STATE EMPLOYEES' RETIREMENT SYSTEM Adoption of the Group Annuity Mortality Tables; Resolution No. 2001-103 [31 Pa.B. 7087] Whereas, the State Employees' Retirement System (SERS) has historically utilized the 1971 Group Annuity Mortality (GAM) tables to determine actuarial equivalence; and
Whereas, based upon the recommendation of the SERS actuary, the State Employees' Retirement Board (Board), at its March 2001 meeting resolved to change to the use of the more recent 1983 GAM (GAM 83) tables for the purpose of benefit calculations; and
Whereas, in order for SERS to begin utilizing the GAM 83 tables for the purpose of benefit calculations beginning on January 1, 2002, the Board must formally adopt the GAM 83 tables through this Resolution; and
Whereas, a detailed explanation of the planned utilization of the GAM 83 tables follows as an exhibit; and
Whereas, this resolution and the following detailed explanation will be published in the Pennsylvania Bulletin prior to January 1, 2002;
Now Therefore, in consideration of the foregoing recitals that are made a part hereof, it is hereby:
Resolved, this 5th day of December, 2001, that the Board adopt the GAM 83 tables for the purpose of benefit calculations as explained in the document entitled: ''Applying Actuarial Equivalence Factors-Methodology for Determining ''Better-of,'' which follows as an exhibit to this resolution.
Applying Actuarial Equivalence Factors Methodology for Determining ''Better-of'' Background
The Pennsylvania State Employees' Retirement System (SERS) uses a complex array of actuarial tables to determine actuarial equivalence for optional settlements and early retirement. The State Employees' Retirement Board has decided to move SERS from the 1971 to 1983 Group Annuity Mortality (GAM) table to determine actuarial equivalence to conform to the change in the tables used by the actuary for costing bases.
Currently, there are three sets of actuarial tables used for determining actuarial equivalence for SERS covered employees. The reason for the different sets of actuarial tables is the United States Supreme Court decisions in Arizona Governing Committee v. Norris, 463 U.S. 1073, 103 S. Ct. 3405, 77 L.Ed. 1236 (1983), decided in August 1983 and Florida v. Long, 487 U.S. 223, 108 S.Ct. 2354, 101 L.Ed. 206 (1988), decided in 1988. These decisions required that pension plans pay equal benefits for men and women. The issue was further complicated by the requirement that the remedy (the equalization of benefits) could not be accomplished by reducing higher benefits to lower levels for current employees.
In conformance with the Supreme Court decisions, the approach taken by SERS was to equalize benefits earned after August 1983 for members who entered SERS before August 1983. SERS currently uses ''blended'' sex-neutral actuarial tables for employees hired on or after August 1, 1983. For employees hired before August 1, 1983, the benefits resulting from service credited before August 1, 1983 continued to be calculated using the existing sex-distinct actuarial tables, while benefits resulting from service credited on or after August 1, 1983 uses ''topped up'' actuarial tables. The reason for the three sets of tables is that the Supreme Court stated that Norris only had to apply to benefits resulting from service after August 1983 and because the prohibition against ''blending'' tables to produce equal benefits does not apply to employees hired on or after August 1, 1983.
For the December 31, 2000 valuation SERS used the 1983 GAM sex-distinct tables for all post-retirement mortality for the purpose of determining employer costs to fund the pension plan. As a result, SERS has decided to move to the 1983 GAM tables for actuarial equivalence and for determining benefits. Actuaries can, and do, use sex-distinct tables to produce the best estimate of costs. But, SERS' use of the 1983 GAM unisex (blended) table for actuarial equivalence purposes is still controlled by Norris.
Therefore, SERS's adoption of the 1983 GAM tables for actuarial equivalence parallels SERS' adoption of the 1971 tables after Norris.
A. Beginning with retirements with an effective date on or after January 1, 2002, employees hired on or after August 1, 1983 will have their benefits calculated using the 1983 GAM unisex tables rather than the 1971 GAM unisex tables. Almost all post-August 1, 1983 members will receive benefits under the 1983 GAM tables equal to or better than they would have received under the 1971 GAM tables.
B. For employees hired before August 1983, benefits will be calculated two ways and the employee will receive the benefit based on whatever method produces the ''better'' benefit. (Determination of the ''better'' benefit is described below.) Benefits will be calculated using the 1983 GAM unisex table applied to all benefits, regardless of whether the service underlying those benefits was credited before or after August 1, 1983. This benefit will be compared to a benefit calculated using the current methodology applying 1971 GAM tables to post-July 1983 service and sex-distinct 1971 tables to pre-August 1983 service).
Determination of ''Better of''
A member's Maximum Single Life Annuity (MSLA) will be used to determine the ''better'' benefit when comparing GAM71 and GAM83 tables. When establishing the benefit of a pre-August 1983 hire, SERS will compute the member's MSLA using the GAM71 tables (sex distinct and ''topped up'') and the GAM83 tables. The tables that generate the larger benefit will be the basis for any option the member may select and any other actuarial calculation. If the two methods produce the same MSLA, then the set of actuarial tables that produces the higher present value of MSLA will be used.
Return to Service--How to Apply the ''Better-of'' Comparison
When a member retires, a ''better of'' comparison will be performed and his annuity will be determined using the factors that provide him with a higher benefit. At return to service, the member's annuity is frozen. The calculation of the benefit at subsequent retirement depends on the credited service after the return. If the member has less than 3 ''new'' service credits, the member will receive a two-part benefit. The first part of the benefit is a monthly benefit derived from the frozen present value using the same tables that were used at first retirement, based on the member's current age. The second part of the benefit is a benefit for the new service; the better-of test is performed on the new service benefit only.
If the member has 3 or more service credits at subsequent retirement, the member will receive the larger of the following:
* A benefit based on ''thawing'' the frozen present value, where the total years of service and a new high-three average salary are determined. The new benefit is offset by the actuarial value of the retirement benefits received while on the first retirement. The better-of test is performed again, on the total benefit.
* The two-part benefit described above. The benefit derived from the frozen present value is based on the same tables that were used at first retirement, and the second part of the benefit is a benefit for the new service. The better-of test is performed on the new service benefit only.
Change of Option--How to Apply the ''Better-of'' Comparison
A member who elected a joint and survivor option is allowed to change the retirement option under two circumstances:
1. a change in marital status; or
2. the designated survivor predeceases the member.
The ''better of'' table that was originally determined to produce the better retirement benefit would be used as the basis for all optional benefits, including any changes in optional benefit elections.
Domestic Relation Orders--How to Apply the ''Better-of'' Comparison
This would only be an issue for members who received a benefit estimate under a Domestic Relations Order (DRO) before the change in actuarial equivalence factors and retire after the change. However, since all elements used to determine the DRO were estimates, it will not be necessary to revisit the calculation.
Class C Supplemental Benefit--How to Apply the ''Better-of'' Comparison
When calculating the member's Maximum Single Life Annuity under GAM71, male ''topped up'' factors for the Class C Supplemental Benefit, as is done currently, will be used. When calculating the member's Maximum Single Life Annuity under GAM83 however, GAM83 factors for all benefits, including the Class C Supplemental Benefit, will be used.
Death Benefits (Act 500)--How to Apply the ''Better-of'' Comparison
Following the same procedure as used when the Norris decision was implemented, all Act 500 death benefits will be calculated using the table that applies to new members of SERS. Therefore all Act 500 death benefits will be calculated using the 1983 Group Annuity Mortality Table for deaths after January 1, 2002.
Disability Retirements--How to Apply the ''Better-of'' Comparison
For the reasons stated above, the 1983 GAM tables will be used for determining disability retirement benefits for members hired after August 1983. For pre-1983 hires, the member's MSLA will be used to determine the better retirement benefit when comparing GAM71 and GAM83 tables. However, the benefit determination for disabled retirements does not include the calculation of the MSLA.
It is to the member's advantage to have the early retirement portion of the basic disability benefit be the better benefit. Therefore, the table that generates the largest early retirement benefit, of the total basic disability retirement benefit, will be used as the basis for any option the member may select. If all tables produce the same early retirement benefit, then the table that produces the higher present value for the early retirement benefit will be used as the basis for any option the member may select.
If a hazardous duty member (enforcement officers and State Police) has less than 5 years of service, the disability supplement is the difference between the annual basic disability benefit described above and the member's share of the annual disability benefit. The member's share is the accumulated deductions divided by the initial reserve factor, unless the disability is a joint and survivor disability.
Members who have less than 5 years of service at disability would most likely have entered after 1983, and therefore the member's share of the basic disability benefit would be determined using the 1983 GAM table.
Members who would like to obtain a copy of the mortality tables may visit www.sers.state.pa.us or call a retirement counselor at 1-800-633-5461.
NICHOLAS J. MAIALE,
Chairperson[Pa.B. Doc. No. 01-2351. Filed for public inspection December 28, 2001, 9:00 a.m.]
