggests that a $1 million pooled claim during a month early in the calendar year, when unpooled claims tend to be low, should somehow drive a larger pooling charge than the same claim later in the year when unpooled claims are higher). Ms. Casaday defended this approach (April 12 hearing, p. 101, lines 1-8) by stating that a weighted average would inappropriately give more weight to earlier periods because of higher enrollment. However, Ms. Niehus’s exhibit did not weight by enrollment; it weighted by unpooled claims. Unpooled claims were $109.7 million during the first 24 months of Ms. Niehus’s 48 month exhibit and $105.7 million during the second 48 months. So it seems reasonable to conclude that Ms. Niehus is assigning a similar weight to the earlier periods and the later periods.
Another flaw in Anthem’s Supplemental Exhibit 7 is that it was not based on rolling 12-month periods and in fact, Anthem selected periods for calculating the unweighted average pooling charge that overweighted those months when pooling charges exhibit a seasonally high level. Mr. Wei acknowledged these flaws during the hearing (April 12 hearing, p. 147, lines 11-18).
The Superintendent also finds that that the alternative calculations proposed by Ms. Niehus are flawed. It is not reasonable to exclude the high pooled claim experience for January, 2009, as initially suggested by Ms. Niehus. The essence of insurance is to assume the risk of events with a low probability of occurrence. An insurer must be permitted to charge an adequate premium for all risks assumed, not just those above a certain minimum probability level. While pooled claims in excess of $1 million in a month are clearly unusual, it may be true that a frequency of once in a 48-month period is a reasonable expectation for a claim of this magnitude. There is a further reason why it is not reasonable to eliminate a few high and low occurrences from the probability distribution, as Ms. Niehus proposed in her Attachment BN-1. Inspection of the monthly pooling charges in Column 7 of that exhibit reveals that the distribution of values is not symmetric. The highest values are further away from the mean than the lowest values, so eliminating the highest values and an equal number of the lowest values will almost certainly lower the mean result, which is an unacceptable bias.
During the hearing, Ms. Niehus presented Attachment BN-6, which enhanced her original Attachment BN-1 to show the experience for 48 months from September 2006 – August 2010. This exhibit showed an average pooling charge for the 48 months of 8.1%. Ms. Niehus also speculated that the 37.1% pooling charge indicated in her exhibit for the month of January, 2009, could be an outlier and could be excluded from the calculation. However, she did not offer any analysis to support that speculation. Later in her testimony, Ms. Niehus indicated that her previous estimate of 7.4% was probably too low (April 13 hearing, p. 79, line 17) and that a partial, but not full, adjustment should be made for the month of January, 2009 (April 13 hearing, p. 90, line 25, to p. 91, line 2). Ms. Niehus indicated that such an adjustment might result in a pooling charge of 7.8%.
The Superintendent would approve a pooling charge of 8.1%, based on the weighted average of the 48 months of data shown in Attachment BN-6, in accordance with the evidence in the record.
6. Claim Adjustment for Enrollment Shift by Benefit.
Anthem did not refresh this calculation for this rate filing. Anthem determined that one year’s result lacked credibility and determined the proposed value for this rate filing by averaging the values from the last two filings; i.e. .964 for 2010 and .945 for 2009.
There was no evidence presented to support a different assumption, so the Superintendent approves Anthem’s proposed value of .955. However, in its next HealthChoice rate filing, Anthem should provide the calculation of this factor, done in the manner used for the 2009 and 2010 filings, for both this filing and the next filing.
B. Adjustments to Claims
In response to a CAHC information request, Anthem provided highly confidential data about historic and prospective provider cost levels by type of service.
In her pre-filed testimony, Ms. Fritchen commented on the lack of transparency and the difficulty of obtaining adequate 2010 cost information to use in her analysis. Ms. Fritchen also noted that in last year’s proceeding, Anthem asserted that a 0.6% upward adjustment for provider contracting was appropriate, but she stated that information in this year’s filing indicated that a downward adjustment of 0.7% would have been more correct to use in last year’s filing. Ms. Fritchen opined that Anthem’s methodology should be monitored over time “given the significant difference between the estimated impact included in last year’s filing and the actual results.” Given her reservations about Anthem’s approach, Ms. Fritchen recommended no provider contracting adjustment.
In her pre-filed testimony, Ms. Niehus observed that Anthem had used a longer experience period to evaluate trend, but had determined the provider contracting adjustment based on a comparison of growth rates over a much shorter time frame. Ms. Niehus testified that there should be a downward adjustment of 0.3%.
Immediately before the hearing, Anthem revised its rate proposal to set the provider contracting adjustment to 0.0% instead of 0.7%. The explanation provided in Ms. Casaday’s testimony (April 12 hearing, p. 25, lines 13-23) was that provider contracting “impacts our unit cost trend approximately 3.4 percent on average during the 30-month period we looked at for the regression methodology, and considering that we expect a similar impact to provider contracting, our unit cost trend during 2011 and into 2012, that means the incremental impact is immaterial, and that’s why we’ve removed the impact or set it to zero percent in our analysis.”
During the hearing, Ms. Fritchen reiterated her concern about the significant discrepancies in Anthem’s projections in last year’s filing and the actual experience that emerged subsequently, resulting in an overstatement of the provider contracting impact included in the rates last year. Also during the hearing, Ms. Niehus agreed to revisit her calculation because her regression was performed over a different time period than Anthem had used in their trend analysis. A revised calculation was provided by CAHC post-hearing showing a downward adjustment factor of 0.4%. Ms. Casaday testified (April 12 hearing, p.25, lines 2-5) that it was not appropriate for Ms. Niehus to apply a regression analysis to the provider contracting impacts because they are calendar year over calendar year estimates and not a time series.
The Superintendent’s consultants reviewed Ms. Niehus’ revised analysis and determined that the weights used for the projection period should be changed to use endpoint to endpoint methodology, which then requires an assumption for a 2012 provider contract trend. The annual trends should then be combined by compounding, rather than simply adding, to develop the composite trend for the rating period. If the historical average of 3.4% referenced above is assumed for 2012, the impact beyond what is implicit in trend is 0.0%. Based on this evidence, the Superintendent approves Anthem’s revised provider contracting adjustment of 0.0%. The Superintendent shares the concern expressed by Ms. Fritchen about the disparity between what was projected in last year’s filing for 2010 and what actually occurred. Anthem should review its process for projecting these effects and identify ways to improve it.
2. Cost Containment.
Anthem did not propose any specific adjustment for cost containment efforts. Both CAHC and the Attorney General posed questions about cost containment initiatives during discovery. On April 4th, Anthem responded with an exhibit that showed its estimates of the actual and projected value of cost containment efforts during 2008 – 2011.
The data in this exhibit was the basis for analysis presented by Ms. Niehus in her pre-filed testimony. Ms. Niehus opined that her analysis supported a downward adjustment of 0.5% for cost containment. Ms. Fritchen commented on the cost containment adjustment in her pre-filed testimony, but did not offer an alternative.
During the hearing, Mr. Whitmore testified about the difficulty of accurately quantifying the financial impact of cost containment initiatives, especially prospectively but also retrospectively. (April 12 hearing, p. 19, lines 4-22). Ms. Casaday disagreed with Ms. Niehus’s calculations in the prefiled testimony (April 12 hearing, pp. 20-21) and provided an alternate calculation based on the same data that suggested an impact of 0.05% (Anthem supplemental Exhibit 8). Ms. Fritchen also agreed with the difficulty of precisely quantifying these affects, although her testimony suggested that an attempt should be made to do so. (April 12 hearing, p. 232).
In closing argument, Ms. Fritchen proposed an adjustment of 0.0% although her proposal was due more to her inability to develop a more firm estimate given the deficiencies in the data provided by Anthem than a determination that 0.0% was the correct value. CAHC argued that Anthem’s responses to discovery were inadequate and formulated a proposed downward adjustment of 0.6% to trend based on Ms. Niehus’s analysis.
The Superintendent agrees that no adjustment for cost containment should be made. The Superintendent finds that accurate quantification of impacts is very difficult and a significant portion of the discussion of this issue supports the idea that the impact on trend, if it could be measured, would be minimal in either direction.
The projected savings which are the basis for Ms. Niehus’ recommendation are estimates and projections, not firm results. The estimates are often the result of an analytic process based on multiple assumptions and there is a real possibility that multiple assumptions underlying these estimates will not be achieved. Actual impacts can be very difficult to measure retrospectively, because actual claim costs would be measured against hypothetical claim costs in the absence of these cost containment initiatives, and there is no way to know what these hypothetical claim costs would have been. As Anthem points out, many of the cost containment initiatives are voluntary, which makes their impact difficult to predict.
Any successful cost savings initiatives will ultimately be reflected in claim costs and will be beneficial to HealthChoice subscribers. It would be counterproductive and potentially unfair to impose an adjustment on Anthem that was based solely on Anthem’s own projections of impacts that are very difficult to predict and also difficult to quantify retrospectively. Based on the evidence in the record the Superintendent approves Anthem’s use of projected claims with no cost containment adjustment.
Anthem reduced its projected claims per contract per month (PCPM) by $7.78 to reflect prescription drug rebates. Neither the amount nor the methodology used to calculate it was disputed. However, because the trend is used in the calculation, the adjustments to trend discussed above will reduce the amount slightly to $7.73. Similarly, there is no dispute with respect to Anthem’s $0.36 adjustment to reflect the difference between the estimated and actual rebates for the 2009-2010 rating period.
Anthem added $4.94 PCPM to claims to reflect “healthcare management fees included in benefit expense.” Anthem explained that this was previously included in administrative expenses, but has been reclassified as a benefit expense. Anthem further clarified in its April 19 post-hearing response that this change was for GAAP accounting purposes. For statutory accounting, Anthem explained that in its Supplemental Health Care Exhibit, all healthcare management expenses are reported as either quality improvement expenses or administrative expenses, neither of which is part of incurred claims. For purposes of this rate filing, these expenses should be listed as administrative expenses rather than as an adjustment to claims. Doing so would not affect the calculated rates, but would affect the calculated loss ratio. Future filings should include this item under Retention rather than under Adjustments to Claims.
Anthem estimated the cost of new state mandated benefits to be $0.81 PCPM. Anthem explained in response to discovery that the estimate was based on group business and there is no study available reflecting the impact on individual business. Because the impact is “not material,” Anthem chose not to include it.
Anthem made adjustments to reflect new benefits required by the federal Affordable Care Act (ACA). The hearing panel questioned the methodology used, both in discovery and at hearing, and Anthem provided satisfactory answers. The estimates were not disputed and the Superintendent approves Anthem’s adjustments.
The aging over time of the HealthChoice subscriber population continues to affect both claims trend and premiums. The calculated claims trend appropriately reflects the effect on claims as long as the impact of aging continues at the same rate as in the past, but an adjustment is necessary if the effect of aging changes.
With regard to premiums, because Anthem’s HealthChoice rates for any given subscriber increase as the subscriber ages, the collected revenue for the subscriber in a given benefit plan increases over time even if the underlying age-based premium rates do not. Anthem argues that the effect on premiums is appropriately accounted for in its enrollment projections. After determining required revenue for the rating period, Anthem calculates rates that will produce that revenue based on projected enrollment. If the projected enrollment reflects aging at the same rate as in the past, no further adjustment is necessary. However, the amount of aging implicit in Anthem’s enrollment projection for the future rating period is lower than it has experienced historically. So Anthem’s trend and claim projections, which are used in the calculation of required revenue, assume more aging of the subscriber base than its revenue projections based on current rates do, which leads to a larger than needed increase.
Ms. Fritchen discussed this issue in her prefiled testimony (Fritchen prefiled, p. 9, line 3), in which she developed an estimate of 0.7% as the annualized impact of aging on Anthem’s revenue collections. There was additional testimony on this issue provided by Ms. Casaday (April 12 hearing, pp. 27-29) and Ms. Fritchen (April 12 hearing, pp. 217-218 ).
During the hearing, the hearing panel requested that Anthem demonstrate that the projections reflect aging beyond what has occurred through December, 2010 (April 12 hearing, p. 174). In its May 2 post-hearing response, Anthem provided the premium factors shown below for the base period, 12 months ending September, 2010, and the rating period, 12 months ending June, 2012. In the spreadsheet provided, Anthem calculated the factors by weighting the age band factors (used to develop age-banded rates) by the contract months in each age band for the given time period.
Weighted Average Premium Factors for Base and Projection Periods:
The Superintendent’s consultants first calculated the annualized percentage change in the premium factor, due to aging, over the 21-month period from the base period to the rating period to be 0.4% as displayed in the chart above. Then the consultants determined the following past premium factors in the same manner as described above, using information Anthem provided to the Attorney General in its March 16 e-mail entitled “updated data files.”
Weighted Average Premium Factor for Historical Periods:
The average annual change over the four-year period from 2006-2010 is 0.7%. Furthermore, the year to year revenue growth has fluctuated in a narrow range during that time. These results are consistent with the results cited by Ms. Fritchen in her prefiled testimony. There is also no evidence that Anthem has made a similar assumption that the rate of aging will moderate in the future.
The Superintendent concludes that Anthem’s claim projections that are used to calculate the required revenue (calculated in Exhibit I of the rate filing) appropriately assume aging in the future at the same annual rate as in the past, 0.7%, while its revenue projections that rely on projected enrollment (calculated in Exhibit III.A of the rate filing) assume an annualized aging impact of only 0.4%. To bring the two revenue projections onto a consistent basis, an annual factor of 1.007/1.004 should be applied for 21 months to the projected revenue in Exhibit III.A. This higher projected revenue reflects the revenue that would have been generated from the rates and projected enrollment if the projected enrollment had assumed an annual 0.7% aging impact rather than only 0.4%. The Superintendent would approve rates set so that the resulting adjusted projected revenue in Exhibit III.A is approximately equal to the required revenue in Exhibit I.
D. Administrative Expense
Anthem included administrative expenses of $31.14 PCPM, excluding healthcare management fees, commissions, premium taxes, and health access payments, all of which are listed separately. As discussed above, healthcare management fees of $4.94 PCPM were listed as an adjustment to claims but should be listed under retention with the other administrative expense items. Healthcare management fees were included in administrative expenses in prior
There is no dispute that the amount is low relative both to past levels and industry norms. There is considerable dispute, however, as to whether Anthem has met its burden of proof to demonstrate that it should not be lower still. The Attorney General and his consultant, Beth Fritchen, expressed considerable frustration with their ability to obtain answers to their questions regarding how this component of the rate was determined, both during the discovery process and during the hearing.
Anthem is part of WellPoint, a large group of affiliated companies, with many administrative functions handled centrally and then allocated among the several companies and lines of business within those companies. Anthem explained that it uses a system called “Hyperion” to do this allocation. It is understandable that a system like this would be complex and difficult to understand. However, Ms. Fritchen testified that she has participated in rate reviews for many companies in other states and asked similar questions and received satisfactory explanations. It seems likely that some of those companies were also part of larger groups of companies facing allocation issues similar to Anthem.
In light of the relatively low level of the administrative expense per contract component of this filing, despite a slight decrease in the number of contracts over which to spread expenses, the Superintendent finds it very likely that the PCPM administrative expenses approved last year would not be excessive in the current rates, and the Superintendent would approve that amount for 2011 rates. However, with respect to any increase in that amount, Anthem’s inability to provide adequate answers regarding its expense allocation system results in a finding it has not met its burden of proof.
Throughout the proceeding, Anthem emphasized that its administrative expenses have decreased over time and that the proposed amount PCPM reflects no increase over actual 2010 expenses. Anthem did not point out that this level was higher than anticipated in last year’s approved rates, nor did any other party. This fact was not readily apparent in part because healthcare management fees were listed separately this year. When the hearing panel became aware of this fact, the Superintendent issued a post-hearing inquiry seeking an explanation. In its May 11 supplemental response, Anthem hypothesized that the difference was due in large part to costs associated with implementation of the ACA. While this is a plausible explanation, there is no evidence in the record to support the conclusions that administrative expenses associated with first year ACA implementation would continue at the same level during this rating period.
Anthem’s response also argued that the comparison of this year’s proposed PCPM administrative expenses to last year’s is not meaningful for three reasons. First, this year’s filing includes two additional products, HMO and HealthChoice HDHP, which were not included in last year’s filing. However, these products are only a small portion of the total enrollment and there is no evidence in the record to suggest that administrative expenses for these products would be significantly different from HealthChoice and Lumenos.
Second, Anthem argues that because 2010 actual expenses were higher than had been projected, any comparison should be to actual levels. However, as noted above, there is no evidence in the record to support that higher expense level.
Third, Anthem argues that its intent last year was to reflect actual 2009 expense levels on a PMPM basis and that this was converted to a PCPM basis using an estimate of 1.68 members per contract. According to Anthem, this year’s assumption of 1.70 members per contract would produce a higher amount. However, if the intent last year was to reflect actual 2009 levels, it should have been lower. Anthem provided a revised filing on April 8, 2010, in last year’s proceeding that substituted actual 2009 results for the estimates in the initial filing. The revision showed a higher ratio of members to contracts, 1.70, but also a larger number of members and a lower amount of administrative expense. Based on the actual 2009 levels of $4,776,000 of administrative expenses, 232,978 members, and 137,046 contracts, the administrative expenses were $34.85 PCPM. However, Anthem did not reduce its proposed administrative expense level.
For the above reasons, the Superintendent finds that an administrative expense component exceeding $35.56 PCPM would be excessive based on the evidence in the record. For future filings, Anthem should find a way to make its process more transparent. This does not mean using a simpler, easier to understand allocation system. That would likely result in a less accurate allocation. Rather, Anthem should, for example, find better ways of ascertaining and explaining why changes in various expense items occurred.
E. Rate Relativities
1. Child Rates.
The “Sponsored Dependent” rate, which is 65% of the community rate, is only available when the parent is enrolled in a group policy with Anthem and the dependent child is not eligible for coverage on the group policy; otherwise, each child age 19 or under who is not covered under a parent’s policy must pay the full “one adult” rate applicable under age 30. Anthem explained: “The discounted factor for the sponsored dependent rates has been in place for a number of years. Anthem has no specific reason to offer as to why these policyholders should pay a lower rate than other children or dependents other than the significantly lower loss ratio … for these contracts.” No loss ratio information was provided for children paying the adult rate.
The Superintendent finds no justification for charging higher rates to some children than others. All policyholders age 19 and under, unless they have dependents covered under their policies, should be eligible for the “one or more child” rate, although Anthem will not be required to cover more than one child under the same contract except to the extent it already does so. Anthem stated there are currently 36 policyholders age 19 and under paying the under-30 adult rate. Using a factor of .65 reduces rates for these 36 individuals by 18.75% (1-0.65/0.80). This will not have a material impact on Anthem’s projected revenue.
The 2009 increase in rates for Lumenos plans was capped at 20%, resulting in lower rates relative to HealthChoice than was indicated by Anthem’s analysis. As explained in the Decision and Order in that proceeding, Anthem had delayed rate action on Lumenos and the cap was intended to protect renewing policyholders from the very large indicated increases that would have resulted from Anthem’s earlier inaction.
Both the 2010 rates and the proposed 2011 rates maintain the rate relativities that resulted from the cap. Assuming the relativities indicated in the 2009 rate filing remain valid, relativity currently used to determine Lumenos rates is inappropriate and as a result, rates for other products are higher than they need to be. However, this inequity should not be accomplished in one step, as that would result in Lumenos policyholders receiving a larger rate increase than shown in the notice of proposed increase they received. The Superintendent finds that a 3% differential in the Lumenos base rate would be appropriate for the current filing.
F. Federal Income Tax
Exhibit IX of the rate filing for years 2005, 2006, and 2009 lists negative operating gains. In those years the operating gain is shown as the same before and after federal income tax (FIT). Anthem’s response to the Superintendent’s First Discovery Request on this issue was that the Operating Gain After FIT is an estimate since federal income taxes are levied at an aggregate level. Initially when questioned during the hearing Ms. Casaday said that “my assertion would be that we don’t have to pay any federal income tax on this line of business when we have losses, so it would be appropriate to apply zero under this estimated calculation of tax.” The hearing panel suggested that if taxes are levied at an aggregate level then the losses on that line would offset the gains on other lines. Ms. Casaday and Mr. Wei agreed with that statement. Correcting for the inaccuracy in Exhibit IX does not affect this year’s rates directly. For future filings, Anthem should develop a procedure to report the appropriate estimate for after-tax operating gains when the before-tax operating gains are negative, based on assigning the appropriate portion of aggregate tax offsets to the individual line of business. Unless a different methodology is shown to be appropriate, the same factor used to determine after-tax gains in positive years should be applied in negative years.
G. Litigation Costs
As in Anthem’s 2009 rate development for 2010 rates, Anthem’s 2010 rate development for 2011 rates again raises the issue of expense recovery for “litigation costs.”7 As previously explained by the Superintendent in last year’s 2010 rate decision:
Embedded in Anthem’s 2011 proposed rates is a total cost of $3,765,000 in projected administrative expenses for the rating period July 1, 2011, through June 30, 2012. Anthem represented that no Litigation Costs were included in this projection.10 Thus, according to Anthem, no adjustment was made to projected expenses during the rating period for Litigation
Costs.11 Therefore, based on Anthem’s representations, which are subject to validation on examination, the requested 2011 rates do not directly or indirectly incorporate Litigation Costs, so no corrective adjustment by the Superintendent to the proposed 2011 rates is necessary in this regard.
In 2009, the Superintendent approved a 0% risk and profit margin, as recommended by the Attorney General, based in part on a unique economic situation resulting in extreme financial hardship for subscribers and the extreme financial health of the company.13 In 2010, the Superintendent approved a .5% risk and profit margin,14 although the Attorney General had again recommended a 0% margin arguing that the conditions cited in 2009 still had not changed.15
Since 2009, the risk and profit margin issue has been thoroughly briefed and argued before the Maine courts and in administrative proceedings before the Superintendent. In a nutshell, Anthem has argued that its individual health insurance rates in this State must be designed to include a “fair and reasonable rate of return” (from Anthem’s perspective, a required positive profit margin).16 Whether and to what extent Maine law requires regulated individual health insurance rates to include a projected profit margin as Anthem maintains, the Superintendent’s determination of what is an approvable rate for a one-year period (including what, if any, built-in expected profit to provide) involves a balancing of investor and consumer interests. In other words, the amount at which to approve a built-in expected profit in regulated rates, if any, must balance the need for a rate not to threaten the company’s or enterprise’s financial integrity against the legitimate governmental interests of protecting the viability of the insurance pool, keeping insurance premiums as reasonable as possible, and minimizing adverse-selection.17 There is no bright-line test. The analysis involves a factual inquiry based on the evidence in the record at the time of the rate review.
The evidence in the record demonstrates that, on average since inception (i.e., for the 12-year period 1999 through 2010), Anthem’s individual health insurance business in this State has achieved a weighted positive profit margin of about 2.1%, or in excess of $15.5 million.
See Anthem Exhibit IX (revised 3/31/2011). While below the nationwide industry average, based on an American Academy of Actuaries policy statement,19 the table demonstrates that over the long term Maine individual policyholders have contributed positively – totaling over 2% of
There has been ongoing debate among parties to the Anthem rate proceedings surrounding the risk and profit charge since the Superintendent’s approval of 0% and .5% in 2009 and 2010 rates, respectively. As explained, the Superintendent reached these decisions following a balancing of competing interests. The Superintendent balances the competing interests again in this 2011 rate proceeding. As demonstrated above, over the 12-year period that Anthem has owned the company, the pre-tax operating gain derived from its individual insurance business in this State totaled over $15.5 million and averaged 2.1% of total revenue. Anthem’s pre-tax profit from the individual product line was 2.5% in 2010, or a gain of over $1.5 million.23 The individual policyholders have contributed positively to Anthem’s company-wide surplus, which grew from 2009 to 2010, as did Anthem’s RBC levels during this same time period. Since at least 2007, Anthem has made significant annual dividend payments to its sole shareholder from its company-wide surplus.24
As to the competing interests, the nearly 40 Anthem policyholders who provided sworn testimony universally opposed Anthem’s proposed rate increase as excessive (as did the Attorney General and CAHC).25 The policyholders testified as to the continuing financial hardships they face due to the economic downturn, the severe impacts annual rate increases have on their budgets, and their corresponding ability (or inability) to stay insured. In recognition of these factors and the protection of the public interest in these circumstances, the Superintendent aims to maintain affordable individual health insurance rates to the fullest extent possible. Specifically, some of the legitimate state interests considered by the Superintendent in conducting rate review proceedings and determining whether or to what extent to allow an explicit expected margin for risk and profit in rates include the need to keep premiums as affordable as possible, and the concern that rising rates have caused adverse selection in Anthem’s individual insurance business. Including a 3% built-in expected profit margin in rates, thereby increasing the rates even more than the average increase specified by the Superintendent in this Decision and Order, would only exacerbate this problem. The Superintendent’s decision that she would approve a 1% built-in expected risk and profit margin in 2011 rates for a single rating cycle reflects the legitimate governmental interests of protecting the viability of the individual insurance pool, keeping premiums as reasonable as possible, and minimizing adverse-selection.
While Anthem’s 3% risk and profit margin in its 2011 rate development might be appropriate under a different evidentiary record, the Superintendent finds that it would contribute to making this year’s 9.7% requested average rate increase excessive. Balancing all of the foregoing considerations and the entire record in this proceeding, the Superintendent would approve a risk and profit margin of 1% in Anthem’s rates for the one-year period July 1, 2011, through June 30, 2012. The Superintendent further finds that the 5.2% average rate increase specified by this Decision and Order, which includes a built-in expected 1% risk and profit margin, is not inadequate.26
I. Grandfathered Policyholders
Anthem explained that for certain plans that have limited enrollment in each plan design, Anthem chose to add the benefits required for non-grandfathered policyholders to all policies. Among those plans was the HealthChoice HDHP product. Anthem provided in their post-hearing response notices mailed to HealthChoice HDHP policyholders regarding grandfathered status. Notices mailed October, 2010, and January, 2011, as well as a contract amendment explained the grandfathered status of the policyholder’s plan. Ms. Casaday explained at hearing that grandfathered policyholders retained their grandfathered status despite having these benefits added. However, the notification of the proposed rate increase included rate sheets provided in the rate filing and by Anthem in their response that stated: “Attachment B: Rates per Contract by Benefit Option for Non-Grandfathered Options.” This caused confusion to at least one policyholder, who testified about this issue. Anthem is directed to correctly identify future rate sheets, including those sent to HealthChoice HDHP policyholders notifying them of the final 2011 rates. An appropriate label would be, “Rates per Contract by Benefit Option for Grandfathered and Non-Grandfathered Plans.”
V. FINDINGS AND CONCLUSIONS
On the basis of a preponderance of the credible evidence in the record, and for reasons set forth in Section IV above, the Superintendent finds and concludes that Anthem’s proposed rates are excessive and unfairly discriminatory. If the changes to the rates proposed by Anthem are applied consistent with this Decision and Order, as discussed in Section IV, the Superintendent could lawfully approve the resulting rates. The necessary revisions to the proposed rates can be achieved by the steps detailed in Attachment A. The rates resulting from these changes are shown in Attachments B1, B2, and B3.
The Superintendent finds and concludes that such revised rates, appropriately developed per this Decision and Order, would not be excessive, inadequate, or unfairly discriminatory; and would likely yield a loss ratio of at least 65%.
As a result of the changes specified by the Superintendent in this Decision and Order, the total average rate increase proposed in Anthem’s filing of 9.7% would be reduced to 5.2%. For the Mandated HealthChoice options, there would be no rate change. For the Non-Mandated HealthChoice, HealthChoice HDHP, HMO Standard, HMO Basic, and Lumenos Consumer Directed Health Plan options, the average increase would be 5.3%, with the specific rate increases ranging from 3.1% to 13.2%.
Pursuant to the provisions of 24-A M.R.S.A. §§ 12-A(6), 2736, 2736-A, 2736-B and authority otherwise conferred by law, the Superintendent hereby ORDERS:
VII. NOTICE of APPEAL RIGHTS
This Decision and Order is final agency action of the Superintendent of Insurance, within the meaning of the Maine Administrative Procedure Act, 5 M.R.S.A. § 8002(4). It may be appealed to the Superior Court in the manner provided for by 24-A M.R.S.A. § 236, 5 M.R.S.A. §§ 11001 through 11008, and M.R. Civ.P. 80C. Any party to the proceeding may initiate an appeal within thirty days after receiving this notice. Any aggrieved non-party whose interests are substantially and directly affected by this Decision and Order may initiate an appeal within forty days after the issuance of this Decision and Order. There is no automatic stay pending appeal. Application for stay may be made in the manner provided in 5 M.R.S.A. § 11004.
PER ORDER OF THE SUPERINTENDENT OF INSURANCE
1 These comments appear in the transcript and are part of the record of this proceeding. The sworn comments have been admitted into evidence pursuant to 5 M.R.S.A. § 9057(3). Unsworn comments, if any, shall be considered for their persuasive value to the extent that they are relevant to facts in the record.
2 Following a brief executive session, it was ultimately determined by Anthem and agreed to by the Superintendent that only one phrase of Anthem’s witness testimony was confidential and in need of redaction from the public transcript.
3 By agreement of the parties, Mr. Kirsch testified and was examined telephonically.
6 As explained in each MD&A, the regulatory filing “provides management’s assessment of the current financial position, results of operations, changes in capital and surplus, and cash flow and liquidity for Anthem Health Plans of Maine, Inc.” The MD&As summarize and supplement Anthem’s Annual Statements. Anthem’s 2010 Annual Statement is already a part of the record, and the 2010 MD&A is a supplement to that Annual Statement. As to Anthem’s 2007 through 2009 MD&As, these regulatory filings provide historic backdrop. Portions of the 2007 MD&A were a part of the record in the 2009 rate proceeding.
7 In an April 9, 2010, information request in the 2010 Anthem rate proceeding, the Superintendent defined the term “Litigation Costs” as “the total costs, actual and/or estimated, ... associated with or related to Anthem’s appeal of the Superintendent’s 2009 Decision and Order (INS-09-1000).”
8 In its Amended Petition for Review of Final Agency Action filed with the Superior Court, Anthem’s request for relief was that the Superintendent’s 2009 rate decision be vacated and remanded to approve rates with a pre-tax profit and risk charge of 3% (as opposed to the 0% approved by the Superintendent). The Superior Court affirmed the Superintendent’s decision and the Law Court dismissed Anthem’s appeal as moot. See footnote 13.
10 The Superintendent explained and inquired “Projected administrative expenses for the second half of 2011 total $2,071,000. How much of this amount consists of Litigation Costs?” Superintendent Information Request 1-37(d). Anthem responded “$0.00 PMPM.” The Superintendent explained and inquired “Projected administrative expenses for the first half of 2012 total $1,694,000. How much of this amount consists of Litigation Costs?” Id. 1-37(e). Anthem responded “$0.00 PMPM.”
11 Anthem represented, however, that actual administrative expenses for the individual business in 2010 totaled $4,244,000, of which approximately $4,300 were Litigation Costs. Response to Superintendent Information Request 1-37(a).
At hearing, Anthem’s counsel stipulated that “clearly the litigation costs were greater than $4,300” and the Company’s witness testified that certain of the Litigation Costs were allocated elsewhere (outside of the Maine individual insurance business) because “Anthem as a corporation viewed [the 2009] rate case as being something which was important not to Anthem but to the entire industry. For an insurance superintendent to suggest that an insurance carrier is not entitled to any risk or profit included within their rates, [Anthem] didn’t look at that as just a Maine individual rate line of business [issue]. … So [the Litigation Costs weren’t] all allocated to [the Maine individual insurance business] members.” (April 12 hearing, p. 91, lines 8-18; see also pp. 130-132.)
12 With this proposed built-in projected 3% profit margin for the second half of 2011 (July 1, 2011, through January 30, 2011) in conjunction with the currently approved built-in projected .5% profit margin for the first half of 2011 (January 1, 2011, through June 30, 2011), Anthem determined that the rates would produce calendar year 2011 projected pre-tax profit of 6.4%, or $4,387,000. Anthem Exhibit IX (1/28/11, revised 2/17/11). For the one-year rating period (July 1, 2011, through June 30, 2012) with a built-in projected 3% profit margin, Anthem projected pre-tax profit to be $2,042,000. Id.
13 Anthem appealed the Superintendent’s 2009 rate decision to the Superior Court and the Supreme Judicial Court. The Superior Court affirmed the Superintendent’s decision, upholding a 0% expected profit margin in rates to be in compliance with statutory and constitutional law. Anthem Health Plans of Maine, Inc. v. Superintendent of Insurance, BCD-WB-AP-09-36 (April 21, 2010) (Humphrey, C.J.). The Supreme Judicial Court sitting as the Law Court dismissed the appeal as moot. Anthem Health Plans of Maine, Inc. v. Superintendent of Insurance, 2011 ME 48 (April 21, 2011). On May 4, 2011, Anthem moved for reconsideration by the Law Court pursuant to M.R. App. P. 14(b)(1).
14 Anthem appealed the Superintendent’s 2010 rate decision to the Superior Court, which matter was stayed upon the motion of Anthem, over the Superintendent’s objection, pending a final ruling by the Law Court in the 2009 appeal. The record shows that with an approved .5% built-in expected profit margin in 2010 rates, Anthem’s actual pre-tax profit was 2.5%, or $1,542,000. Anthem Exhibit IX (revised 3/31/2011).
15 The Superintendent agreed that the conditions cited by the Attorney General in 2009 still existed in 2010, as fully supported by the evidence in the record. While these were among the factors supporting a one-time 0% risk and profit margin in 2009, the Superintendent explained in the 2010 rate decision that it did not necessarily follow that a 0% margin is appropriate on a long term basis.
16 Anthem’s enunciated profit standard is nowhere articulated in Maine’s individual health insurance laws. See 24-A M.R.S.A. Ch. 33. The requirement under the Maine Insurance Code is that individual health insurance rates shall “not be excessive, inadequate or unfairly discriminatory.” Id. at § 2736(2). The “not inadequate” standard under section 2736(2) is not defined by statute. Cf. section 2382(3) defining “inadequate rates” for purposes of workers’ compensation insurance (“A rate is not inadequate unless insufficient to sustain projected losses and expenses and the use of the rate has had a tendency to create a monopoly or, if continued, will tend to create a monopoly in the market or will cause serious financial harm to the insurer.”) In a Bulletin issued in 2001, Superintendent Iuppa explained that “Maine law requires the Superintendent to allow [a proposed individual health insurance rate] increase if it is found to be adequate to pay anticipated claims and is found not to be excessive or unfairly discriminatory.” Maine Insurance Bulletin 311, Individual Medical Insurance: Notice of Rate Increase and Right to Request Hearing, dated September 28, 2001. More globally, in December, 2005, the Actuarial Standards Board adopted Actuarial Standards of Practice (ASOP) No. 8 entitled “Regulatory Filings for Health Plan Entities.” CAHC Exhibit BN-7. ASOP No. 8 provides as a “recommended practice” that when preparing a health insurance rate filing “the actuary should consider which assumptions are necessary for the filing” which “may include . . . expected financial results, such as profit margin, surplus contribution, and surplus level.” Id. pp. 3-4, 3.2.2(g) (emphasis added). The Board’s actuarial standard does not mandate that individual health insurance rates shall include a built-in risk and profit margin, it is only a rating assumption which the actuary “should consider” for a health insurance rate filing. Id.
17 As found by the Superintendent in 2009, the amount of built-in expected profit in the short term (i.e., a single rating cycle) can be “0%”, and the Superior Court affirmed that determination.
18 A significant portion of the losses in 2009 were due to the Superintendent’s disallowance of certain cost recovery in rates, which Anthem did not challenge, and which were the result of Anthem’s own action or inaction. (See 2009 Decision and Order where the Superintendent disallowed in rates what was projected to be just under $1 million). Similarly, the 2010 gains would have been larger but for the Superintendent’s $1 million disallowance (the disallowance spanned two calendar years, 2010 and 2011). If one properly accounts for the projected $1 million disallowance, Anthem’s 12-year accumulated profits from its individual health insurance business would have been nearly 2.3%, or over $16.5 million.
19 See Anthem Exhibit 6 (March 2010, “Premium Setting in the Individual Market”). The Academy policy statement explains at p. 1:
Anthem’s experience illustrates this point. In the three years where Anthem incurred a loss on its individual insurance business, surplus was available – funded in part by the nine years of profit from its individual insurance business – to cover those losses.
20 Anthem Health Plans of Maine, Inc. (doing business as Anthem Blue Cross Blue Shield) is a single corporate entity that maintains one surplus for all of its operations. (See April 12 hearing, p. 62, lines 15-19. In response to a question asking whether Anthem maintains a common surplus for all lines of business, or segregates surplus by line of business, Anthem’s witness responded: “I’m not aware that [Anthem] segregate[s] surplus by line of business.”) There is no separate individual business surplus, no separate group business surplus, etc. In other words, all of Anthem’s individual insurance policyholders contribute to a common surplus that Anthem holds. (Id. at p. 69, lines 13-19.) All income earned from all lines of business goes into that surplus. That consolidated, company-wide surplus is available both to meet all financial obligations of the corporation, including all insurance claims from all lines of business, and to pay shareholder dividends to the parent corporation.
23 For the period January, 2011, through June, 2011, the 6-month projected pre-tax profit margin is 21.5%, which is in excess of $6.7 million (derived from Anthem Exhibit IX, using current rates and available data as of February, 2011).
24 It is noted that A.M. Best Company rated Anthem Health Plans of Maine, Inc. “A” (Excellent) as of December 31, 2007, “reflect[ing] the agency’s opinion as to the Company’s financial strength, operating performance and ability to meet [its] claim obligations.” A.M. Best assigned the “A” (Excellent) rating to Anthem following two years in which the Company lost money on its individual insurance business (2005 and 2006). Anthem’s A.M. Best ratings have remained “A” (Excellent) for every year since, including the period ending December 31, 2010. This period included a year with a loss (2009) and years with 0% and .5% expected profit margins built in to rates. The A.M. Best ratings demonstrate that the Superintendent’s 2009 and 2010 rate decisions have not adversely impacted Anthem’s financial strength, operating performance, or ability to meet its claim obligations.
25 The Attorney General opposed the level of Anthem’s requested rate increase and argued that the Superintendent should limit the expected profit margin to the 0.5% reflected in current rates. CAHC also opposed the level of the overall increase and recommended no built-in expected margin for 2011 rates.
26 Based on Anthem’s Exhibit IX, revised to reflect rating modifications pursuant to this Decision and Order, the projected pre-tax profit margin for 2011 is 5.7%, or nearly $4 million (with the 1% expected margin built-in to rates). If one includes the projected 5.7% pre-tax profit for 2011 into the cumulative analysis, Anthem’s 13-year pre-tax gain would be nearly 2.6%, or over $20 million, for its individual health insurance business.
All changes below are intended to be applied to the Excel file Anthem provided in a response to the AG on
Last Updated: August 22, 2012
|Copyright © 2006 All rights reserved.|