ALJDEC decisions subject to certification as final

18F-007-ARB · Arizona State Retirement System · 2018-05-11

IN THE OFFICE OF ADMINISTRATIVE HEARINGS

Julie Schweigert,

Appellant,

v.

Arizona State Retirement System,

Respondent.

No. 18F-007-ARB

ADMINISTRATIVE LAW JUDGE DECISION

HEARING: April 20, 2018, at 8:00 a.m.; the record was held open until May 4, 2018, to allow the Administrative Law Judge to have the benefit of the court reporter’s transcript in making her recommendation to the Arizona State Retirement System’s Board’s Appeals Committee.

APPEARANCES: Julie Schweigert (“Appellant”) appeared on her own behalf; the Arizona Statement Retirement System (“ASRS”) was represented by Jothi Beljan, Esq., Assistant Attorney General.

ADMINISTRATIVE LAW JUDGE: Diane Mihalsky

_____________________________________________________________________

FINDINGS OF FACT

Appellant is a member of ASRS. After Appellant retired, she appealed ASRS’ determination that her average monthly compensation for purposes of calculating her retirement benefit was $8,748.20 on her retirement date of July 17, 2017.

ASRS referred the matter to the Office of Administrative Hearings, an independent state agency, for an evidentiary hearing. A hearing was held on April 20, 2018. Appellant testified on her own behalf and submitted sixteen exhibits. ASRS presented the testimony of Member Advocate Jenna Golab and submitted 20 exhibits.

Applicable Statutes

A.R.S. § 38-757 is entitled “Normal Retirement” and provides in relevant part as follows:

A. After application on a form prescribed by the director, a member may retire on reaching the member's normal retirement date. . . .

B. [A] member who meets the requirements for retirement benefits at normal retirement shall receive a monthly life annuity that equals the result of paragraph 1 of this subsection multiplied by paragraph 2 of this subsection when those paragraphs are defined as follows:

1. The number of whole and fractional years of credited service times the following:

. . . .

(c) 2.20 per cent if the member has at least 25.00 years of credited service but not more than 29.99 years of credited service.

(d) 2.30 per cent if the member has at least 30.00 years of credited service.

2. The member's average monthly compensation.

A.R.S. § 38-758 is entitled “Early Retirement.” A.R.S. § 38-758(A) provides that “[a] member who has attained age fifty and who has five years of total credited service is eligible to elect, in a form and manner prescribed by the board, to receive a reduced retirement income.”

A.R.S. § 38-711 defines the terms “average monthly compensation,” “early retirement,” “normal retirement age,” and “normal retirement date” in relevant part as follows:

5. "Average monthly compensation" means:

. . . .

(b) For a member whose membership in ASRS commenced on or after January 1, 1984 but before July 1, 2011, the monthly average of compensation on which contributions were remitted during a period of thirty-six consecutive months during which a member receives the highest compensation within the last one hundred twenty months of credited service. . . .

. . . .

11. "Early retirement" means retirement before a member's normal retirement date after five years of total credited service and attainment of age fifty.

. . . .

27. "Normal retirement date" means the earliest of the following:

(a) For a member whose membership commenced before July 1, 2011:

. . . .

(iii) The first day that the sum of a member's age and years of total credited service equals eighty.

A.R.S. § 38-715(C) requires ASRS’ director to “[a]dminister this article” and to “[p]rescribe procedures to be followed by members and their beneficiaries in filing applications for benefits.” A.R.S. § 38-715(D)(4) allows ASRS’ director to “[m]ake retirement under this article effective retroactively to on or after the day following the date employment is terminated if the member was unable to apply before the retroactive effective date through no fault of the member.”

A.R.S. § 38-755(A) requires ASRS to “make information concerning a member's account accessible to the member in written or electronic form. This information shall include the member's current account balance, contact information, beneficiary election, estimated retirement date and estimated benefit amount.”

Hearing Evidence

Appellant is an accountant. Appellant started working as a Staff Auditor I for the State of Arizona Auditor General, an ASRS employer, on June 6, 1988.

Between July 2007 and December 2009, Appellant was employed by the Maricopa County Stadium District, an ASRS employer. Appellant worked full-time and was well compensated during the time she worked for the Maricopa County Stadium District.

After December 2009, Appellant started a family. Appellant was employed by the Scottsdale Unified School District (“SUSD”), another ASRS employer. Appellant chose to work part-time to have more time for her children, her husband, and other family responsibilities.

Because Appellant was well-compensated while she worked full-time for the Maricopa County Stadium District, she testified that she planned to retire on a date that would maximize her lifetime retirement benefit by having her benefit calculated using the average monthly compensation under A.R.S. § 38-711(5)(b) based on her higher earnings at the Maricopa County Stadium District between July 2007 and December 2009, before she started working part-time.

Appellant stopped working for SUSD after the 2015/2016 school year because she planned to retire when she turned 50 in September 2016.

Ms. Golab testified that a member’s retirement annuity payment depends on the member’s age and years of service, which determines her eligibility for retirement as well as the correct multiplier, and average compensation as defined in A.R.S. § 38-711 as reported by her employers. Ms. Golab testified that some members work longer to obtain a higher multiplier and that for most members, their highest salary for purposes of calculating average monthly compensation will be at the end of their state service.

On May 31, 2016, at Appellant’s request, ASRS sent her an estimated calculation of her retirement benefits if she retired on September 15, 2016. The estimate was based on an average monthly salary of $9,053.011, 29.21 years of service, and a 2.20 multiplier, and an age at retirement of 50 years. The estimate calculated that Appellant would have 79.21 points at retirement and would receive a straight-life annuity in the amount of $5,680.41 or a 10-year certain life annuity of $5,624.23. The bottom of the May 31, 2016 estimate noted that “[t]his estimate reflects and EARLY retirement, which means a permanently reduced benefit.” The estimate included the following disclaimer in bold capital letters at the top of the page: “*** THIS IS AN ESTIMATE ONLY, USING UNAUDITED DATA—NOT AN INCOME COMMITMENT***.”

Ms. Golab testified that ASRS generates 400-800 estimates each month. Ms. Golab testified that if ASRS were required to audit the information included in each estimate, it would need a much larger staff than the six persons it currently employs to provide benefit estimates.

Ms. Golab testified that estimates are computer-generated based on the information on record for a member. For programming convenience, the average monthly salary on the estimate is based on the highest salary fiscal year, which runs from July 1 to January 30, not individual 36 months. However, staff is able to manually adjust or manipulate the average monthly salary. Because most members’ average annual salary remains steady or increases over time, such adjustment is not always made. Ms. Golab testified that ASRS has only 3 members who, like Appellant, had significantly higher earnings early in their careers. There is no requirement that staff manipulate the data used to calculate average monthly salary to achieve the most accurate amount in individual cases.

Ms. Golab testified that the May 31, 2016 estimate was incorrect because it assumed that Appellant was still actively employed and that she would earn service credit for both August and September 2016. The error likely was not caught due to the proximity of the date on which Appellant terminated her employment with SUSD. Ms. Golab testified that the computerized estimator typically does not stop projecting service until four to six weeks after ASRS fails to receive a contribution from the employer.

Appellant requested another estimate. On June 6, 2016, ASRS sent another calculation that showed that if Appellant retired on June 6, 2016, she would be 50 years old, would have 28.99 years of service, and would have an average monthly salary of $9,982.98. The June 6, 2016 estimate showed a straight-life annuity monthly payment in the amount of $6,174.03 and a 10-year certain life annuity monthly payment of $6,112.97. The bottom of the June 6, 2016 estimate noted that “[t]his estimate reflects and EARLY retirement, which means a permanently reduced benefit.” The June 6, 2016 estimate included the following disclaimer in bold capital letters at the top of the page: “*** THIS IS AN ESTIMATE ONLY, USING UNAUDITED DATA—NOT AN INCOME COMMITMENT***.”

On June 21, 2016, ASRS sent another calculation that corrected the June 6, 2016 estimate to show that if Appellant retired on April 9, 2016, she would have 28.99 years of service, her average monthly salary would be $9,563.42, and she would receive a straight-life annuity monthly payment of $5,914.55 or a 10-year certain life annuity of $5,889.83. The June 21, 2016 estimate also included a note that provided in relevant part as follows:

This is a corrected estimate. Please disregard the estimate(s) dated 06-06/2016. This estimate reflects an EARLY retirement, which means a permanently reduced benefit. You can work as late as Nov, 2016 and still get the highest months in. The final pay period must be in Nov, not Dec, or a high month will drop off.

The June 21, 2016 estimate included the following disclaimer in bold capital letters at the top of the page: “*** THIS IS AN ESTIMATE ONLY, USING UNAUDITED DATA—NOT AN INCOME COMMITMENT***.

Because Appellant questioned the statements on the May 31, 2016 and June 21, 2016 estimates that if she retired on June 6, 2016 or September 15, 2016 she would receive a permanently reduced retirement benefit due to her early retirement, Appellant contacted Member Services benefit advisor Frances Dyer. On September 13, 2016, Ms. Dyer sent an email to Appellant that stated in relevant part as follows:

This is in response to your email received earlier last week. . . . I am the subject matter expert on the benefit estimates so your email was referred to me.

One of the problems we have with an automated benefit estimator is due to employer contribution reporting and pay cycles. With nearly 700 employers that each have their own accounting and payroll systems it is impossible to fully automate the benefit estimate. Schools sometimes pay year round, and sometimes for only a school year cycle which makes it extra complex. In your case, the biggest problem is related to your dual employment and changing from a non-school employer to a school employer. This makes estimating future retirement benefits especially challenging and reduces the accuracy of the estimates. Because we are providing estimates using unaudited data, the results are subject to change. When you apply for retirement a careful and more precise calculation will be done.

I have attached a spreadsheet showing the salary we used to estimate your average monthly compensation if you retire September 16, 2016. I highlighted the highest 36 consecutive months (in blue) to validate the AMS of $9,563.42 provided on your most recent benefit estimate (dated June 21, 2016). We do not use fiscal years as a whole when calculating salary. We use 36 consecutive months of salary in the last 120 months, as shown on the worksheet.

Ms. Dyer attached to the email a spreadsheet showing the salaries that Appellant’s employers had reported to ASRS between July 2007, and June 2016 and highlighting the 36 consecutive months between December 2007and November 2010, in which Appellant had earned the highest reported salary.

On September 12, 2016, Appellant filed an appeal to ASRS to question the advice and estimates that she had received, in relevant part as follows:

I knew that my retirement benefits were based on my highest 36 months within my last 120 months of service. Based in this I called ASRS, during the spring of 2016, to find out the best date to separate my employment in order to maximize my benefits. I called on 3 more occasions and was told several times that if I continued to work past June 2016 my retirement benefits would decrease. . . . Based on this information and knowing that the last month of credited service for a school district employee at SUSD is April, I resigned from my position with Scottsdale Unified School District in April 2016.

Based on ASRS’ employees’ alleged poor advice, Appellant requested to purchase the service that she could have earned at SUSD between August and November 2016, thereby making her eligible for a normal retirement and increasing her monthly retirement benefit. Appellant also requested that ASRS increase her record of the compensation that she received from Maricopa County in December 2009 based on its alleged failure to credit family medical leave taken during that month.

On September 21, 2016, Dave King, ASRS’ Assistant Director for the Member Services Division, sent a letter to Appellant, denying her requested relief because no statute allowed Appellant to purchase future service credit or to correct the compensation that Maricopa County had reported for December 2009. Mr. King informed Appellant in relevant part as follows:

Essentially, [A.R.S. § 38-736(B)] dictates only the employer can send in contributions for a specific time period and the ASRS has no authority to amend the contributions and apply them to a different pay period than specified by the employer. As a result, we cannot move a contribution received for January 2010 to be included with December 2009 to change your highest 36-month period, even if you believe the family medical leave should have been paid on December 27, 2009. Based on the contributions history, we maintain your projected Average Monthly Compensation 36-month period is December 2006 through November 2009. We have included a more detailed breakdown of your estimated, unaudited salary for the past ten years for your review.

While we realize this is not entirely the outcome you sought through your request, we sincerely hope you understand the reason why we cannot provide specific salary calculations on an estimated amount and that you have a better understanding of the statutes that define how compensation is determined in calculating your pension annuity.

A formal appeal process exists for members who do not agree with staff decisions. However, you have not yet retired and therefore no formal action has been taken by the ASRS against your legal rights. As such, no agency appealable action has occurred in order to continue the appeal process at this time. . . .”

Ms. Golab testified that the ASRS included the spreadsheet of Appellant’s salary history due to her unique salary situation, so that she could see if the months of high past salary would drop off the 120-month period used to determine her average monthly salary high past salary if she decided to go back to work to accrue additional years of service to reach normal retirement.

On September 26, 2016, Appellant returned to work for SUSD in a different position.

On October 23, 2016, Appellant contacted Ms. Dyer to verify that Maricopa County had corrected her December 2009 compensation as it informed her that it would do. Ms. Dyer confirmed that Maricopa County had made the adjustment and requested that ASRS correct its records. Appellant testified that Ms. Dyer informed her that after the correction, she would reach 80 points for a full retirement in January 2017.

Ms. Golab testified that a member must work and contribute in a month to earn service credit for that month, so that if a member was employed at the end of the month, her pay and contribution would be posted in the following month. The flip side would be that if a member terminated her employment at the end or maybe the beginning of a month, her pay would get posted in the next month and she would receive service credit for that month. Ms. Golab testified that ASRS’ correction of Appellant’s December 2007 salary from Maricopa County Stadium District did not change her highest 36 consecutive months of salary in the 120 months before her retirement.

At Appellant’s request, on November 8, 2016, ASRS provided an estimate for her retirement benefit if she retired on December 30, 2016, when she would be 50.29 years old with 29.32 years of service, for a total of 79.61 points. The estimate provided an average monthly salary of $9,572.36 and a monthly straight-life annuity payment of $6,102.31 or a 10-year certain life annuity monthly payment of $6,041.96. The estimate contained the note that the estimate reflected an early retirement, which meant a permanently reduced benefit, and the disclaimer in in bold capital letters at the top of the page: “*** THIS IS AN ESTIMATE ONLY, USING UNAUDITED DATA—NOT AN INCOME COMMITMENT***.”

Ms. Golab testified that the November 8, 2016 estimate with a retirement date of December 30, 2016, accurately projected 29.32 years of service and a $9,572.36 average monthly salary.

On November 8, 2016, ASRS also provided an estimate that if Appellant retired on February 28, 2017, she would be 50.45 years old and have 29.55 years of credited services, for the 80 points required for normal retirement. The estimate stated that Appellant’s average monthly salary was $9,572.36, her estimated monthly straight-life annuity payment would be $6,222.99 or her estimated 10-year certain life annuity payment would be $6,161.44. The estimate contained the note that it reflected “a NORMAL retirement which means an unreduced benefit” and disclaimer in in bold capital letters at the top of the page: “*** THIS IS AN ESTIMATE ONLY, USING UNAUDITED DATA—NOT AN INCOME COMMITMENT***.”

Ms. Golab that the estimate for a February 28, 2017 retirement date correctly noted that Appellant would be eligible for a normal retirement on that date. However, staff who completed the estimate did not appropriately adjust the average monthly salary to account for the additional months that Appellant would have worked on March 1, 2017 in the 120 month history.

Appellant testified that on November 8, 2016, Ms. Dyer called her and apologized about providing incorrect information because Appellant had not received any service credit for September 2016, because the pay period ended in October 2016. Appellant testified that Ms. Dyer told her that she would receive the highest monthly retirement benefit if she worked until July 2017, because she would then qualify for a multiplier of 2.3, rather than 2.2, and would have reached 30 years with ASRS employers. Appellant played into the record Ms. Dyer’s explanation of why she thought Appellant would be entitled to higher retirement benefits if she reached 30 years with an ASRS employer, even though some of her early higher salary months would drop off the 120 months used to calculate average monthly salary:

MS. DYER: Well, so, yes, part of that higher salary is dropping off, but it’s not like you lose a whole year of it at a time. You only drop off a month at a time. So like every month you go forward, it only – it only changes the salary by that month, you know, not -- not like the entire year. So, you know, so in this particular case, even though your average monthly salary is slightly lower, it’s not horrendously lower, but the average monthly salary was lower, but because your service is higher, and even most importantly, because you’re now at that 2.3 percent multiplier, that’s why, you know, working, you know, just a few more months and getting it to that higher multiplier, it – it makes that much difference. It – it makes a higher difference than if you retire earlier and keep the higher salary months.

MS. SCHWEIGERT: That’s interesting, because my salary went from like $9,000 a month to $2,000 a month.

MS. DYER: Well, yeah, but like I said, you’re not – you’re not losing that entire year salary. It’s only – you know, it’s only dropping a month at a time, and it’s not – and it’s not picking up, you know, what you’re working right this minute, either. It’s still – there was still some gradual changes . . . .

Ms. Dyer promised to send a new estimate to Appellant that would show the benefit of waiting until July 2017, to retire, which Ms. Dyer stated would result in an average monthly salary that was $250.00 lower, but a significantly different monthly retirement payment. Appellant testified that she disagreed with Ms. Dyer because if she waited to retire until July 2017, some of her higher earning months would drop out of the ten-year prior period.

At Appellant’s request, on November 21, 2016, ASRS provided an estimate of her retirement benefit if she retired on June 30, 2017, when she would be 50.79 years old with 30.1 years of service. The estimate provided an average monthly salary of $9,328.22 and a straight-life annuity payment of $6,457.93 or a 10-year certain life annuity payment of $6,394.06. The estimate contained the disclaimer in in bold capital letters at the top of the page: “*** THIS IS AN ESTIMATE ONLY, USING UNAUDITED DATA—NOT AN INCOME COMMITMENT***.”

Ms. Golab testified that the estimate for the July 1, 2017 retirement date included an average monthly salary that was lower than the projected average monthly salary for the March 1, 2017 retirement date, but that that the average monthly salary still included months that were outside the 120 months that could be considered.

On March 27, 2017, Ms. Dyer confirmed in an email to Appellant that she needed to work until July 2017, to get credit for 30 years of service and a higher multiplier. On April 11, 2017, Ms. Dyer advised Appellant to work as long as she could, up until the end of July 2017, to get 30 years of service.

On April 11, 2017, at Appellant’s request, ASRS provided an estimate of her retirement benefit if she retired on May 1, 2017, that provided an average monthly salary of $9,328.22 and a monthly straight life annuity payment of $6,132.00 or a 10-year life annuity payment of $6,070.31. The estimate contained the disclaimer in in bold capital letters at the top of the page: “*** THIS IS AN ESTIMATE ONLY, USING UNAUDITED DATA—NOT AN INCOME COMMITMENT***.

On April 11, 2017, ASRS also provided an estimate of Appellant’s retirement benefit if she retired on June 1, 2017, which provided an average monthly salary of $9,328.22, and a monthly straight-life annuity of $6,154.57 or a monthly 10-year life annuity payment of $6,092.66. The estimate contained the disclaimer in in bold capital letters at the top of the page: “*** THIS IS AN ESTIMATE ONLY, USING UNAUDITED DATA—NOT AN INCOME COMMITMENT***.”

On June 30, 2017, Appellant submitted an online request to ASRS to retire on July 17, 2017. The online request elected to receive a 10-year life annuity. The average monthly salary generated by the only retirement application was $9,053.00 and the estimated monthly pension amount was $5,911.00. The retirement estimate included the disclaimers “*DOLLAR AMOUNTS SHOWN HERE ARE GROSS AMOUNTS, NOT INCLUDING ANY APPLICABLE FEDERAL AND STATE TAXES” and “THIS IS AN ESTIMATE ONLY, USING UNAUDITED DATA – NOT AN INCOME COMMITMENT.”

Ms. Golab testified that the lower average monthly salary on the retirement application estimate was due to the computer system completely dropping off the 2007 fiscal year from the 120-month period.

On July 17, 2017, Appellant retired from ASRS employment. On August 31, 2017, ASRS informed Appellant that the average monthly salary that it used to calculate her benefits was $8,679.26 and that, based on her average monthly salary, 30.1 years of service, and the 2.3 multiplier, her monthly annuity retirement benefit would be $5,946.10..

On October 13, 2017, Ms. Dyer sent an email to Appellant, that provided in relevant part as follows:

I am still working with the benefits accounting but want to give you an update.

Your pension salary has been increased to $8748.20 which will increase your gross benefit to $5993.33.

I have also been doing additional calculations to see if retiring earlier would make a difference and I believe it will make a significant difference and that you could [have] retired as early as March and your benefit would be significantly higher. Before I give you numbers I am asking someone to verify my calculations. Since I made such a large mistake before I don’t want to repeat that.

On October 16, 2017, ASRS informed Appellant that after a routine post-retirement audit, it had determined that her monthly annuity retirement benefit would be $5,993.33.

On November 27, 2017, Appellant filed an appeal on ASRS’ calculation of her retirement monthly benefit and Ms. Dyer’s poor advice.

On December 5, 2017, ASRS Assistant Director Dave King denied Appellant’s appeal in a letter that provided in relevant part as follows:

The average monthly salary is derived from the highest 36 consecutive months of salary in the past ten years of contributing service. In your specific case, with your elected retirement date of July 17, 2017, the highest 36 months are August 2007 through July 2010 and your ten-year timeframe, as outlined in statute, is August 2007 through July 2017. This timeframe is primarily comprised of older, 12-month employment salary from Maricopa County with the last several months of your 36-month period also included a 9-month employment salary from [SUSD]. The estimates you received in November 2016 and April 2017, did not drop off the higher salary pay periods spanning from December 2006 through July 2007. Specifically, the November 2016 benefit estimate, assuming a July 1, 2017 retirement date, needed in order to reach 30 years and the higher multiplier, included May 2007 through July 2007, which are outside the 10-year window required by law with a July 2017 retirement date. The ASRS staff who prepared these estimates did not realize this error and incorrectly used an estimated 36-month salary range with older, ineligible salary. Due to this inclusion, your estimates from November 2016 forward were overstated as it included ineligible salary in the average monthly salary calculation.

We sincerely regret that the benefit estimates contained an overstated average monthly salary, which resulted in an overstatement of your final gross monthly retirement benefit versus the estimates provided. Though you did previously appeal salary information in September 2016, you were still actively contributing and your salary was never audited during that review of a June 2016 benefit estimate and denial to purchase future unearned service credit. The response actually indicated your account was unaudited and therefore we could [not] provide an exact computation for your average monthly salary. Since benefit estimates are produced using unaudited data prior to retirement, the ASRS is always concerned with members relying on a specific dollar amount as their monthly benefit. That is what we clearly publish the following disclaimer in bold-face type on each benefit estimate we prepare: “THIS IS AN ESTIMATE ONLY, USING UNAUDITED DATA—NOT ANY INCOME COMMITMENT.”

On February 1, 2018, Appellant appealed Mr. King’s determination to ASRS’ Director because “[t]his difference of $400-$500 per month is not because of the use of unaudited data as stated in the disclaimer. This difference was due to errors on the part of ASRS staff whose job it is to provide, and upon which we rely, accurate information.”

On February 14, 2018, ASRS’ Deputy Director and Chief Operations Officer Anthony Guarino denied Appellant’s appeal in a letter that provided in relevant part as follows:

As we are mandated to derive the average monthly salary only from “a period of thirty-six consecutive months during which a member receives the highest compensation within the last one hundred twenty months of credited service,” we cannot calculate your highest 36-months assuming an ‘optimum’ retirement date within April 2016, September 2016, or some other timeframe you believe to be most advantageous. You earned service credit and contributed through July 2017. If we were to include April 2006 through July 2007 in your last ten years of salary history, we would actually be using a period of over ten years from which to derive your highest 36 consecutive months of salary. Using more than a ten year period to determine the highest salary range is not permitted in statute.

Further, A.R.S. § 38-765 states that “if any change or error in the records results in any member or beneficiary receiving from ASRS more or less than the member or beneficiary would have been entitled to receive if records had been correct, ASRS shall correct the error and as far as practicable shall adjust the payments in a manner so that the actuarial equivalent of the benefit to which the member or beneficiary was correctly entitled is paid.” This means the ASRS is statutorily required to pay a member only what they are eligible for under the law based on their chosen retirement date and retirement options elected.

The ASRS has addressed this specific matter with the area responsible for completing benefits estimates to ensure the estimate review process is followed for benefit estimates whose highest 36-month range falls in the latter half of a member’s ten-year period in all cases going forward. Though the ASRS sympathizes with your situation, we maintain it is our fiduciary responsibility to uphold statute and pay only what is actuarially correct and what you are actually entitled to receive. Benefit estimates are provided as educational tools we provide to our membership in an effort to assist with retirement planning and are not income promises prior to the actual audit of your account upon retirement. While we understand your disappointment over the incorrect estimates you received from ASRS staff, we cannot award you a higher monthly benefit beyond what statute permits due to an error on an estimate based on unaudited data.

Appellant acknowledged that, as an accountant, she audited financial statements for state agencies. Appellant also acknowledged that she understood the formula set forth in statutes for the calculation of retirement benefits. Appellant acknowledged that she knew in December 2009, that her decision to work less would affect her salary and that the months in which she received higher compensation would start dropping off the ten-year period used to calculate benefits in December 2016. Appellant also acknowledged that the spreadsheet that Ms. Dyer provided in September 2016, provided accurate information about her salary history.

Ms. Golub testified that the Appellant reached the 80 points needed for a normal retirement based on the total of her age and years of service on March 1, 2017. By that date, however, some of her higher earning months from 2007 were dropping from the 120 months that ASRS is allowed to consider to calculate average monthly salary. Ms. Golub testified that ASRS cannot tell members when to retire under statute and that sometimes employers offer employees incentives to retire, such as insurance, that cause some members to choose to retire at a time that does not fully optimize their retirement benefits.

Ms. Golab performed calculations based on audited data from the December 31, 2016 retirement date, which was the last estimate that was provided with an accurate projection of Appellant’s average monthly salary, and March 1, 2017, which was the first date on which she achieved the 80 points necessary for a normal retirement.

For the December 31, 2016 retirement date, Ms. Golab calculated that Appellant would receive a monthly life annuity 10-year term certain of $6,030.50, which was about $11.00 less than the $6,041.96 projected on the estimate. Ms. Golab opined that her calculation was so close to the estimate because the average monthly salary of $9,572.36 on the estimate correctly captured the historical salary data.

For the March 1, 2016 retirement date, Ms. Golub arrived at an average monthly salary of about $9,444.50, based on a 120-month range of March 2017 through April 2007. Ms. Golub calculated a life annuity 10-year term certain estimate of $6,192.00 per month, which was $61.00 less than the projected $6,161.44 amount. Ms. Golub testified that “[t]he average monthly salary on this March 1st, 2017, estimate is using the same exact salary as used for December 31st, 2016, estimate, and does not appropriately account for old historical higher salary dropping off to account for the additional months worked.”

Ms. Golub noted that the number Appellant used in her appeal to ASRS from the estimate was the straight-life annuity monthly payment projected on the July 1, 2017 estimate of $6,457.93, even though she had elected to receive the life annuity 10-year term certain. Ms. Golub testified that the life annuity 10-year certain estimate of $6,394.06 was only $400.73 more than Appellant’s $5,993.33 monthly benefit, which equated to about a 5.6% difference.

Ms. Golab testified that Ms. Dyer was a benefit advisor at ASRS. She was a non-supervisory front-line staff member who was not an attorney. Ms. Golab testified that although ASRS staff try to help member prepare for retirement, it is ultimately up to the member to make the decision that best optimizes what is important to the member.

CONCLUSIONS OF LAW

This appeal lies within the ASRS’ Board’s jurisdiction.

Appellant bears the burden of proof to establish that ASRS incorrectly or improperly determined her average monthly compensation for purposes of calculating her retirement benefit as $8,748.20 on her retirement date of July 17, 2017, by a preponderance of the evidence.

“A preponderance of the evidence is such proof as convinces the trier of fact that the contention is more probably true than not.” A preponderance of the evidence is “[t]he greater weight of the evidence, not necessarily established by the greater number of witnesses testifying to a fact but by evidence that has the most convincing force; superior evidentiary weight that, though not sufficient to free the mind wholly from all reasonable doubt, is still sufficient to incline a fair and impartial mind to one side of the issue rather than the other.”

Appellant argued that she detrimentally relied on statements by ASRS employees and the various ASRS estimates that will cause her financial harm for the rest of her life and ASRS failed to advise her of the significance of the date she chose for retirement. Appellant argued that ASRS should be estopped from paying her any less than the optimal retirement benefit she otherwise might have been entitled to receive because Ms. Dyer and the various computer-generated estimates gave bad advice.

“[E]stoppel may apply against the state only when the public interest will not be unduly damaged and when its application will not substantially and adversely affect the exercise of government powers.” If ASRS was required to pay Appellant more than she was entitled to receive on the date that she eventually selected for her retirement, she will receive more than other similarly situated retired members receive under A.R.S. §§ 38-747(B), 38-758, and 38-711.

In addition, for estoppel to lie against a state agency, the agency’s communication giving rise to the estoppel must be formal, usually in writing, and made by “a person authorized to act in the area under consideration.” “In general, the state may not be estopped due to the casual acts, advice, or instructions issued by nonsupervisory employees.” Most of Ms. Dyer’s advice was not in writing. Ms. Dyer was a nonsupervisory employee and was not authorized to bind ASRS by her advice to Appellant. In addition, the clear language of the bold capitalized disclaimer on every estimate sent to Appellant warned her that the amounts were estimates, not income promises. Despite Appellant’s claimed reliance on Ms. Dyer’s advice, ASRS is not estopped from calculating her retirement benefit according to governing statute on the retirement date that she elected.

No statute requires ASRS to do more than make information about Appellant’s account accessible to her. ASRS provided accurate information about Appellant’s age and years of service according to the information in its records and corrected erroneous information when necessary. ASRS twice provided accurate information about Appellant’s salary history to allow her to make her own calculations to determine when the optimal date for her retirement would be, in light of the 36 highest earning months in the last 120 months before her retirement, achieving a normal retirement, and the multiplier. Appellant was an accomplished, well-compensated accountant who was more capable than most members of performing the mathematical calculations required by statute to calculate her potential retirement benefits at any given time and optimal retirement date under the governing statutes. Appellant acknowledged in her September 2016 appeal that she knew her benefit would be based on the highest 36 months’ salary in the 120 months before her retirement. At a minimum, Appellant should have questioned the accuracy of the estimates in early November 2016, when the average monthly salaries for the December 30, 2016 retirement date and the February 28, 2017 retirement date were both shown as $9,572.36, even though two high months of salary should have dropped off to arrive at the February 2017 estimate. She acknowledged that she questioned Ms. Dyer’s advice but inexplicably failed to perform her own calculations.

The fact that Appellant apparently still has not performed those calculations to retroactively arrive at an optimal retirement date does not make ASRS responsible for paying whatever optimal benefit she might have received based on that date. Because Appellant is responsible for the choices she made, A.R.S. § 38-715(D)(4) does not require ASRS to give her an optimal retroactive retirement date.

RECOMMENDED ORDER

Based on the foregoing, it is recommended that the ASRS Board affirm its determination that Appellant Julie Schweigert’s average monthly compensation for purposes of calculating her retirement benefit was $8,748.20 on her retirement date of July 17, 2017, and denial her appeal.

In the event of certification of the Administrative Law Judge Decision by the Director of the Office of Administrative Hearings, the effective date of the Order will be five (5) days from the date of that certification.

Done this day, May 11, 2018.

/s/ Diane Mihalsky

Administrative Law Judge

Transmitted electronically to:

Paul Matson, Director

Arizona State Retirement System