Memo: PERC Confirms High Cost to Taxpayers of Closing Pennsylvania’s Defined Benefit Pensions

Date of Press Release: 
June 26, 2013


To: Members of the Pennsylvania General Assembly

From: Dr. Stephen Herzenberg, Executive Director, Keystone Research Center

Date: June 26, 2013

Re: PERC Confirms High Cost to Taxpayers of Closing Pennsylvania’s Defined Benefit Pensions

Download a PDF Version of this Memo

Yesterday, the Pennsylvania Public Employee Retirement Commission (PERC) transmitted actuarial notes analyzing the financial impact of House Bill 1352, which would transition future school employees from the state’s current defined benefit (DB) pension system to 401(k)-type defined contribution (DC) retirement accounts, and House Bill 1353, which would transition future state employees to a DC plan. Both House proposals also incorporate most of Governor Corbett’s proposed cuts in the additional benefits that current school and state employers would earn for future years of service. PERC’s Actuarial Note, based on analysis by consulting actuary Tony Parisi of Cheiron Corporation, concluded that the House proposals to close the defined benefit pension plans could cost as much as $37 billion on net: $33.8 billion of this total would be the cost of closing the Public School Employees’ Retirement System (PSERS) defined benefit plan to new school employees.

Cheiron’s analysis relied on actuarial studies of the House proposals by the SERS and PSERS actuaries, Hay Group and Buck Consultants. The House proposals that Hay and Buck studied are similar to Senate Bill 922. (One difference is that the original Senate proposal included the Governor’s proposed cut, from 2.5% to 2%, in the multiplier for future years of service earned by most current SERS and PSERS members.) PERC’s overall conclusions—based on Cheiron’s determination that the Buck and Hay studies are sound and reasonable—are similar to Keystone Research Center’s (KRC’s) conclusions based on the Buck and Hay [1] In particular:

1. Transition costs of closing the state’s defined benefit pension plans—$44.2 billion. The biggest-ticket cost of the House proposals results from shutting SERS and PSERS to new members. Buck Consulting actuaries estimate this cost to PSERS (based on HB 1342) at $34.1 billion. Hay Consulting estimates this cost to SERS (based on HB 1343) at $10.1 billion. Cheiron judges as reasonable the Buck and Hay assumptions of lower investment returns once the PSERS and SERS plans close to new members.

2. Higher retirement costs for new employees—$2.1 billion. In the Pension Reform Act of 2010, Pennsylvania lawmakers lowered the cost of pensions for new employees to only 3% of payroll on average.[2] The House proposal would increase future pension costs to 4% of payroll. The added cost over the next three-plus decades would be $2.1 billion, according to the actuarial studies of the House proposal—a higher cost of $6.9 billion for PSERS and a savings of $4.9 billion for SERS. (The difference is $2.1 billion not $2 billion because of rounding errors.)

The SERS and PSERS proposals estimate that savings from cutting the future pension benefits of current employees would be nearly $10 billion. If the courts ruled the benefit cuts for current employees unconstitutional, however, these savings would disappear, and the House proposals would be projected to cost closer to $50 billion.

One other important cost from a switch to defined contribution plans, not addressed in the PERC transmittals, results from the high costs of such plans relative to defined benefit plans. There is an overwhelming body of evidence showing that DC retirement savings plans cost a lot more to achieve the same level of retirement security.[3] Individual accounts cost more than Pennsylvania’s current defined-benefit pensions because they have higher administrative and financial management fees. They also achieve substantially lower investment returns: as just reported by Forbes magazine, defined benefit pensions “beat 401(k) savers silly” for the simple reason that professional managers outperform employees making their own investment choices.[4] Put all the inefficiencies of 401(k)-type plans together and, according to the National Institute on Retirement Security, these plans cost 57% to 83% more than Pennsylvania’s current pensions to achieve the same level of retirement security.

In a memo to lawmakers earlier this week, KRC relied on the Hay and Buck studies of the Senate Bill 922 to suggest shifting new employees from the current state DB plans to DC retirement accounts would hurt taxpayers as well as reduce the retirement security of future Pennsylvania teachers, nurses, and other public servants. Now yet another actuary (Cheiron) and PERC have spoken. It’s time for Pennsylvania lawmakers to go back to the drawing board on pensions.


[1] The Keystone Research Center's summary of the findings of the Buck and Hay studies of the Senate proposal were contained in KRC pension primer #7, A $40 Billion Dollar Oversight: Actuarial Studies Document High Cost of Governor Corbett’s Pension Plan,

[2] The Governor’s pension report lists the current (2012) “normal cost” of pensions for new employees as 2.2% for PSERS and 5.1% for SERS. (Normal cost is the cost of projected benefits for the current plan year; normal cost for new employees is thus the cost of projected benefits for new employees for the current plan year. See Public Employee Retirement Commission (PERC), Funding and Reforming Public Employee Retirement Systems, January 2013, p. 235; online at Since PSERS has about 2.5 times as many participants as SERS, the average of the PSERS and SERS rates weighted by the size of each plan is roughly 3%. See PA Office of the Budget, The Keystone Pension Report: A Discussion of Structural Reform and Relief to Pennsylvania’s Retirement System for Long Term Sustainability, p. 5 and p. 2; accessible online at

[3] For details, see Keystone Research Center, Less Bang for Pa.'s Buck: Governor’s Pension Proposal Would Force Taxpayers (and Employees) to Foot the Bill for Retirement Plans with High Fees, Low Returns,