Adopt a Retirement Incentive for State Employees
The Legislature should increase the benefit rate from 2 percent to 2.25 percent for employees eligible to retire in the Employees Retirement System during the 1994-95 biennium.


Background
As fiscal conditions have tightened in many states, retirement incentives have been offered by the public sector to reduce payroll costs (salaries and benefits), reduce the number of employees in the workforce, or alternatively, to reorg anize staffing in agencies and departments. Private employers have turned to retirement incentives for similar reasons.

Retirement incentives are temporary increases in retirement plan benefits designed to attract a certain group of employees to retire. Retirement systems, such as the Employees Retirement System of Texas (ERS), can offer a retirement window of opportunity. This is a limited, designated time period for retiring employees to receive enhanced benefits.

Retirement incentives or windows vary in form, but the most common incentives for defined-benefit pension plans (see definition below) include: (1) increasing retirement benefits by increasing the unit-benefit percentage, increasing years of service or changing the calculation of fina l average salary; (2) expanding retirement eligibility by reducing the age or years of service required for retirement, or reducing or eliminating penalties for early retirement; (3) paying health or life insurance premiums; and (4) paying lump-sum retirem ent bonuses. 1~

In Texas, ERS administers a defined-benefit pension plan for state employees. Under this plan, the state commits to pay a specific annuity to a retiree based on a unit-benefit formula that considers an employee s years of service, a benefit rate of 2 percent for each year of service and an employee s salary level. For example, an employee who retires with 20 years of service would receive a monthly annuity in the amount of 40 percent (20 years of service x 2 percent) of his or her average monthly salary for the three years of highest earnings.

Members of ERS are eligible for normal retirement at age 60 with 5 years of service, age 55 with 25 years of service or age 50 with 30 years of service. ERS does not provide members the option of retiring early with reduced benefits. However, as discussed in another recommendation of this report, commissioned peace officers and custodial officers are eligible to retire at age 50 with 20 years of service or at age 55 with 10 years of service.

Two key issues regarding retirement incentives are whether they actually save money over the long run and whether they should be targeted to certain groups of employees. On the issue of long-term savings, William C. Whitbeck, director of Michigan s Office of the State Employer, has noted: You achieve savings to the extent that you are able to permanently reduce the size of the workforce. 2~ Additionally, Minnesota s deputy commissioner of labor relations says retirement incentives are a good idea, if you don t have to refill the positions. 3~

Other experts say that many of the adopted retirement incentives ended up costing millions of dollars because many of the programs were poorly planned and many retirees were replaced. 4~ Additionally, according to David Kehler, director of the New Jersey Public Affairs Research Institute: Governments tend not to document the effectiveness of many of their [retirement incentive] programs. 5~

Regarding whether retirement incentives should be targeted to certain groups of employees, the Texas Performance Review (TPR) found that most states offering retirement incentives offered them to all eligible state workers, regardless of job responsibilities. As a result, the loss of too many difficult-to-replace and supervisory employees has been experienced, and some states have been forced to rehire the retirees as consultants or part-time employees, reducing payroll savings.

In the State of New York, a 1983 incentive program expected to save $50 million ended up costing $50 milli on because too many employees opted for the incentive and were subsequently replaced. To avoid what happened in 1983, New York narrowly targeted its 1990-91 incentive program to those employees in less critical jobs as determined at the discretion of agenc y management. The new program is expected to save $120 million by eliminating the positions of 6,000 employees who participated in the program. 6~ New York s attorney general determined that the narrowly-targeted program was not discriminatory, and no lawsuits have been filed by employees to date. 7~

Experience in Texas
The Legislature adopted a retirement incentive for state employees in August 1986 in response to an estimated deficit of $2.9 billion in the state budget. 8~ Governor Mark White called a special session to address the budget deficit, and the Legislature passed a retirement incentive (House Bill 40) for members of ERS in an effort to reduce the work force without having to consider mandatory work force reductio ns.

House Bill 40 allowed membe rs of ERS who were already eligible for normal retirement to retire during a window open between November 30, 1986, and May 31, 1987, with a benefit rate of 2 percent for all years of service. The incentive increased benefits by factors ranging from 6.7 pe rcent for members with 40 years of service to 33.3 percent for members with ten years of service. The bill had a restriction that a retiring employee could not return to work as a state employee.

The normal benefit formula was based on a 1.5 percent bene fit rate for the first ten years of service and a 2-percent rate for subsequent years (1.5 percent x average salary x first 10 years of service) + (2 percent x salary x subsequent years). Members could retire with full benefits at age 60 with 10 years of s ervice or at age 55 with 30 years of service. Members could also retire early with reduced benefits at age 50 with 30 years or at age 55 with 25 years of service.

At the time the legislation passed, ERS estimated that 6,787 employees with combined salari es of $179.5 million were already eligible for retirement. Assuming a 50 percent utilization rate and that no positions would be refilled, the estimated payroll savings from the incentive was $89.8 million, offset by lump-sum payments for accumulated vacat ion leave.

The number of employees that eventually retired during the window was 2,918 43 percent of eligible employees with total annual salaries of $75.6 million. The incentive added $186.2 million to ERS s unfunded actuarial liability and increased the amortization period by eight and a half years. The incentive did not increase the state contribution rate to ERS.

The incentive probably produced a short-term savings, but no one in the state monitored the savings. Additionally, before the window closed, the 70th Legislature passed a bill that allowed employees retiring during the window to return to work for the stat e. ERS membership figures in the following year indicated that most of the retiring employees had been replaced.

Experience in Other States
The National Association of State Budget Officers recently conducted a survey on state retirement incentives. 9~ Results indicated that 25 states offered various types of retirement incentives with mixed results. Some states estimated the short-term savings and costs associated with the incentives, but the majority did not monitor whether vacated positions were re-f illed. The survey did not address whether the 25 states actually reduced state expenditures by the total amount or a portion of the salaries of employees retiring under the window.

For the states that tracked the cost effectiveness, the net savings reported in the first year ranged from $1.5 million to $40 million. Some states reported negative effects on agencies due to loss of skilled and knowledgeable employees and the difficulty in re-filling certain professional and technical positions. Some states re-filled most of the vacant positions; others re-filled half.

The length of retirement windows ranged from two and a half months to more than one year, averaging seven months. About half the states offered early retirement incentives by expanding retirement eligibility. Other states increased benefits in the range of 5 to 10 percent by increasing years of service or the benefit percentage.

About half the states placed various restrictions on retirement incentive programs, such as requiring minimum age and years of service, delaying payments for accumulated leave, mandating that retiring employees could not return to state employment and r estricting earnings of re-hired annuitants. An average of 42 percent of eligible employees opted for retirement during the window of opportunity, ranging from 22 to 75 percent. As an alternative measure, nine states reported that retirement incentives doub led the normal level of retirements.

Experience in the Private Sector
More and more private companies are offering retirement windows as economic conditions have required staff reductions and streamlined operations. According to a 1992 Towers Perrin surve y of private employers, more than one-fifth of the 534 U.S. companies surveyed offered one or more retirement windows in the last four years. 10~ The windows were most common in large companies and in the manufacturing, utilities and banking industries. Eighty-eight percent of respondents said that the windows had accomplished their objectives.

The survey found that eligibility for retirement incentives was based primarily on age and service, rather than pay level. On the average, about 10 percent of the wo rkforce was eligible for the special retirement benefits. Companies typically offered more than one incentive to encourage employees to retire, and the average window length was six weeks to three months. The windows were often targeted at specific units o r divisions. About 37 percent of all eligible employees took advantage of the windows, and a greater percentage of highly-paid employees opted for the windows (47 percent) than lower-paid employees (34 percent). 11~

Among others, IBM, Ford and American Telephone and Telegraph (AT&T) have recently offered retirement incentives. IBM s voluntary workforce reduction program, the Individual Transition Option, ended in the U.S. on July 31, 1992, and resulted in a reduction of 32,000 employees worldwide, with a bout two-thirds of these employees from U.S. divisions. The IBM incentive package included one week of salary for each six months of employment (not to exceed 52 weeks of salary), career transition services and a transitional medical program. The program a llowed for regular retirement, a bridge to retirement or resignation. 12~

In 1991, Ford offered a narrowly-targeted early retirement window after experiencing the loss of many valuable employees in an incentive program offered in 1980-81. This plan excluded those with scarce skills and required management approval for employees who were eligible to opt for the window. Ford also set specific reduction targets for job categories in overstaffed operations. 13~

AT&T s 1989 incentive plan that initially cost $160 million to eliminate 12,000 managers now saves the company $450 million per year. As a result, the company offered another window in 1990 for 15,500 non-management employees. 14~


Recommendations
A. The Legislature should increase the benefit rate from 2 percent to 2.25 percent for state employees eligible to retire in the Employees Retirement System (ERS) during the 1994-95 biennium.

Assuming the bill is effective in time, employees eligible to retire during fiscal 1994 should have 60 days to elec t the incentive between July 1, 1993, and August 31, 1993. Otherwise, the election period should be moved forward to September and October of 1993. Employees eligible to retire during fiscal 1995 should also have 60 days to elect the incentive between July 1, 1994, and August 31, 1994. In order to receive the increase in retirement benefits offered during the two-year window beginning on September 1, 1993, employees must retire on the first date they are eligible to retire . For example, all employees eligible to retire on the first day of the biennium September 1, 1993 must retire on that date in order to receive the increase in benefits.

A provision should be included in the legislation authorizing the incentive so that employees retiring during the window could not return to work as a state employee for a certain time period, such as two years. Employees retiring during the window should be required to sign an agreement with the ERS expressing their intent to retire.

B. To fund the cost of the retirement incentive, the 30-year state contribution rate to ERS, as recommended by TPR in another part of this report, should be increased by the incremental cost of funding the retirement incentive over a period of 30 years, as determined by ERS actuary.

C. To ensure that the retirement incentive would produce a net savings and a reduction in payroll expenses, state agency appropriations should be reduced by one-half of the amount of the final annual salary for each employee who retires during the window.

To accomplish the budget reduction, ERS should report to the Comptroller of Public Accounts the final annual salary for each employee who elects the retirement incentive. The Comptroller s office should be authorized to reduce state agency appropriations by one-half of the amount of the final annual salary when each employee retires during the window. The budget reduction should be based on the number of months remaining in the biennium after each retirement. To avoid excessive budget cuts for small age ncies, some agencies should be exempted from the budget reductions, including those with total appropriations of less than $1 million per year. The Legislature would identify other types for exemption as may be appropriate.


Implications
The recommendation would produce a short-term payroll savings and a long-term actuarial cost. Retirement benefits would be increased for employees retiring under this incentive program. The incentive would increase the actuarial cost to ERS and reduce stat e agency budgets by one-half of the annual salary of each retiring employee.


Fiscal Impact
To assist in determining the fiscal implications of the recommendation, TPR contracted with the actuarial firm, Leggette & Company, Inc., to provide actuarial cost estimates and payroll savings. 15~

In November 1992, Leggette obtained employee records from ERS for both regular employees and commissioned peace officers. The data indicated that 12,630 employees will be eligible for retirement during the 1994-95 biennium with total p ayroll of more than $369 million. Of this amount, 8,507 employees will be eligible for retirement as of September 1, 1993, and the remaining 4,123 employees will become eligible at various times during the 1994-95 biennium. The average annual salary for el igible retirees is $29,280.

Leggette assumed that the incentive would be granted for all past service up to the date of retirement as long as the date occurred during the 1994-95 biennium. The actuarial cost represents the annual cost to fund the incentiv e over a period of 30 years. The annual cost to ERS to fund the incentive over 30 years could be offset by future actuarial surpluses resulting from favorable experience in investments and other factors.

The recommendation would increase standard retirement benefits by 12.5 percent. The present value cost of the additional benefits is estimated to be $121.9 million. The 30-year state contribution rate to ERS (6.35 percent of payroll), as recommended elsewhere in this report, would need to be increased by 0.35 percent to fund this incentive over period of 30 years, and this cost is reflected in the estimates shown below.

Payroll savings are estimated at 50 percent, which represents the amount that would be reduced from agency appropriations during each year of the biennium. It is assumed that retirements would occur during the biennium as employees are eligible and that 25 percent of the retirees would not be replaced. The cost estimates of lump-sum termination payments assume an average annual leave balance for each retiree of 200 hours, which is the average balance of employees retiring from the Comptroller s office in fiscal 1992.

Assuming rates of retirement at 175 percent of the standard retirement rates provided by ERS actuary, 3,118 employees with total payroll of $93.1 million would retire in fiscal 1994 and 2,239 employees with payroll of $65.1 million would retire in fiscal 1995. Thus, 42 percent of eligible retirees would retire during the window. About 59 percent of the net savings/costs would accrue to the General Revenue Fund.

The following table summarizes the net savings/cost of the recommendation assuming the biennial ERS state contribution rate is set at the proposed 30-year rate of 6.35 percent. If the biennial state contribution rate to ERS returns to 7.4 percent the statutory rate the cost of the retirement incentive would be funded out of surpluses in the ERS retirement fund, nullifying the savings shown in another recommendation in this report to reduce ERS contribution rate.

Net Savings/Cost With ERS State Contribution Rate of 6.35 Percent

Lump-Sum
Fiscal Actuarial Payroll Termination Net
Year Cost to ERS Savings Payments Savings/(Cost)

1994 $11,344,000 $42,775,000 $8,778,000 $22,653,000
1995 11,344,000 63,831,000 6,304,000 46,183,000
1996 11,344,000 * 0 (11,344,000)
1997 11,344,000 * 0 (11,344,000)
1998 11,344,000 * 0 (11,344,000)


Gain/(Loss) Gain/(Loss) Net
Fiscal to the General to Other Dedicated Gain/(Loss) Change in
Year Revenue Fund 001 Accounts or Funds to All Funds FTE s

1994 $13,411,000 $ 9,242,000 $22,653,000 -780
1995 27,340,000 18,843,000 46,183,000 1,340
1996 (6,716,000) (4,628,000) (11,344,000) *
1997 (6,716,000) (4,628,000) (11,344,000) *
1998 (6,716,000) (4,628,000) (11,344,000) *

Five-Year Total $34, 804,000

* It is anticipated that some payroll savings would continue after the 1994-95 biennium to the extent that agencies do not replace retirees or replace them with lower-paid employees; however, t his savings would be offset by productivity losses and cannot be estimated.



Endnotes
~1 National Conference of State Legislatures, Fiscal Affairs Program, National Association of State Budget Officers (NASBO) Reports on States Experience with Early Retirement, The Fiscal Letter, March/April 1992, p. 4. (Newsletter.)
~2 Anne Jordon, The Gold Watch Plan For Cutting Budgets, Governing, (July 1992), p. 25.
~3 Ibid.
~4 Michael deCourcy Hinds, Early Retirements to Reduce Budgets Cost States Money, New York Times (November 16, 1992), p. A11.
~5 Ibid.
~6 Ibid.
~7 Telephone interview with Lou Rafaele, Associate Budget Examiner, New York State Division of Budget, Albany, New York, December 11, 1992.
~8 State Pension Review Board, Early Retirement Study Outline, n.d., pp. 63-66.
~9 National Conference of State Legislatures, The Fiscal Letter, pp. 4, 11.
~10 Towers Perrin, More Employers Opening Retirement Windows, Monitor, September 1992, Issue 134, pp. 1-3. (Newsletter.)
~11 Ibid.
~12 International Business Machines, documentation dated September 3, 1992.
~13 Joann S. Lublin, Workplace: Bosses Alter Early-Retirement Windows to be Less Coercive-And Less Generous, Wall Street Journal (April 1, 1991), p. B1.
~14 Selwyn Feinstein, Labor Letter: A Special News Report on People and Their Jobs in Offices, Fields and Factories, Wall Street Journal (October 30, 1990), p. A1.
~15 Leggette & Company, Inc., Final Report to the Comptroller s office, December 9, 1992.