SENATE GOVERNMENTAL AFFAIRS 13 DECEMBER 1983 RETIREMENT FORUM

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP86B00338R000400510001-3
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 160
• Document Creation Date: December 21, 2016
• Document Release Date: August 21, 2008
• Sequence Number: 1
• Publication Date: January 25, 1984
• Content Type: MEMO

Approved For Release 2008/08/21 : CIA-RDP86B00338R000400510001-3
2 5 JAN 1984
MEMORANDUM FOR: Deputy Director of Personnel for Special
Programs
Deputy Director of Personnel for Policy,
Analysis, and Evaluations
Office of Legislative Liaison
raison Division
SUBJECT: Senate Governmental Affairs 13 December 1983
Retirement Forum
1. Attached for your information and analysis is a
complete copy of the witness list and prepared testimony
submitted for the subject retirement forum. This forum, the
first of several that will be conducted during the coming
months, was sponsored by Senator Ted Stevens (R., AK), in
his role as Chairman of the Civil Service, Post Office, and
General Services Subcommittee of the Committee on Govern-
mental Affairs. The purpose of these retirement forums is to
develop a broad base of understanding and knowledge of
retirement and pension systems in both the public and
private sectors in preparation for consideration of
supplemental Social Security retirement legislation.
2. I would normally have attended this forum but was
not available on the date it was conducted. I will plan to
attend all subsequent forums, the next of which is scheduled
for 16 February 1984. If the Office of Personnel (OP) is
interested in having one of its own at this and succeeding
retirement forums sponsored by Senator Stevens, please
advise and I will try to make the necessary arrangements. It
is my understanding that Mr. Ed Hustead of Hay Associates,
Inc. attended the 13 December 1983 session.
3. Please advise if this office can be of further help
in this or any other regard.
Attachment
As stated
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Distribution:
Original - Addressees
1 - C/LD/OLL w/o att
rl - C/LEG/OLL w/att
1 - TBC Chrono w/o att
1 - TBC Subject w/att
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 1. M /(ACV. Ill
II V(NS. ALASKA
~(S MC C ML7N1AS. JR.. MD
.M S CON(N, MAINE
OUR, NCI NG(R. MINN.
1 N ! RUDMAN, N H.
C UANI ON IN, MO.
COCHRAN. MISS.
AM L ARMSIRONG. COLO.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
-0 qP
1NOMAS I IAGIfION. MO
NINNY M JACKSON, WASH
LAWTON CNILES, FLA
SAM U.N. NGA
JOHN GI INN0OHIO
JIM SASS(R. TINN.
CARL LI VIN, MICH
JIFF DING AMAN, N. MIX
COMMITTEE ON
GOVERNMENTAL AFFAIRS
SUBCOMMITTEE ON
CIVIL SERVICE. POST OFFICE. AND
GENERAL SERVICES
SuRCO-11((
1(0 v11VLNS. LIAS(A. CN.:NMAN
CNANILS MC C M47MIAS, JR. MD JIIF W.G.M.N. N MIX
WILLIAM L ARMSTRONG COLO JIM SASSIR, I(NN
WAY"( A SCNILY. S.AIF DIRICIOR
Policy Forum
Economic Security Programs
December 13, 1983
9:30 a.m. - 4:30 p.m.
Dirksen Senate Office Building Room SD106
I. The Present Federal Retirement System Philip Royal Shipp, Jr.
Congressional Research Service
II. Sources of Economic Security
III. Defining Retirement Income Objectives
IV. Defined Benefit Plans
V. Defined Contribution Plans
Dallas L. Salisbury
Employee Benefit Research
Institute
Vincent M. Tobin
Buck Consultants, Inc.
Dan M. McGill, Ph.D.
University of Pennsylvania
Teachers Insurance and
Annuity Association/
College Retirement
Equities Fund
VI. Defined Benefit/Defined Contribution: Robert D. Krinsky
A Comparison Martin E. Segal Company
MODERATOR: Paul S. Berger
Arnold F, Porter
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
United ~5tates 5enflce
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
ed Stevens
assistant Majority Leader
FOR IMMEDIATE RELEASE
Dec. 5, 1983
Contact: Press Office
(202) 224-1039 '
FEDERAL PENSION POLICY FORUM PLANNED
A Senate Governmental Affairs Subcommittee will sponsor
a series of educational policy forums on federal government
pension systems according to Senator Ted Stevens, Chairman of
the Subcommittee on Civil Service, Post Office and General
Services.
The forums are part of a broad range of studies requested
by the Governmental Affairs Committee to be used in developing a
new supplemental civil service retirement program.
Beginning in 1984, all newly hired federal employees will
be required to participate in the social security system as
well as the interim plan designed to relieve financial hardships
that result from paying into two systems. The interim plan
allows new federal workers to make reduced contributions to the
civil service system until a new retirement system is developed.
"The purpose of the forums will'be to bring together in
an informal setting various interested parties to discuss the
issues involved in developing a new pension system," Stevens
said. "Discussions will be focused on a broad range of issues
such as pension plan designs, benefit levels and funding and
would serve to educate participants and clarify issues."
The first forum will be at 9:30 Dec. 13, in SD-106.
(more)
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
W *
-2-
Pension experts presenting papers at the forum will include
Philip Royal Shipp, Congressional Research Service; Dallas
Salisbury, Employee Benefit Research Institute; Vincent Tobin,
Buck Consultants, Inc.; Dan McGill, Ph.D., University of
Pennsylvania; John McCormack, Teachers Insurance and Annuity
Association/ College Retirement Equities Fund and Robert Krinsky,
Martin Segal Company.
The forum will also include a round table discussion with
40-50 participants.
Persons interested in attending are asked to call the
Subcommittee staff at 224-2254.
-30-
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
SENATOR THOMAS F. EAGLETON
OPENING REMARKS
Before the
POLICY FORUM, ECONOMIC SECURITY,P.ROGRAMS
December 13, 1983
In 1920, when Congress passed the Civil Service Retirement
Act, some 330,000 federal employees were covered. Today, only
A3a
ears later, that original - y , Act has been modified more than
150 times by either Congress or an Executive Order, and some
2.7 million active civil servants and postal service workers,
plus 1.8 million retirees, are now covered.
What have these changes and this phenomenal growth meant?
Obviously, they are a reflection of the growth of the federal
government. They are also,'however, a reflection of major
policy changes about federal retirement.
When enacted in 1920, the civil service retirement program
sought to improve government service by improving benefits to
employees, but it also sought to keep costs down; by the 1940s
and the end of World War II, the purpose had changed a bit and
improvement of morale and the protection of employees became
a major policy concern; then, beginning in the `50s, the federal
retirement system acquired characteristics of Social Security
and social welfare, while at the same time setting an example
of a comprehensive employee income protection plan.
All of these changes were gradual. Today, however, since
Congress mandated inclusion of all new federal employees in
Social Security, beginning January 1, Congress will be forced
to totally review the civil service retirement program.
. Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
In one shot, we will enact.a new, perhaps similar, but perhaps
totally different, civil service retirement program.
Obviously, this will mean focusing on policy. Will we
want a system that simply "holds down costs," the basic policy
when the system started in 1920? Will we want a pension plan --
like most private sector plans and civil service in the '40s
-- which emphasizes employee protection and simply complements
the social welfare aspects of Social Security? Or, as we have
at present, will we want an exemplary plan that provides
comprehensive employee income protection?
Before we begin to address these policy questions, I feel
we must educate ourselves about what has happened and what is
happening now in'the pension area. The pension landscape of
1983-bears scant resemblance to retirement programs of the 1920s.
Today, pension planning and pension investment are more
than big business -- they are so big that pension fund billions
are becoming the tail that wags the nation's'economic dog. And
the plans are so complicated and varied it takes computers
simply to keep track of all the options: IRAs, Keoghs, 401(K)s,
thrift plans, standard annuities, stock options, and the like.
Finally, the laws themselves, particularly the Employee
Retirement Income Security Act of 1974, have so changed the
pension ballgame that to me it would be almost malfeasance
to draft major retirement legislation without a thorough study
of the realities of today's pension world, both public and
private.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Thus, I am delighted to welcome the pension experts and
others in this room who came to be enlightened about this
-complicated and important area.
As you may know, Senators Roth, Stevens, Bingaman and I
began our educational efforts in late summer, when, on behalf
of the Senate Governmental Affairs Committee, we asked the
General Accounting Office, the Congressional Budget Office,
and the Congressional Research Service to undertake a coordinated
effort to gather actuarial and economic pension data. We felt
it was vital for the Committee to have such data prior to
.drafting-legislation.
At the same time, we concluded that we, and others who
would be interested, simply needed to learn more about pensions.
The most efficient and effective way to cover the topic seemed
to be to invite "experts" to present information that could be
discussed and debated in a public forum. Today opens the first
of what we intend to be a series of five forums. We hope these
forums will acquaint the Congress and the public with pension
issues of the day so that when the politics emerge -- the lobbying,
the hearings, the actual legislative drafting -- we will all speak
with a better understanding of some basic facts and data.
To that end, I return this important forum back to Mr.
.Berger, adding only my personal thanks and the thanks of the
Committee to those individuals who made today's event possible.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Congressional Research Service
UAW The Library of Congress
Prepared by
P. Royal Shipp
Senior Specialist in Social Legislation
for the
1983-84 Policy Forums on Federal Retirement
Sponsored by
Subcommittee on Civil Service, Post Office and General Services
Committee on Governmental Affairs
United States Senate
December 13, 1983
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
The objective of this paper is to present an analytical framework for
thinking about the issues in the civil service retirement system. (CSRS).
Developments in this pension system over the past few years have resulted in
pressures for making major changes (or reforms) in the CSRS.
One development was the passage of the Social Security Amendments of 1983
which covered certain groups of Federal employees (principally those joining the
Federal workforce after January 1, 1984) by social security. Such coverage sug-
gests the need for a supplemental pension system, coordinated with social se-
curity, for those newly covered employees.
In addition, critics of the CSRS have claimed that it is too costly, too
generous, underfinanced, and that it has an undesirable effect on the ability
of the Federal Government to attract and retain competent and experienced em-
ployees. These criticisms have been accompanied by proposals by the admini-
stration and other groups for changing the current system. These criticisms
of the current CSRS will affect the design of a pension for employees newly
covered by social security.
The paper begins the discussion of an analytical framework by describing
briefly the origin, history, and basic provisions of the current CSRS. This
is followed by an analysis of the principal criticisms made of the current
CSRS. Most actuarial assessments of the CSRS have concluded that it is more
costly than pensions "typically" available in the private sector using the
widely accepted entry age normal cost system. Whether this leads to the con-
clusion that the current system should be reduced or that the cost of the new
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
system should be less than the current system is a difficult, and should be a
deliberate, policy choice. This policy choice should be driven, not just by
program cost, but by taking into account that while the value of Federal pen-
sion benefits is greater than the private sector, total compensation (including
pay and other benefits) may have fallen behind. Beyond that, it is not clear
that studies to date have adequately reflected recent changes in private pen-
sion and deferred compensation arrangements which have developed over the
past few years. The General Accounting Office (GAO) and the Congressional Re-
search Service (CRS) have studies.under way which may shed additional light on
this question.
The paper continues by grouping the specific analytical issues into three
basic categories.
First, there is a discussion of workforce goals, designed to indicate how
the choice of specific pension design features can induce greater employee re-
tention or greater turnover. The central design.feature at issue is the "por-
tability" of benefits; that is, the extent to which employees moving to employ-
ment outside the Federal Government will be able to take along already "earned"
pension rights. Since social security will be portable between Federal and pri-
vate employment for newly covered employees, this change by itself will facili-
tate movement back and forth.
Second, the design of a new pension system should consider the goal of ben-
efit adequacy. The concept of "replacement rates," that is,the proportion of
final year's earnings "replaced" by retirement income, is used to analyze this
issue. President Carter's Commission on Pension Policy proposed a national goal
of'retirement income sufficient to ensure. that retirees suffer no loss in income
standard-of-living upon retirement. Studies show that this goal is met by full-
career employees covered by the current CSRS and by other full-career employees
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
CRS-v
in jobs with basic social security supplemented by pension coverage. In con-
trast to social security, the current CSRS does not provide higher replacement
rates for lower income employees. An important policy decision will be how
much, if any, of this social security "tilt" toward lower-income beneficiaries
to retain in the design of the new CSRS.
The third major issue area is cost and financing of benefits. It is im-
portant to distinguish between these two. The analytical focus should be on
the cost of total benefit obligations of the system rather than on ways of fi-
nancing these benefits and on the differences between funded and unfunded lia-
bilities. A trust fund, with assets invested in the Federal Government's own
securities performs some important functions, but it does not ensure payment
of benefits and does not constitute a prefunded system in the sense that pri-
vate pension benefits can be prefunded. Investment of funds outside the Fed-
eral Government raises additional issues and is a possibility which should be
carefully explored.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
PREFACE ..................................................................
I. HISTORY AND PROVISIONS ............................................. 1
II. CRITICISMS OF THE CURRENT CIVIL SERVICE RETIREMENT SYSTEM .......... 12
A. Major Criticism: Too.Costly and Too Generous .................. 14
B. Other Criticism: Reduced Efficiency of the Workforce ......... 16
III. CURRENT AND FUTURE ISSUES .......................................... 19
A. Workforce Goals ................................................ 20
B. Benefit Adequacy ................................................ 23
C. Costs and Financing ........................................... 40
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? is
The Pendleton Act of 1883 marked the birth of the modern Federal Civil
Service. The Act sought to eliminate patronage and to attract and retain highly
qualified employees by basing the new classified appointments to Federal service
on open and competitive examinations. It also sought to provide continuity and
political neutrality in Federal service by awarding classified employees more
job security than normal private sector employment. The Act did not, however,
establish a retirement system for Federal civilian employees.
At the time Congress was considering the Pendleton Act, Federal government
workers neither expected nor planned to retire. Few retirement systems existed
and most employees worked all their lives in order to support themselves and
their families. Many people considered it unfair to fire employees whose only
failing was old age, and the removal of longtime civil servants frequently met
with a hostile public outcry.
Given these circumstances, government supervisors were understandably re-
luctant to remove elderly employees from their staffs, particularly since many
were Civil War veterans. Instead, an unofficial and unauthorized "public pen-
sion" plan emerged. Employees were essentially retired on the job--oldtimers
were carried on the rolls and paid full salaries, but performed little or no
work. This impromptu. plan impaired the performance of government operations
by blocking the flow of appointments and promotions required to recruit and
retain capable employees. In addition, it was very expensive.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
. 0
Beginning at the turn of the century, a variety of Civil Service pension
proposals began to emerge, sponsored by representatives of labor, management,
and the public. The first Federal pension bill was submitted to Congress in
1897, and in the next 23 years over 100 bills dealing with Federal pensions
were introduced. The Civil Service Commission began recommending a retirement
plan in its annual report in 1898, and retirement plans for civil servants
were included in the Republican Party platform in 1912 and the Democratic
Party platform in 1918.
Meanwhile, the Federal Civil. Service was growing rapidly. Between 1883
and 1918, the Civil Service grew from 131,000 employees to 917,000 employees.
With the end of World War I, there was increasing talk of cutting the Federal
bureaucracy back to pre-war size. Faced with the need to reduce the size of
the Federal payroll and to manage Federal personnel more efficiently, Congress
passed the Civil Service Retirement Act in 1920.
The 1920 Act generally limited its coverage to permanent competitive
employees in the executive branch of the Federal Government and to regular
employees of the District of Columbia Government. During the first year after
passage, approximately 330,000 employees were covered under the Civil Service
Retirement System (CSRS).
The Act established three retirement ages for different types of employment--
62, 65, and 70. In general, retirement became mandatory at the age prescribed,
regardless of length of service. Employees separated before reaching retirement
age could not receive annuity payments unless they had served at least 15 years.
Employees with at least 15 years of service could retire at the specified
ages, or at any age if they became totally disabled for useful and efficient
service in their position or some other position to which they could have been
assigned.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 s
The Civil Service Retirement System was established as a defined benefit
plan. In other words, the retirement benefit earned by an employee was based
not on how much money he (together with his employer) had contributed to the
system, but upon two other factors: the employee's average annual basic salary
during his last 10 years of work, and the employee's length of service, up to a
maximum of 30 years. The annuity could not be less than $180 per year and
could not exceed $720 per year. To fund this benefit, the Act required a de-
duction of 2 1/2 percent from the salary of each employee, to be credited to a
Civil Service Retirement Fund. The Secretary of the Treasury was authorized
to invest any Civil Service Retirement Trust Fund money not immediately needed
for payment of benefits in interest-bearing securities of the United States.
The interest from such investments became part of the fund.
The 1920 Act contained no. provisions for optional retirement, no survivor
benefits,..and no protections for employees who were involuntarily separated
through no fault of their own. The pensions provided were far from generous.
Strict eligibility requirements and small pensions meant employee contributions
more than covered the demands on CSRS funds during the first decade of the
plan. The Government, in fact, did not contribute any money at all to the
CSRS Fund until 1928.
The retirement system created by the 1920 Act was designed primarily as
a management tool. The law provided for only two types of retirement, mandatory
and disability. Mandatory retirement for most employees-was at age 70., unless
continuance in service was authorized by the employee's supervisor. This provi-
sion applied to all workers, even those who lacked sufficient years of service
to qualify for an annuity. An application for disability retirement could be
initiated either by the employee or by the employer.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 0
The initial objectives of CSRS, however, were soon modified by changing
economic conditions and new Government policies. The Great Depression exposed
the vulnerability of all workers to shifts in the economic climate, and the New
Deal expanded the Government's role in providing and promoting economic security
for workers and their families. The Government began to add additional benefits
to CSRS and to place additional employees under the plan. Coverage was expanded
from 331,000 workers in 1920 to over 2 million workers in 1942. Optional
retirement, survivor benefits, and protections for involuntarily separated
workers were all added to the system.
Events of the 1940s also made retirement benefits a more prominent part of
compensation for employees in commercial and industrial enterprises. During this.
time pension coverage for employees of State governments also was expanding, and
by 1947 every State had a retirement system for most categories of State employ-
ees.
In 1942, the Internal Revenue Service ruled that employer contributions to
qualified pension plans could be deducted from gross income as an ordinary bus-
iness expense. At the same time, Government wage controls imposed during World
War II prompted employers to offer retirement benefits as a means of retaining
and compensating workers for increased production efforts without contributing
to inflation. In 1947, the Supreme Court upheld a National Labor Relations
Board ruling that employees have a legal obligation to bargain in good faith
over the terms of pension plans. This decision, coupled with the steel indus-
try's fact-finding board ruling in 1949 that pensions were.an appropriate re-
sponsibility of industry, led to a rapid increase in the number of pension plans.
The 1930s and 1940s marked the emergence of a second generation of CSRS
objectives: the maintenance of income to employees. The concept of retirement
benefits as an indirect or deferred form of compensation slowly joined the
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
concept of retirement benefits as a management tool. A retirement pension be-
came viewed as a right earned through past service. Workers were held to be
entitled to an income sufficient to live in dignity upon completion of a full
career.
In the years following World War II, additional changes were made to CSRS.
A more generous formula for computing annuities was established, minimum bene-
fit levels were set to match those established by social security, survivor
benefits were improved, most Government employees not already covered by other
Federal pension plans were brought under CSRS, and those still excluded (mostly
temporary and intermittent workers) became covered by social security. Beginning
in 1962, automatic cost-of-living adjustments were added to CSRS.
As various new benefits were added to the Civil Service Retirement System
and as the system began covering more and more retirees and survivors, new in-
come to the CSRS Trust Fund was needed to finance benefits. Beginning in 1928,
Government appropriations were added to employee contributions to ensure that
adequate assets existed in the CSRS Trust Fund to meet the growing financial
demands placed on the system. Periodically, it became necessary to increase the
revenue flowing into the fund to assure its continuing ability to meet CSRS
obligations. Additional revenue was provided by increasing employee contribu-
tions and government payments through existing sources, and by establishing new
sources of income to the fund's account.
The current financing system began in 1969, with the passage of P.L. 91-93.
Under the provisions of P.L. 91-93, CSRS income derives from four main
sources:
1. Employee contributions. Employees covered by CSRS contribute 7 percent
of their basic pay through payroll deductions. (Congressional employees and haz-
ardous duty employees contribute 7 1/2 percent, Members of Congress contribute
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
8 percent, and in return receive more generous benefits). These deductions
are compulsory for most Federal workers; they are optional for Members
of Congress and congressional employees;
2. Matching contributions by employing agencies;
3. Interest on trust fund assets; and
4. Direct appropriations from the U.S. Treasury to fund additional lia-
bilities resulting from wage increases; extension of coverage to previously un-
covered employees; new or liberalized benefits, except COLAs; and credit granted
under CSRS for military service. An amount equal to the interest on the "un-
funded liability" (which will be discussed later) is also transferred from the
Treasury to the Trust Fund.
The current financing system ensures that annual trust fund income in a
given year always will exceed annual fund outlays that year--that the pay-as-
you go system will handle immediate needs each year. (As the discussion of
unfunded liability will show, the assurance that each year's trust fund revenue
will exceed its outlays does not mean that the plan's liabilities are "funded"
in an actuarial sense. Because P.L. 91-93 authorizes the automatic transfer
of funds from the Treasury to the CSRS Trust Fund to cover the last two cate-
gories of revenue, an annual appropriation by Congress is not required, except
to approve the matching contribution of each Federal agency contained in that
Agency's budget.
The current benefit structure of CSRS is as follows: Employees become
vested in the system after five years of covered employment. Workers covered
by CSRS are eligible for an unreduced nondisability pension at age 55 with 30
years of service, at age 60 with 20 years of service, or at age 62 with 5 years
of service (eligibility requirements and benefit levels for law enforcement
officers, firefighters, air traffic controllers, and Members of Congress and
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
congressional employees are more liberal than the.general provisions). Employ-
ees whose jobs are abolished or whose agencies are undergoing a major reorgani-
zation or reduction-in-force (RIF) are eligible for involuntary retirement at
any age with 25 years of service or at age 50 with 20 years of service, but bene-
fits are reduced by 2 percent for each year the retiree is under age 55 at the
time of his or her retirement.
Pensions are computed as a percentage of average annual salary during the
highest three consecutive years of earnings (High-3). Employees earn 1.5 per-
cent of their High-3 average annual salary for each of their first five years
of service, plus 1.75 percent of their High-3 average salary for each of their
next five. years of service, plus 2 percent of their High-3 salary for each year
of service thereafter. The total benefit may not exceed 80 percent of the re-
tiree's High-3 average annual pay; this ceiling is reached after about 42 years
of service.
CSRS'provides disability pensions to vested employees who are unable to
continue working due to disease or injury and who are not qualified for assign-
ment to any vacant position in the same agency at the same pay scale and grade.
CSRS also provides automatic survivor benefits for survivors of employees with
at least 18 months of service. Benefits for survivors of retirees are automati-
cally provided unless the retiree elects not to have such coverage at the time
of his retirement; the selection of survivor benefits results in a reduced
monthly annuity.
Retired pay and survivor annuities are protected against inflation. The
Federal Statute (5 USC 8340) requires that each March a cost-of-living adjustment
(COLA) equal to the increase in the Consumer Price Index (CPI) over the previous
calendar year be applied to the annuities of Civil Service retirees and their
survivors. In 1982, however, Congress suspended standing laws pertaining to
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
COLAs in pensions and survivor annuities for all Federal retirees for Fiscal
years 1983, 1984, 1985. The Omnibus Reconciliation Act of 1982 (P.L. 97-253)
provides that, during these years, COLAs will be effective one month later each
year (i.e., April 1983, May 1984 and June 1985), and that non-disabled retirees
under age 62 will receive a reduced COLA. Pending legislation would create a
new measuring period with COLAs applied in December of each year.
CSRS benefits are taxable as income after the beneficiary has received
monthly payments equal to the amount withheld from the employee's paycheck
(this portion of the benefit was taxed when it was originally earned). Benefits
generally become taxable after about 14 months.
Employees covered by CSRS are not currently covered by social security,
and do not pay taxes into the social security Old Age and Survivors Insurance
or Disability Insurance trust funds. Since the beginning of 1983, Federal
employees have been paying taxes into the Health Insurance Trust Fund and
earning credits toward medicare coverage. The Social Security Amendments of
1983 (P.L. 98-21), however, do extend for the first time social security cover-
age to all new Federal employees, and selected categories of current workers;
most current Federal workers remain exempt from such coverage. The Amendments
require the following categories of Federal civilian employees to be covered by
social security, in addition to CSRS, beginning on January 1, 1984:
1. All Federal employees hired on or after January 1, 1984, including
those with previous periods of Federal service if the break in Federal service
exceeds 365 days;
2. Congressional employees on the same basis as other Federal employees,,
as well as current congressional employees who are not participating in the
Civil Service Retirement System on December 31, 1983;
3. The Vice President and Members of Congress; and
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
4. Most Executive Level and Senior Executive Service political appoint-
The Social Security Amendments of 1983 also provide that retirees who be-
come eligible for both social security benefits and a government Pension after
1985 will be subject to a reduction in their social security benefits, phased
in over a five year period. Federal civilian employees exempted from this
reduction include current Federal workers and Members of Congress who become
covered by social security because of the Act, and workers with 30 years of
of social security covered employment; those with 25 to 29 years of covered
work will have their social security benefit reduced by a smaller percentage.
At the time the Social Security Amendments were passed in March 1983, no
new Civil Service Retirement System had been designed to supplement social se-
curity for newly-covered employees. Without the passage of additional legis-
lation, beginning January 1, 1984, Federal personnel affected by the Social Se-
curity Amendments would have begun paying 7 percent of their salary into the
social security system in addition to the 7 to 8 percent they already contri-
bute to CSRS. This formula would have placed an excessive burden on these Fed-
deral employees, and could have created problems in the recruiting and reten-
tion of personnel..1/
Legislation passed in November 1983 eliminated this problem. Title II of
the Federal Physicians Comparability Allowance Amendments of 1983 (P.L. 98-168),
known as the Federal Employee's Retirement Contribution Temporary Adjustment Act
of 1983, contains language governing treatment of Civil Service employees newly
1/ U.S. Congress. House. Committee on Post Office and Civil Service and
for the Senate Committee on Governmental Affairs. Effect of Requiring New Fed-
eral Employees to be Temporarily Covered by both~Social Security and Civil Ser-
vice Retirement. U.S. Govt. Print. Off., Aug. 31, 1983. (GAO/OCG-83-1)
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
covered by social security until a .modified CSRS incorporating social security
coverage is enacted. In general, the Act requires that Civil Service employees
brought under social security by the 1983 amendments make the full social se-
curity contribution, but have only 1.3 percent of their salaries withheld
for the Civil Service Retirement System over the next two years. Unless these
employees deposit the difference between the amount withheld at the 1.3 percent
rate and what would have been withheld at the 7 percent rate before they begin
collecting their pension, however, the benefit earned under CSRS during the
years in which they contributed only 1.3 percent of their salary will be offset
by that portion of their social security benefit earned during that period.
Members of Congress and Executive Level and Senior Executive Service
political appointees employed by the Government on or before December 31, 1983,
may select their level of participation in the Civil Service Retirement System.
They may choose not to participate at all in CSRS, in which case they will re-
ceive a refund of past contributions to*CSRS and not receive any pension from
that system; they may participate at the 1.3 percent rate, in which case they
will receive a pension offset by their social security benefit unless they de-
posit what would have been withheld at the normal 7 or 8 percent rate; or they
may participate at the full 7 or 8 percent rate, in which case they will receive
both full social security and full CSRS benefits. If they elect to forfeit
their right to CSRS benefits and receive a refund of their past contributions
to CSRS, they may resume their participation at a later date, but will not re-
ceive credit for their years of service before they terminated coverage unless
they first repay their contributions for those years, Dlus interest.
The Act also provides that any employee who first takes civilian office in
the Government on or after January 1, 1984, who participates in CSRS, and who
is affected by the Social Security Amendments of 1983 shall be subject to a new
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Government pension system when one is established. In addition,, any credit
earned toward the existing CSRS by any employee who is participating in CSRS
on or after January 1, 1984, and who is required to join the new CSRS will be
creditable under the new CSRS.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Only a few years ago features of the Civil Service Retirement System were
praised as a model for other pension systems. Recently, however, the CSRS has
come under attack with critics pointing to different program deficiencies.
During the 1970s, criticisms of CSRS as well as many other government programs
mounted. These were prompted by a combination of factors: (1) Public confidence
in the Federal government was shaken during the 1970s as the nation struggled
with a troubled economy, the end of the Vietnam War, and the trauma of Water-
gate. (2) Large Federal deficits focused increasing attention on the cost
of Federal programs; (3) the cost of the CSRS grew rapidly during this decade,
causing some critics to assert that it had become overly generous; and (4) in
combination with limits on Federal pay the CSRS made retirement especially at-
tractive and hampered efficient government management of personnel.
Until recently, critics made another major complaint against the civil ser-
vice retirement system. They objected to the exclusion of Federal workers from
social security. They said it was inappropriate that even the top level mana-
gers of the social security system, and those who legislated it, were not sub-
ject to its rules.
After several years of study and discussion, the social security amend-
ments of 1983 brought new Federal workers, as well as all House and Senate
members and top level Executive branch appointees into social security. How-
ever, in bringing new Federal workers into social security, thus silencing
one criticism leveled against it, Congress is left with the onerous lob of
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
designing a new Federal pension system for newly-covered employees. During
the debates over the Social Security Amendments of 1983, which brought
Federal workers into social security, several legislators said that Congress
intended to enact such a supplementary pension system. Senator Ted Stevens,
Chairman of the Senate Subcommittee on Civil Service, Post Office and General
Serevices, said, "Obviously a new pension system for Federal employees needs to
be developed to coordinate with the social security system." 2 /
Congressman William Ford, Chairman of the House Committee on Post Office
and Civil Service, stated that "...we believe that new Federal employees who be-
come covered under social security should be provided retirement benefits com-
parable to those under the civil service retirement system." 3 /
In addition to designing a new system, the Congress still is faced with con-
tinued criticisms of the current system: critics claim that it costs too much
and that it has adverse effects on the quality of the Federal workforce. Deal-
ing with these two problems is made more difficult because possible solutions
are intertwined with each other. For example, the designers of a new pension
system face the following dilemma: they must decide to what extent to pattern the
supplemental pension system after the current CSRS. If the new system replicates
the current one, it may be subject to the same criticisms. On the other hand,
if the new system is substantialfly less expensive than the CSRS, problems arise
of co-workers receiving unequal benefits--some more, some less.
Issues of designing a new system will be developed later in this paper.
To help lay the foundation for this later discussion of issues, the nature of
the criticisms against the current system are briefly identified here.
2/ Congressional Record, v. 129, Mar. 2, 1983. p. S4251.
3/ Ibid., p. H765.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
A. Major Criticism: Too Costly and Too Generous
Most studies done to date agree that the current civil service retirees
receive more costly benefits than their counterparts from the private sector.
In a recent study published by the Brookings Institution, Dr. Robert
Hartman, now of the Congressional Budget Office, stated:
"The conclusion to be drawn... is inescapable. The CRS system
is much more costly to the Federal government as employer (and to
the taxpayer) than the'combined social security and private pension
package is to employers in the private sector." 4 /
For the past two years the Office of Personnel Management has barraged the
public with prese releases, Director's speeches, and other instruments to make
the following point: The Civil Service Retirement system is too costly and
should be cut. For example:
...at one time, more generous benefits for Federal employees
could be justified on the basis that Federal employees were underpaid
compared to the private sector. However, several public and private
studies now indicate strongly that Federal employees no longer trail
the private sector in pay..." 5 /
The Administration, acting on the perception that CSRS is too generous, has
proposed changes that would reduce the cost of the CSRS by 37 percent and in-
crease employee payments to finance the system by 57 percent. 6 /
When critics voice the concern that the current CSRS is too costly, they
may be referring to any one or a combination of the following points:
1. The absolute dollar amounts are large. Outlays for Civil
4/ Hartman, Robert W. Pay and Pensions for Federal Workers. The Brookings
Institution, 1983. p. 71.
5/ Devine, Donald. Director of the Office of Personnel Management. Office
of Personnel Management, press release. Feb. 22, 1983.
6/ U.S. Office Management and Budget. Fiscal Year 1984 Budget. Income
Security Section. pp. 5-119.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Service retirement will exceed $21 billion during the current fiscal year..
A recent OPM press release stated that the CSRS is the fourth largest entitle-
went program, ranking only behind OASDI, Medicare and Medicaid. The release went
on to point out that it may exceed Medicaid and rise to third place and that CSRS
outlays are greater than AFDC and food stamps combined. It is not clear why the
CSRS is being compared to food stamps and AFDC rather than the Federal payroll,
or the military payroll for that matter. But in a time of $200 billion budget
deficits, the size of all programs becomes an issue.
2. Growth has been rapid--both real and nominal. Program outlays
were $2.5 billion as recently as 19.70, and increased by 607 percent in the next
12 years. This substantially exceeds the growth of GNP (213 percent) and growth
in total budget expenditures (272 percent). Looking at "real" rates of growth
(with the effects of inflation removed) also shows CSRS outlays increased more
rapidly than GNP during this period. CSRS outlays increased by 208 percent com-
pared to GNP at 36 percent and the entire budget at 61 percent in real terms.
Rapid rates of recent growth often leave the specter that growth may con-
tinue at such high levels in the future.
However, CSRS outlays are unlikely to grow faster than the economy in the
future. The rapid growth of the past decade and a half resulted from (1) the
maturing of the pension system as the large buildup of employees of the Great
Depression years and World War II came onto the retirement rolls; (2) the effects
of benefit improvements enacted after World War II, including full cost of
living protection; and (3) the upward effects on Federal pay of rapid economic
growth during the 1950s and 1960s which increased subsequent pension amounts.
These conditions are unlikely to be repeated. In fact, based on current esti-
mates, the CSRS outlays as a percent of GNP would decline over the next years,
from 0.66 percent in 1983 to 0.60 percent in 1988.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
S ?
3. CSR costs are high relative to similar pension practices in the
private sector. This point will be discussed in detail later in this paper.
Studies to date almost unanimously favor the position that the CSRS does more
than pension systems that provide benefits to similar employees in the private
sector.
4. The unfunded liability keeps growing. Although not directly
related to cost, OPM has characterized the CSRS's unfunded liability as repre-
senting a menace to the economic well-being of the nation. This colorful point
also will be discussed in greater detail in later sections. Unfunded liability
sounds like a straightforward concept, but it is not. It is a complex actuar-
ial concept that only those in that fraternity completely understand; and they
may disagree on its implications and measurement. This paper.maintains that
it will be more productive to concentrate on the total size of the obligation
or liability incurred by;the CSRS than to focus on the question of what share
of this total obligation to label "funded" as opposed to "unfunded." 7 /
B. Other Criticisms: Reduced Efficiency of the Workforce
A different group of CSRS critics maintain that the retirement system
reduces the efficiency of the Federal workforce. Experts agree that the type of
pension system does affect workforce type whether in the private or public sec-
tor. In the private sector it is a common practice to design a comprehensive
compensation package of wages, fringe benefits and pension rights, that aims to
7/ Munnell. Alicia. The Economics of Private Pensions. Discusses funded
and unfunded liabilities and demonstrates that given a particular size of obli-
gations, different actuarial methods will yield different proportions classified
as funded (as opposed to unfunded). Brookings Institution, 1982.: pp. 150-154.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
yield a desired kind of workforce. 8 /
The current CSRS has been criticized because the retirement system and the
way it interacts with the Federal pay system tend to restrict mobility of work-
ers into and out of the government, as well as to induce the highest level gov-
ernment managers to leave government service as soon as they are eligible to be-
gin drawing their pensions--often during their most productive years.
An employee with 15 to' 20 years of work for the Federal government faces
the likelihood of severe financial loss if he leaves Federal government before
becoming eligible for a pension. If he leaves, he is allowed to withdraw his
own contributions, but he receives no interest and no part of the matching
sums paid. for his pension by his employer, the Federal government. At age 50
after 20 years, an employee's own contributions would amount to only about one-
third of the total benefits accrued by him.
Another alternative upon moving to other employment outside the government
is to leave the contributions in the system and to begin drawing an annuity upon
eligibility. This would be at age 60 with 20 years service and at age 62 with
more than 5 years of service but less than 20 years. In either case, the amount
of initial benefit would be determined by salaries earned at the time the'Federal
employment was ended, with no allowance for wage or price rises thereafter. De-
pending on the rate of inflation and the number of intervening years, the real
value of the initial benefit might shrink dramatically. There are two way to
avoid this loss: (1) Employees can wait to leave Federal employment until they
can establish eligibility through involuntary early retirement or disability;
or (2) employees can go back to work for the Federal government before retirement
8/ Rock, Milton. Handbook of Wage and Salary Administration. McGraw-Hill,
1972. p. 6-4.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
to build up the salary base upon which benefits are calculated.
The current CSRS interacted with the Federal pay system to lose the Feder-
al government many of its most experienced employees during recent years of
high inflation. Many top-level government managers retired as soon as eligi-
bility was reached; for many at the peak of their productive careers. In recent
years retirement benefits have been fully indexed for inflation and more, (with
what was called a "1 percent kicker" and semi-annual adjustments) but Federal
salaries have been held down and have fallen in real value. Thus, after a few
years some top level civil servants who had retired were able to receive more
in retirement benefits than they would have been paid in salary if they had con-
tinued working.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
To summarize, Civil service retirement looms as a major issue for the Con-
gress for two reasons: (1) recent criticisms leveled at the cost, generosity,
and other features of the current system, and (2) the need to design a supple-
mental pension system for Federal workers recently covered by social security.
This includes not only all new Federal workers, but some current workers--inclu-
ding Members of Congress, and at their choice, congressional staffs.
The remainder of this paper will discuss in more detail and attempt to
provide a simple structure for thinking about issues in civil service retire-
ment--as they affect the current system, the design of a new system, and as
the interaction of the two. This discussion of issues will be organized a-
round three themes: (1) workforce goals, (2) adequacy and equity of benefits
and (3) costs and financing.
Most Federal programs, and civil service retirement is no exception, have
more than one goal.. The goals of civil service retirement are not unique to it,
but characterize other retirement income systems as well. Nor are these newly
discovered goals. Proponents of social insurance in general, and of a pension
system for civil servants, discussed the different objectives during debates on
both subjects early in this century. The basic arguments have not.changed much
since then, although at different times different goals have received emphasis.
The objectives of civil service retirement have been, first, to support the de-
sired type of workforce by either making it easier or more difficult to move
from Federal to non-Federal employment; second, to meet the goals of adequate
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
?
and equitable income for retired persons; and third, to do so at a cost that bal-
ances the interests of those receiving benefits to taxpayers paying the costs.
A. Workforce Goals
In managing an organization's workforce, specific organizational policy is
directed to inducing or reducing employee turnover. This question cuts both
ways. In the first place, organizations (public or private) incur substantial
training costs whenever they hire new employees. These costs should be amor-
tized over the productive years of employment, and the organization suffers an
economic loss if the employees leave during their productive years. Thus, it
is in the organization's interest to design compensation packages (including
deferred compensation or pensions) to induce employees not to seek other employ-
ment. On the other hand, it is in the interest of some types of employers to be
able to have employees leave at relatively young ages. Examples include the mili-
tary and police and firefighters. As the military puts it, they have a need for
a work force with sufficient "youth and vigor." The deferred compensation pack-
age designed to encourage rapid turnover will look much different from one de-
signed to encourage experienced workers to stay on the job. In the military,
for example, the pension benefits (referred to as "retired pay" in the Federal
laws) are not vested for 20 years. People who leave the military with fewer than
20 years of service get nothing from the pension system. However, after 20 years,
the employee not only vests, but is eligible to start drawing benefits immediately
upon ending his employment with the military. Thus, a person with 20 years of
military service, regardless of age, can retire and begin receiving benefits
(at 20 years the benefits will be one-half the previous base pay.) The average
age for enlisted men to begin receiving retirement benefits is 39 and for offi-
cers is 43.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
Some rather arcane pension concepts describe the techniques used by pen-
sion planners to design a plan that meets the particular policy goals of an
organization's workforce management policy. The most important of these con-
cepts is "portability." "Portability" means different things to different peo-
ple. Some like to equate it with social security where a worker can change jobs
and continue to earn benefits under social security. In its technical sense,
portability refers to the ability to transfer the present monetary value of an
individual's vested pension benefits to a succeeding plan or an Individual Re-
tirement Account (IRA). Emphasis.is placed on the word "vested" because unless
the participant has a vested right in the pension, there would be no credits to
transfer (i.e., the terminating employee would forfeit any benefits that had
been accrued.) The practice of portability between employer-sponsored plans is
virtually nonexistent in the private sector. However, defined contribution
plans (those whose benefits are determined by amounts contributed and interest
earned) often offer a terminating employee the lump sum credited to his pension
account upon terminating employment. Also, companies offering defined benefit
plans "cash out" small vested benefits by offering terminating employees a
lump sum reprsenting the present value of the future benefit. While amounts re-
ceived can be "rolled over" (invested) tax free into an IRA and maintained on a
tax deferred basis until withdrawn at retirement, company pension plans gener-
ally do not have provisions for accepting these amounts upon hiring the em-
ployee. It should be noted, however, that there are about 2,600 multi-employer
pension plans operational in certain industries and/or geographic areas. A
multi-employer plan is a collectively-bargained plan where many employers par-
ticipate in the same plan. Many of these plans are large.. Employees can change
jobs and keep earning pension credits and service towards fulfilling the plan's
vesting requirements providing that they continue working for another employer
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
participating in the multi-employer plan. However, since their value is based
on salary level (in addition to years of service), price inflation after leav-
ing a job erodes the value of initial pension benefits. Civil service retire-
ment benefits are portable among government agencies, but not between government
employment and employment outside the Federal government.
A second feature of pensions that can influence workforce behavior is their
"accrual rate"--the rate, expressed as a percent of pay, at which benefits are
earned. Just as employees receive specified amounts of cash income, so also
they receive specified amounts of non-cash deferred compensation. And as the
amounts of cash income differ from company to company, so also do the amounts of
deferred compensation. Furthermore, it is possible for deferred compensation to
be earned at different rates depending on how long employees have been with the
organization. The way this usually works is to multiply the accrual rate times
the years of service, times salary (can be final salary, average of the highest
three or more years, or an average over the years of work). These rates have'
a great effect on costs and replacement rates of pension systems. They also
affect the workforce turnover, since some employers have larger accrual rates
for longer periods of employment. A pension system whose benefit accrual rates
vary directly with years of service, such as.the CSRS, is called "backloaded."
Federal workers have an accrual rate of 1.5 percent for the first 5 years of
employment, 1.75 percent for the second 5 years and 2.0 percent for all employ-
ment over 10 years. A backloaded pension is thought to tie employees with
lengthy service more tightly to the employer.
Features that provide strong financial incentives to stay with an employer
have been called "golden handcuffs." This term has taken on a pejorative mean-
ing in recent years as the civil service retirement system, along with government
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
in general, have come under attack. However, strong "golden handcuffs" should
not be criticized unless they hinder the desired goal. If the policy decision
is that Federal employment, or certain categories of Federal employment, should
be looked upon as a lifetime career, and that long time experienced employees
are most capable in carrying out the government's business, then the pension
system should continue to tie employees to government service with "golden hand-
cuffs."
On the other hand, if the policy decision is that the government would
function better if there were more people moving back and forth between the pri-
vate sector and the Federal government, then the pension system should, and
could be, changed to promote this end. Valid arguments can be made for both
points of view, and it is possible to design a pension system to promote either
end. The point to be emphasized is that whether an explicit decision on this
policy issue is made or not, the design of the pension system will influence
the type of workforce that results.
The current CSRS gives employees guaranteed rights to benefits (vested
rights) relatively early, but the benefits earned are not portable outside the
Federal government. Hence, Federal workers are more "tied" to their jobs than
are workers with social security coverage, or other portable pension rights.
It follows from this analysis that covering Federal workers by social security
will, by itself, increase mobility in and out of the Government because of the
portability of social security benefits.
B. Benefit Adequacy
The second objective for civil service retirement, as well as for many
other pension programs, and for the largest social insurance and income transfer
programs, is to help finance a decent standard of living for persons too old to
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? is
continue working for pay. This goal raises questions of measurement and compari-
son; this goal also requires the examination of CSRS in a broader context that
takes into account all the sources of income that are or should be available to
retired workers, their dependents and their survivors.
A "three-legged stool" has become a common metaphor for visualizing a poli-
cy model of sources of retirement income in the United States. Social security
makes up the first leg. From the beginning it was not intended to supply total
retirement income, but be a base only. When he signed the Social Security Act
in 1933, Franklin Roosevelt said,
"We shall make the most orderly progress if we look upon Social
Security as a development toward a goal rather than a finished product.
We shall make the most lasting progress if we recognize that Social
Security can furnish only a base upon which each one of our citizens
may build his individual security through his own individual efforts." 9/
Individual savings of all kinds make up the second leg. These can take the
form of savings accounts, financial or other investments, life insurance annui-
ties, IRAs and similar programs, and a variety of other means. Pensions consti-
tute the third leg.
Analysis of civil service retirement should begin with the broad policy-
oriented veiw rather than a program-oriented view, with program adequacy consid-
ered in the context of overall income sources to the retired, insofar as this is
possible. In other words, analysis of program adequacy should consist of more
than one leg of the stool. At the very least, the Federal pensions for employ-
ees newly covered by social security should combine the two in assessing benefit
adequacy.
The first issue to address is how to establish the proper level of adequacy.
9/ The Public Papers and Addresses of Franklin D. Roosevelt. Social Securi-
ty Amendments. V. 8, 1941, p. 80.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? ?
In meant-tested and social insurance income transfer programs a widely-used
measure is the poverty level. One objective of these programs is to transfer
relatively large amounts to those in poverty or near this level. Social se-
curity, while not a means-tested program, nonetheless has deliberate fea-
tures of income distribution built into it, and has from the beginning. In
fact, such redistributive effects define, in part, what social insurance is.
Thus, the social security benefit formula provides more for lower-income earn-
ers than it does for higher, as a percentage of total preretirement earnings.
The effects of this policy are seen by noting that the poverty rate among the
elderly has declined over the past two decades until it now is no higher than
for the population as a whole. But without the social security program, the
percentage of the elderly in poverty would be 250 percent higher than it is to-
day. 10/ The share of final earnings replaced by social security, the "replace-
ment rate," is substantially higher for lower income than it is for higher in-
come beneficiaries.
Pensions, on the other hand, do not typically use the poverty level as a
measure of adequacy, or income redistribution from high to low income as a poli-
cy target. Instead, replacement rates have been widely accepted as a measure
of pension benefit adequacy. Replacement rates are calculated by dividing the
amount of pension benefit received upon retirement by the amount of earnings
received just prior to retirement. The analysis in this paper will distinguish
between "gross" replacement rates (before taxes and other expenses incurred as
a result of working) and "net" replacement rates (after taxes and work expenses).
.. 10/ U.S. Congress. House. Ways and Means Committee. Background material
on Poverty. Committee Print, 98th Conk;., 1 st Sess. U. S. Govt. Print. Off.,
Oct. 17, 1983. p. 107.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? 0
Net replacement rates are higher and offer a more accurate comparison of pre-
and postretirement income standards of living.
In its 1981 report, President Carter's Commission on Pension Policy recom-
mended the establishment of a national policy goal that the level of retirement
income available to individuals be sufficient to maintain the same income stan-
dard of living they had while working. Such a policy would not be redistribu-
tive but would assume roughly the same distribution of net income (after taxes
and work-related expenses) as*exists among the working age population. Indeed,
it does even more. It assumes the pensions can be designed to "integrate"
with social security in such a way that the redistributive effects of social
security can be offset. In other words, with an "integrated" pension, higher
income workers will receive proportionately more in pension benefits than will
lower income workers, just as lower income workers receive proportionately more
from social security than higher income workers. The current civil service re-
tirement system, not being "integrated with social security, is neutral with
respect to its effects on workers with different income levels. Thus, all civil
service retirees with the same number of years of service, regardless of their
salary level, will receive the same proportion of their salaries, whether low
or high, as a retirement benefit.
One of the most difficult issues in designing a new pension system for Fed-
eral employees covered by social security will be how to deal with the "tilt"
in the social security formula. This "tilt" produces a higher replacement rate
for lower earning workers than it does for higher earners. It works like this:
to calculate social security benefits, an average is taken of an employee's en-
tire earnings history covered by social security. (Each year of earnings is
brought up to the present by increasing each year's amount by a factor which
reflects changes in wage rates over time.) Once this average annual lifetime
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
earnings level is calculated, the actual monthly benefit amount is received by
applying different factors to it. The calculations for the workers earning at
the maximum level covered by social security ($35,900) would be as follows:
90 %
of the first ............ $ 254, equals
$ 229
32 %
of the next ............. $ 1274, equals
$ 408
15 Z
of the remainder over ... S 1528, equals $ 63
Thus, the monthly benefit equals $ 700, for earnings at the social security
earnings maximum, yielding a gross replacement rate of 23 percent. The distri-
butional effect of the tilt is shown by comparing replacement rates for this
maximum earner with the low-income and middle-income earner, for social security
only, as shown in the following table:
Low Income Middle Income Maximum Income
Gross replacement
rate 47 %
Such distributive effects are inherent in the nature of social insurance
programs. While it is true that the lower income workers pay social security
taxes on the entire amount of their earnings in contrast to higher income
workers, actuarial calculations indicate that the lower income workers still
recieve a substantially higher "return" on their contributions than do the high-
er income workers. 11 /
The decision facing the Congress in designing a new system, as it faces
all employers with a pension system supplementing social security (and this in-
cludes State and local governments as well as private employers) is what to do,
11/ U.S. Congress. House. Ways and Means Committee.' Staff Data and
Materials Related to Social Security Financing. Committee Print, 98th Cong.,
1st Sess. U.S. Govt. Print. off., Jan. 27, 1983. pp. 45-49.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
if anything, about this income-redistributing tilt in social security.
There are three basic options: (1) Completely offset it. In other words,
offset the pension benefit dollar for dollar with benefits from social security.
(2) Ignore it; do not take it into account at all. In other words, simply add
the amount of any supplementary pension on top of social security. This would
retain the effect of the social security tilt in final retirement income (social
security plus pension benefits). (3) "Integrate" the pension with social securi-
ty in such a way as to offset part of the tilt, but not all of it.
1. Completely offsetting social security. It is technically possible
to design a supplemental pension system that, when combined with social security,
would exactly replicate the current civil service retirement system. Some state-
ments made at the time the 1983 social security amendments were being debated sug-
gested this. possibility for the Federal pension system to supplement social se-
curity for newly covered workers. The technical procedure would be simply to off-
set, dollar for dollar, any social security benefits in calculating the amount
of the supplemental pension, with the objective of making the total amount equal
to what would have been received if the current civil service retirement had
continued in effect.for new Federal workers. The effect of this would be that
the lower paid Federal workers would receive very little of their total benefit
from the new civil service retirement supplemental pension and almost all of
their total benefit from social security. On the other hand, the highest paid
Federal employees would receive most of their total benefit from the civil ser-
vice retirement supplementary pension. The reason is that the current civil
service retirement system has no tilt in its benefit formula, but returns pen-
sions.of equal amounts, relative to earnings, at all levels of earnings.
Internal Revenue Service rules governing private pensions do not permit a
plan to fully offset social security benefits. Both employers and employees
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
S ?
receive income tax advantages from the deferred compensation in the form of pen-
sion benefits if the plan is "qualified," based on standards specified in the
Internal Revenue Code. The IRS would not qualify a plan as eligible for the tax
advantages if it were set up to offset completely the tilt features of the social
security program. Of course, the new civil service pension does not necessarily
have to meet the same "integration" standards as private pensions, but such a
plan would be subject to criticism if it did not.
2. Add-on plans. At the other extreme, a pension plan could be de-
signed so its benefits would simply be additive to social security. As in the
current CSRS, the supplemental pension could pay a flat percent of final salary,
varied only by length of career. Of course, if the new pension system were to
maintain the cost of the current system, the amount of this supplemental pension
by itself would have to be substantially smaller than the current CSRS; but
when combined with social security the average benefit could be about the same.
If this same cost as the current CSRS system were used, the effect would be for
lower income workers to receive substantially more than workers at the same in-
come in the current system, and for higher earning employees under the new sys-
tem to receive smaller retirement benefits than comparably paid employees in the
current retirement system.
This practice, while rare for salary-related plans in the private sector,
is fairly common among state pension systems. 12 / Of 50 state systems, 28
have pure "add-on" plans. One reason for this different behavior is that many
state plans were already in operation when the social security act was passed
in 1935. Furthermore, according to a study by the General Accounting Office,
12/ National Association of State Retirement Administrators. Survey of
State Retirement Systems. Sept. 30, 1982.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
is ?
many state governments prefer this type of arrangement because it is easy to
understand and explain to employees and taxpayers alike. 13 / Any of the
other methods of coordinating pension benefits with social security can be
interpreted by lower income workers as having something taken away from them.
On the other hand, add-on plans reduce benefits of higher earners, compared to
a plan like the CSRS.
3. Integrating social security with pensions. Integration is a
technical term referring to procedures for designing pension plans to offset
part of the effect of the social security benefit tilt on final combined retire-
ment income (pension plus social security). There are different ways to do
this, depending on how much of the social security tilt pensions designers want
to offset and at what levels of income. 14 / For purposes of this paper, it is
sufficient to note that integration permits the design of a pension system with
replacement rates across the entire range of income levels falling between those
resulting from pure offset plans and pure add-on plans.
Of the two principal bills considered to date for a supplementary pension,
one is an add-on and one an integrated plan. Congressman Ehrlenborn has proposed
a pure add-on plan. While this plan has not been costed out, it probably will
be more costly than the current system. This bill includes a thrift plan along
with its defined benefit add-on pension plan. Senator Stevens has proposed a
defined contribution plan with step rate integration, also to be accompanied by
a thrift plan. These thrift, or capital accumulation plans, fairly common in
13/ U.S. General Accounting Office. Federal Employee Demographic and
Integration of State Retirement Plans with Social Security. CAO/FPCD-83-38,
July 27, 1983.
14/ U.S. Library of Congress. Congressional Research Service. Inte-
grated Pension Plans: An Analysis of Earnings Replacement. Typed report,
by Ray Schmitt. Jan. 5, 1981. Washington, 1981.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
is ?
the private sector, can take many forms. The most usual is for the employer
to match at specified rates, voluntary contributions to retirement programs by
employees. Private sector practice supports the intuitive notion that these
would be used more by higher income workers to offset their likely loss of bene-
fits compared to the current system under a pension system integrated with soc-
cial security. 15 /
The discussion of benefit adequacy to this point has been general and ab-
stract. As the debate proceeds, estimated replacement rates for specific design
options can be used as one way of determining their effect and of choosing among
them.
A computer-based model, developed at the Congressional Research Service
(CRS), has been used to estimate replacement rates for different combinations of
pension benefits, including the CSRS. This model is currently being updated to
accomodate the many program change's over the past year, particularly in social
security. Data used in the following analysis are slightly out-of-date, having
been done about one year ago. Nonetheless, they will indicate the magnitudes
involved and will demonstrate how such replacement rate data can be used in do-
ing analysis of proposed Congressional options for designing a new, or for mak-
ing changes in the current, retirement system.
A number of researchers have made estimates of the replacement rate neces-
sary to maintain living standards in retirement. Most agree that retirees need
less than 100 percent of gross earnings replacement to maintain their standard
of living. While no precise earnings replacement objectives have been estab-
lished as a matter of national policy, researchers generally agree that indi-
viduals need between 60 and 80 percent gross preretirement earnings to maintain
15/ Hay/Huggins. Noncash Compensation Comparison. Hay Assocaites. 1983.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
preretirement living standards--with rates at the higher level more appro-
priate for lower income individuals.
Net earnings replacement rates permit a better comparison between pre-
and postretirement income. Differences in tax liability before and after re-
tirement and reduction of work-related expenses such as social security, pay-
roll deductions, transportation, clothing and meals purchased away from home
are taken into consideration. In addition, differences in discretionary con-
sumption and savings, and possible reductions in living expenses in later life
(resulting from lower home financing, educational expenses for children, and
health care), influence living standards.
The following tables provide some rough calculations of replacement rates
for hypothetical workers receiving benefits from different pension systems, in-
cluding civil service retirement. Since tax treatment and other expenses vary
substantially before and after age 65,' net replacement rates are shown. Replace-
ment rates were calculated for both a "low income" worker (final salary of
$16,300 in 1982 dollars) and a high-income worker (final salary of $53,900).
Since social security has provisions for dependent spouses, replacement rates
are shown for both single workers and workers with dependent spouses. These
rates are for the year immediately after retirement and for the same workers 15
years later at age 80. The extent of cost-of-living indexation of benefits and
tax treatment varies from program to program. Without cost-of-living protec-
tion, the value of benefits in real terms declines.
When the replacement rates are higher than 100 percent, the after-tax in-
come of the retiree is actually greater than before retirement. This is true
for lower income workers with dependent spouses because social security provides
this extra amount for dependent spouses and social security enters into the pic-
ture in all programs,(except civil service retirement). Social security by
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
itself (not shown here) does not reach the 100-percent level although it comes
close for low-income workers with spouses. This underscores the notion that so-
cial security should not be looked upon as the sole source of retirement income.
Table 1. AFTER TAX REPLACEMENT RATES (AGE 65 WITH 40 YEARS OF WORK)
Low earner Higher earner
Single Married Single Married
Civil service retirement
(no social security) .......... 94.6 82.5 118.7 98.1
Military retirement
.(includes social security) .... 106.3 127.8 89.6 92.0
Railroad retirement
(includes social security-
type benefit) ................. 84.7 119.5 64.8 86.9
Social security plus
private pension ............... 102.4 117.2 94.4 92.0
Source: Financing Work-Related Entitlement Programs. Prepared for the Senate
Committee on Budget, by the Congressional Research Service. The calculations
are based on a common set of work histories and earnings records. They were
patterned after salaries of Federal workers. The low-income worker had a
final salary of of $16,300 and the high earner a final salary of $53,900.
Civil Service retirement benefits are not taxed until the amount of employ-
ee contributions has been paid back. Ordinarily this occurs after 1 or 2
years. Thus, replacement rates shown here are higher than they would be 1
or 2 years hence. For economic and demographic assumptions, see pp. 52 and
477.
As Table 1. shows, the programs compared here provide very high replacement
rates for individuals with long, consistent work histories. This analysis sug-
gests that the pension and social insurance system provides adequate benefits
(as adequacy is defined here) for those with long and consistent attachment to
the work force, whether they have low or high income.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? ?
In general, the programs compared in these tables provide after-tax income
replacement at a level fairly close to the level of income before retirement,
whether high or low. Except for civil service retirement, the programs shown
in Table 1. all include a combination of social security and another type of
pension. For the combined systems, married couples receive substantially higher
replacement rates as a result of social security's provision of benefits to de-
pendent spouses based solely on the earnings record of the working spouse.
The effect of social security is also seen in the difference between high-
and low-income workers. Social security's benefit formula "tilt" shows up as
higher income replacement for the low-income workers, except for civil service
retirement, which has no social security component. High-income earners under
civil service retirement actually receive higher income replacement than do low-
income workers, because the difference in tax treatment makes a bigger differ-
ence between preretirement and postretirement earnings.
Of the examples considered here, military retirement and civil service re-
tirement have their benefits fully indexed to changes in the cost of living.
Thus, 15 years after retirement, their gross benefits have maintained the same
relationship to gross preretirement earnings as they had upon retirement. For
the railroad retirement and private pensions, however, only the social security
component is fully indexed. The private pension counterpart in railroad retire-
ment and typical private pensions are not fully indexed, and thus over time the
size of the retirement benefit in relation to preretirement income shrinks.
This reduction would amount to about one-fourth in 15 years for high-income
railroad retirement and private pension beneficiaries. (See Table 2.) For lower
income recipients, whose overall benefits include a relatively larger social se-
curity component, the shrinkage is less, about one-seventh for railroad retire-
ment and one-fifth for private pensions. Military retirement and civil service
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
retirement suffer no diminution at all, since their benefits are fully in-
dexed.
Table 2. AFTER-TAX REPLACEMENT RATE (SINGLE WORKER)
At age 65
At age 80
Low
earner
High
earner
Low
earner
High
earner
Civil service retirement ..........
94.6
118.7
82.7
88.8
Military retirement
(includes social security) ......
106.3
89.6
106.3
89.6
Railroad retirement
(includes social security-
type benefit) ..................
84.7
68.8
72.6
50.3
Social security plus
private pension .................
102.4
94.4
82.5
71.3
Source: See Table 1, p. 33. The differences shown in the table for civil
service retirement result from tax treatment. Civil service retirement
benefits are taxable but are not taxed until the amount of employee con-
tributions has been paid back, ordinarily a period of 1 or 2 years. Thus
replacement rates shown are higher at age 65 than at age 80.
The difference early retirement provisions make in replacement rates is
highlighted in Table 3. Civil service retirement and military retirement, both
of which allow early retirement with unreduced benefits, show high replacement
rates even with early retirement. (The hypothetical private pension Plan in-
cluded in these examples for comparison purposes permitted retirement at age
55, but with substantially reduced benefits. Since social security benefits
do not start until at least age 62, at age 55 this shows only the effect of
the private pension.) By age 80, the private pension and military retirement
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
recipients have begun receiving social security in addition to their other pen-
sion, thus boosting their overall replacement rates.
Table 3. AFTER-TAX REPLACEMENT RATES (AGE 55 WITH 30 YEARS OF WORK)
Age 55 Age 80
Low
earner
High
earner
Low
earner
High
earner
Civil service retirement .........
68.3
83.6
60.2
60.8
Military retirement
(includes social security
at age 62) ....................
65.0
74.0
104.6
95.6
Private pension plus
social security ................
16.3
29.8
42.8
29.8
Source: See Table 1, p. 33.
In addition to giving some indication about the size of replacement rates,
the previous analysis should serve to suggest the difficulty of this type of
analysis. For example, replacement rates would seem to be adequate for those
with 40 year work histories who retire at age 65. Shorter work histories and
younger retirement ages, probably more typical now, would show lower.replace-
ment rates. In addition, the replacement rate for the CSRS is much higher
at retirement than it is just a year or two later when the benefits begin
being taxed. Replacement rate analysis can and should be done as one way to
assess changes and options. The presentation here should be viewed only as
an introduction to the topic, suggesting the significance of this type of
analysis.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?.
4. Value of Benefits. Replacement rate analysis is only one way to
assess benefit adequacy. Another, not used as much, can look at the total dis-
counted value of benefits earned at the time of retirement. over a lifetime
of work, an individual accrues rights to a stream of benefits to be received
upon retirement and continuing on until death. This stream of benefits can be
thought of as an asset whose value can be calculated. The asset value is the
sum of the present value of-.the future stream of benefits, plus the compounded
value of benefits already received from programs permitting early retirement.
The following chart displays these asset values for workers retiring at age
65 with 40 years of work and a final income of about S42,600 in 1982 dollars
(salary of the "middle" income Federal workers, GS-13, used in this example for
all programs.)
VALUE OF BENEFIT AT AGE 65
1,200
1,000
Past Government Pension
Future Government Pension
$120.000
Future Private Pension
M;,
$410,000
M.
w:~"? o .a :v :> ~;,...: ir,. '~~ ^"%C. w~X;J'.l::::.
.r:B ,n iii: '"!ai: . ia::;.? '.
Social Railroad Private Civil Service Military Civil Service
Searity Retirement Pension Retirement + Private + Private
Source: See Table 1,, D. 33.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
$1,110,000
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Benefits earned by a 40 year, 65 year old Federal government retiree are
estimated at about $500,000, only slightly higher than his counterpart in the
private sector. These values are substantially above social security by itself
at $120,000--driving home the point once again that social security should
not be thought of as sufficient retirement income by itself. The railroad re-
tirement private pension component is less generous than the other programs con-
sidered for these medium-income workers. The value of the benefit to the hypo-
thetical railroad retiree would be $260,000.
The variation in total program benefits is seen most dramatically for pro-
grams permitting early retirement benefits. For the military retiree, benefits
have been.received for 20 years, compounding to yield an amount which when com-
bined with the future annuity amount, is over $1 million. Early retirement
benefits in civil service retirement are added to future annuities in the same
way, except that it would only be for 10 years. The combination of compounded
payments to age 65 together with the present value of future benefits would to-
tal approximately $670,000 for civil service. These past benefit payments add
greatly to the overall asset values and explain why these two programs are more
costly than the others, as will be discussed in the next section.
In summary, income adequacy as a goal raises difficult program design
issues, particularly in comparing features of the new system with those of the
current system. If the measure used is replacement rates, it will be difficult
to design.a pension program which meets private sector standards for IRS quali-
fication and at the same time provides replacement rates to higher earning em-
ployees at the same level that the current system does without exceeding the
cost of the present system. If the Congress decides on a pension system, which
when combined with social security costs less than the current system, and if
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
the program is to be "integrated" with social security, then higher earnings
employees under the new system could have substantially lower replacement rates
than their current system counterparts.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
1. Measurement of Costs. During these years of $200 billion budget
deficits stretching as far as the eye can see, the cost of all Federal programs
has fallen under close scrutiny. Section II above noted that the cost of the
civil service retirement system has become a serious issue in recent years.
There are several reasons-: (1) Most-studies have concluded that the CSRS costs
more than prevailing practice in the private sector; (2) Outlays for the CSRS
have risen rapidly since 1970; (3) The absolute size of the program is relative-
ly large and it falls in a budget category along with other entitlement programs
that are considered to be relatively "uncontrollable" spending; (4) The size of
the "unfunded liability" has been highlighted by Office of Personnel Management
news releases and (5) Employees pay one-fifth of the cost of their pension, with
the taxpayers paying the other four-fifths.
Accompanying this list of cost problems is a list of cost concepts nearly
impenetrable to the layman. It is safe to predict that during the upcoming de-
bate on civil service retirement different cost totals and concepts will confuse
the issues.
An additional barrier to understanding emerges from the difficulty of dis-
tinguishing the issues of cost from those of financing. In other words, how much
the system costs is a different question from determining how to pay its costs.
When the concept of financing enters the picture, questions about meaning and
effects of trust funds and the meaning and effects of unfunded liabilities, al-
so enter.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
S 0
The analysis in this paper refers to the cost of civil service retirement
as the discounted value of benefits earned. This is a simple concept but because
of the different ages of employees and different amounts of benefits earned at
any particular time, agreement on a particular concept to measure the cost of a
pension system is not clear cut. The cost concept emphasized in this paper is
what actuaries call the "entry age normal cost" system. This methods starts
with a group (cohort) of employees lust starting work and calculates based on
the rules of the current pension system, how much they will "earn" in total
benefits until they all retire or otherwise leave the system, and until the last
one dies. These-earned benefits are discounted to the present using assumed
rates of interest and this discounted value of benefits is compared to a similar-
ly discounted value of the projected total wages and salaries paid to the same
group of employees over their working years (with specific assumptions for how
many become disabled, how many retire early, etc.) The resulting ratio (dis-
counted benefits divided,by discounted wages and salaries) is the cost of the
benefits--stated as a "percent of payroll." (These long run costs are best
stated as a percent of payroll because absolute dollar figures, even with rela-
tively modest rates of inflation, become so large far into the future that the
frame of reference is lost.) Using this method, the entry age normal cost for
the current civil service retirement system is 36.52 percent for 1982, (accord-
ing to the most recent estimate of the Office of Personnel Management.)
This is the generally accepted costing method and is the one most used for
valuing a pension system's long run cost. It is used by the Office of Personnel
Management and other private and public evaluators of the CSRS. These long run
costs are important because employees working today are earning rights for pen-
sion benefits that will become payable in future years.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Another important cost concept that.is also used in CSRS analysis is the
budget cost. In doing budget analysis, projections of costs are made for no
more than 6 years into the future, a long time for most programs, but a very
short time for pension programs, whose actuaries do estimates of costs 75 and
more years into the future. These short run budget figures are also important,
and it is well to keep.in mind that for any paticular year they represent
rights earned in previous years. While Congress has the legal authority to de-
value these earned rights, based on past experience it is unlikely to do so. 16/
The practices seem to have been that changes in pension programs should be
made only as they affect benefit rights earned in the future, whether the em-
ployment be in the public or private sector.
A final distinction is often made in talking about pension costs. Actu-
aries present data in the form of cost to the employer as opposed to the overall
cost of the benefits rights earned. Thus, when the costs of the CSRS are dis-
played, the amount of the employee contributions is often subtracted to yield
what is called the cost to the employer. For example, the cost the Government
as employer of the CSRS is 35 percent minus 7 percent (employee contributions),
leaving a figure of 28 percent.
The amount of employee contributions to-require in a new pension system
would be controversial, although the recent legislation (P.L. 98-168) may have
helped to settle that issue. The significant point to make is that the amount
of employee contributions has nothing to do with the amount of benefit rights
that accrue from a defined benefit pension. This is true not just for civil
service retirement but for social security and defined benefit private pension
16/ Financing Work-Related Entitlements. Legal Rights of Individual Under
the Social Security Program and Federal Retirement Programs as contrasted with
Private Sector Counterparts, by Kathleen S. Swendiman. p. 503.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? . ?
systems as well. In fact, most private pensions require no employee contribu-
tions at all, but the benefit rights earned by employment are just as real.
Furthermore, the military retirement system, similar to the CSRS in that the
Federal Government is the employer in both cases, requires no employee contribu-
tions. This does not depreciate the value of the benefits earned in any way.
2. Financing and Trust Funds. A clear distinction should be drawn
between the cost of a program and its financing. When the discussion turns from
cost as the present value of benefits earned in the future (difficult concepts
by themselves)-to discussions of ways of financing this cost, additional diffi-
cult terminology and concepts come into play.
One reason is that pension funds are most often financed through trust funds.
(The military retirement system, which previously had no trust fund, will be set-
ting one up as a result of recent legislation.) There are two things to keep in
mind about the CSRS trust fund: (1) It does not ensure the availability of funds
to pay benefits at a particular level; and (2) benefits have to be paid from re-
sources available in a particular year. Except for the possibility of affecting
the national savings rate by its trust fund actions, which the civil service re-
tirement trust fund does not do, the presence or absence of a trust fund for a
Federal Government pension, has no real effect on the Government's ability or
commitment to make benefit payments. In social security, as well as Federal pen-
sions, commitments ultimately are backed up by the taxing authority of the United
States Government, not assets in a trust fund.
Deciding on policy goals and designing a pension system to meet them may
be impeded by public confusion and misunderstanding about the way we now pay for
civil service retirement and other Federal pension and social insurance programs.
There is a general, but mistaken, perception that the economic and budgetary ef-
fects of these Government trust funds are analogous to trust funds used to finance
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
S ?
pensions in the private sector. For private pension plans, accrued rights to
benefits earned from years of work and earnings are legally enforceable rights,
and the Employee Retirement Income Security Act (ERISA) has established a variety
of requirements governing the accumulation and investment of reserves. Eventu-
ally, private plans should have sufficient funds to be set aside so that at the
time of retirement the fund will be large enough to pay the benefits accrued;
that is, they are to be "fully funded." Thus, benefits will not have to be paid
out of current operating revenues.
Although similar in appearance, Federal Government trust funds do not and
cannot in themselves ensure availability of funds to pay future entitlement
benefits. Benefit payments for Federal pensions and social insurance programs
have to be paid for in much the same way as other Federal expenditures. In
this context, the payroll and other earmarked taxes are viewed as an overall
source of revenue rather than a source of revenue for a particular purpose.
Trust funds, for those programs that have them, function largely as a mechanism
to ensure sufficient funds for making benefit payments on time without the need
for annual appropriations.
While this analysis emphasizes the budgetary effects of financing Federal
pensions and social insurance through trust funds, it is recognized that this
emphasis ignores some features of and distinctions among trust funds that often
weigh importantly in congressional policy concerning program financing. For
example, a trust fund shortfall puts pressure on Congress to cut benefits. On
the other hand, substantial social security trust fund balances in the past have
made it easier for Congress to raise benefits, indeed, have exerted pressure (or
the temptation) to raise benefits.
The social security trust fund, in contrast to the civil service retire-
ment fund, receives almost all revenues from special earmarked payroll taxes
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
whose only purpose is to pay social security benefits. The special earmarking
of these trust fund revenues may strengthen the political commitment to main-
tain payments as long as trust fund balances are positive.
In addition, while the present funding basis for civil service retirement
may not fully reflect its budgetary effects, it may be desirable psychologically
and, to a considerable extent, actuarially. It makes the public and the covered
employees more aware of the long run costs of the system than if the Federal
Government merely paid, each year, the excess of the outgo over the employee
contributions.
All Federal trust fund assets consist of the Federal Government's own se-
curities which are, in effect, IOUs the Government has written to itself. Under
current law, benefits can be paid only if there are trust fund balance. However,
regardless of the size of trust funds, benefit payments can be made only from
funds available in a specific year--whether they come from taxes, other revenues,
or money borrowed from the public. This is what is meant by a pay-as-you-go
financing system: Benefits are paid by the current generation of workers and
taxpayers to the current generation of retirees. Measured by economic and
budgetary effects, Federal pension and social insurance programs, whether they
have trust funds or not, are de facto pay-as-you-go systems.
The total size of the obligations for benefit payments does not depend on
whether or not program benefits are paid through a-trust fund. It depends in-
stead on the eligibility and benefit rules set up in the program's authorizing
legislation. For instance, as noted above, the entitlement to military retire-
ment benefits is no more or less binding than the entitlement to civil service
retirement benefits, although the military retirement system has no trust fund
at all while the civil service retirement system has a relatively large and
growing trust fund. The requirement to pay beneficiaries their legally specified
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? S.
amount at the legally specified time does not depend on the existence or non-
existence of a trust fund. Further, the entitlement to benefits in both programs
is established by legislation and can be changed for both programs by Congress.
In the civil service retirement and military retirement programs, the Fed-
eral Government is not only the payer of benefits, but the employer as well.
Trust funds become even less significant when used in such programs. Military
retirement will have no trust fund at all until .1985 and payments will simply be
made each year from general fund appropriations. This procedure did not prevent
the calculation. of total benefit obligation--nearly one-half trillion dollars.
If the military retirement system ceased enrolling new members today, payment
of accrued entitlement. benefits would have to be made, subject to subsequent
changes in law, until the last beneficiary was removed from the rolls.
The civil service retirement system does have a trust fund, but calculation
of the total amount of the Government's accrued obligation for civil service re-
tirement is similar to that of military retirement. However, its financial ar-
rangements are substantially different. Only part of the total obligation is
"unfunded" in an actuarial sense, with the remainder being funded (on the books
of the Federal Government) by assets held by the civil service retirement trust
fund and by future earmarked employer and employee contributions. As long as
the trust fund assets are invested in the Government's own securities, adding
to these assets does not constitute budget outlays, and conversely does not
prefund retirement benefits.
Any trust fund size or buildup has no effect by itself on Federal outlays
or deficits now or in the future. These are determined by the benefit to which
workers and retirees become entitled--the same as in the military retirement sys-
tem. Measures taken by the Government as employer to prefund result in IOUs
the Government writes to itself. The obligation to make payments follows from
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? 0
the provisions of the benefit formula, not from trust fund balances. The civil
service retirement and the military retirement program could be fully funded on
the books of the Government without changing the size of Government outlays or
the Federal deficit in any way.
The civil service retirement system financing procedures have confused
many. A commonly held view is that the pensions should be fully paid for by em-
ployees and employers. However, as long as the Government as employer invests
trust fund assets in its own securities, the idea of prefunding has little eco-
nomic or budgetary meaning. Furthermore, in a pay-as-you-go system, increasing
the amount of employee contributions would simply result in requiring Federal
workers to pay more of the cost of benefits to retired employees rather than hav-..
ing these benefits paid by taxpayers in general.
3. The Fund and the Unfunded Liability. If the important analytical
concept is the total value of the civil service benefit obligation accrued, and
if this obligation could be prefunded only with Government IOUs to itself, why
are some critics concerned about the "unfunded liability"? It follows from the
previous analysis that whether the liability is funded or unfunded makes no real
difference when the fund is built from sources that are internal to the.Govern-
ment. A brief explanation of the term, "unfunded liability" will be followed
by a discussion of how the current CSRS financing (through its trust fund) works.
It has been asserted that CSRS's unfunded liability represents.a menace to
the entire economic well-being of the nation. Unfunded liability is defined as
the difference between the projected future assets of the fund and its projected
liabilities. It is argued that this unfunded liability in CSRS represents a
growing, unmanageable, and unconscionable threat to future taxpayers.
The term, "unfunded liability" sounds ominous, and in the private sector
high unfunded liabilities would constitute.a significant threat to beneficiaries
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? 0
as well as plan sponsors. However, the term refers to a complex actuarial con-
cept that in this system is used strictly to regulate the flow of budget authority
into the system's disbursing mechanism. Actually, a far more important concept
is system "liabilities" both funded and unfunded, the measurement of which is
rather sensitive to the assumptions and methods used to establish it. This pa-
per maintains that it will be more productive to concentrate on the effect upon
the total liability arising. from the various operations of the CSRS and compare
these effects to the various alternative operations that could take their place.
The civil service retirement system trust fund has grown over the years.
Sources of revenues to this trust fund are mainly internal to the Government,
as table 4 shows. During 1983 through 1992, total income to the trust fund is
projected at $498 billion. Of this, 13 percent is from employee contributions,
28 percent from interest on trust fund assets, and 59 percent from other Govern-
ment transfers. Only the 13 percent from employee contributions comes from out-
side the Government.
. Changes made to CSRS financing in 1969 enabled the civil service retirement
trust fund to grow rapidly during the 1970s, when trust funds financed by pay-
roll taxes were having difficulty. In 1969, the Congress considered the issue
of the unfunded liability of the civil service retirement system. Congress de-
cided to make paper transfers (issue special notes or IOUs to itself) to the
civil service retirement trust fund in sufficient amount to keep the fund grow-
ing. These transfers were for interest that would have been earned if the un-
funded liability had been funded, amortization of certain unfunded liabilities,
agency contributions, and interest on transferred assets. Interest on the un-
funded liability and interest on assets have been growing, and along with them
the trust fund. Congress also built up the trust fund by amortizing additions
to the unfunded liabilities caused by new or liberalized benefits (except
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
. 0
cost-of-living adjustments), extension of coverage to previously uncovered em-
ployees, and increases in salaries on which benefits are computed. Employee
and agency contributions were increased from 6.5 to 7 percent.
TABLE 4. INCOME TO THE CIVIL SERVICE RETIREMENT FUND
(In billions of dollars)
30-year Inter-
Em- Agen- amorti- est on Mili- Invent- Total
ployee. cy zation un- tary ment in-
Year contri- contri- pay- funded pay- income come
butions butions ments liabili- went
ty
1983 ....
4.7
4.7
5.1
9.8
1.5
9.7
35.5
1984 ....
5.0
5.0
6.0
10.5
1.7
10.8
39.0
1985 ....
5.4
.5.4
6.9
10.9
1.8
11.8
42.2
1986 ....
5.8
5.8
7.7
11.2
1.9
12.7
45.1
1987 ....
6.2
6.2
8.6
11.6
2.0
13.5
48.1
1988 ....
6.6
6.6
9.5
12.0
2.1
14.3
51.1
1989 ....
7.1
7.1
10.5
12.4
2.1
15.1
54.3
1990 ....
7.5
7.5
11.5
12.8
2.2
16.0
57.5
1991 ....
8.0
8.0
12.5
13.2
2.3
16.8
60.8
1992 ....
8.4
8.4
13.5
3.6
2.4
17.7
64.0
Total ... 64.7' 64.7
91.8
118.0
20.0
138.4
497.6
Source: See Table 1, p. 33.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 0
Table 5 shows the operations of the civil service retirement program from
two perspectives: effect on the unified budget and effect on the trust fund.
TABLE 5. RECEIPTS AND EXPENDITURES FOR CIVIL SERVICE RETIREMENT--1983-92:
BUDGET PERSPECTIVE AND TRUST FUND PERSPECTIVE
(In billions of dollars)
Assumptions
Receipts and expenditures:
Budget expenditures .................................... 325.6
Receipts .............................................. 64.7
Excess of expenditures over receipts a/............ 260.9
Trust fund perspective:
Income to the trust fund .............................. 497.4
Budget expenditures ................................... 325.6
Net increase in fund ....................... 171.8
a/ $30 billion of this will come from Postal Service contributions.
Source: See Table 1, p. 33.
Of the total benefit payments of $325.6 billion during 1983-92, one-fifth
or $64.7 billion will come from employee contributions. The remaining $260.9
billion will come from tax revenues or borrowing. During the same period, the
trust fund.financing mechanism will result in a net addition to the trust fund
of $172 billion.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
In effect, the trust fund is owed by the Federal Government to itself.
The civil service retirement program affects the Federal budget deficit and the
economy by the extent to which its outlays for benefit payments each year differ
from employee contributions that year.
The size of the trust fund and the amount of the unfunded liability do not
affect outlays or the budget deficit now or in the future if trust fund assets
are invested in the Government's own securities. Eligibility requirements and
benefit formulas, both established by statute, determine the extent of the
Government 's obligation to make payments.
Before ending this discussion of pension obligations and their funding, it
should be noted that it is possible to think of other Government programs as
establishing obligations for the future also. OPM press releases compare the
unfunded liabilities of the civil service and military retirement to the nation-
al debt. One could just as easily compare the discounted present value of the
food stamp program, or of Federally funded agricultural or medical research, or
of military expenditures to the national debt or to anything else. Without
putting too fine a point on it, pension and social insurance programs are not
the only ones which imply future obligations. Even those who say the Federal
Government is too. large at present are not advocating its elimination, only its
reduction. For example, the current administration has advocated reductions'in
overall Federal expenditures not to zero, but until they reach 19 percent of
GNP. 17/
Nor can it be argued that the obligation for pensions and social insurance
differs in kind from the obligations for other Government expenditures. There
is a difference in degree, certainly, but as noted above, the Congress does have
17/ U.S. Office of Management and Budget. A Program for Economic Recov-
ery. Feb. 18, 1981. p. 11.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 0
the authority, if it chooses, to reduce social insurance and pension entitle-
ments as well as other expenditures.
4. How to Settle on the Cost of a New System. The cost of the cur-
rent civil service retirement system is, by conventional measures, around 35
percent of payroll, using the entry age normal cost system and dynamic assump-
tions. Based on this cost figure, the CSRS has been criticized as too expensive,
as substantially more expensive than prevailing-pension plans in the private sec-
tor. Against these assessments are rebutting arguments made by groups represent-
ing Federal employees during last year's debate on the Social Security Amendments
of.1983 that the CSRS really only costs about 14 percent of payroll, that the
Federal employers and employees each pay 7 percent of this, and that the system
started having problems because the Federal Government did not pay its full share
in the early days of the program. 18/
The President's budget for 1984 included proposals that would have reduced
the cost of the CSRS from its current 35 percent down to 22 percent, and would
have increased employee contributions until they were 11 percent (from the cur-
rent 7 percent) which when combined with a matching 11 percent from employers
would exactly fund the new reduced system.
Without getting too much into the technical details another word needs to
be added about the different kinds of actuarial costs. Actuaries talk about
"dynamic" and "static" costs, in making their estimates. There is a big differ-
ence. The 35 percent entry-age normal cost figure is arrived at using "dynamic"
18/ Moe Biller. President of the American Postal Workers Union. Testi-
mony before the House Subcommittee on Social Security of the Ways and Means
Committee. Feb. 8, 1983.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
S
assumptions, while the 14 percent entry-age normal cost relies on "static" assump-
tions. The big difference is that dynamic assumptions assume continued inflation
and real wage growth while static assumptions do not. Intuitively, one would say
that, of course, if inflation and wage growth add to pension costs they ought to
be included. The bulk of actuarial opinion appears to favor this view. In
fact, all of the large Federal systems, including social security, are valued
using dynamic assumptions.
However, the issue is not so clear cut as it might seem to be. Until re-
cently, the Federal system, contrasted with private and with State and local
government pensions, have been fully and automatically protected against price
inflation-by having cost of living adjustments (COLA) tied to increases in the
consumer price index (CPI). Private and State and local government pension sys-
tems do not have full and automatic COLAs even though most of them do make ad
hoc adjustments to take account of inflation. The point is that the Federal
systems are valued based on economic assumptions that include the effects of
postretirement inflation, while others are not. And this is a large difference.
These points are technical, but bear importantly on the issues, because
cutting back on the COLAs seems to be the way the Congress has seen to be most
appropriate when reductions need to be made in Federal pension and social in-
surance programs. Technically, the Congress has not devalued already earned
rights, because full and automatic adjustments to what are called post entitle-
ment benefits are not earned until the benefits are increased.
Given all these difficulties in measurement, the Congress is still left
with the question of deciding how much the new pension system should cost (and
whether to deal implicitly or explicitly with whether the cost of the current
system is excessive). The parameters would seem to be as follows:
1. To target the cost of the new pension system at about the same as
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
the current system--or 35 percent of payroll (dynamic assumptions). While it
would not be exact, it becomes a fairly simple matter to devise the supplemental
pension which, when combined with social security, would give about this cost.
Many difficult questions of program design would still remain, depending on which
of the system objectives goals were thought to be paramount.
2. To pattern the cost of the new CSR pension, when combined with so-
cial security, at approximately the cost of pensions in the private sector.
This generally implies cutting back on Federal pension costs. This approach has
a great deal of conceptual appeal but has formidable problems associated with
measuring the cost of the "typical" private pension. In the first place,
there is no such thing as typical private practice. There are over 500,000
pension plans in the United States and almost as many variations in them.
Even if it were possible to isolate and agree on a.set of typical provisions
and value. them, it is not clear that the values would be comparable unless
they were valued by using a common set of economic assumptions and a com-
mon demographic data base. The recent proliferation of company sponsored capi-
tal acumulation plans (such as thrift plans, 401(k) plans, etc.) would make it
even more difficult to ensure that all these costs are captured in the compari-
son. (They are part of the retirement income package offered by many large
corporations and, though not offered to Federal employees, should be considered
in the value comparison.) The data base itself raises another problem. What
set of private sector employers would it be appropriate to compare Federal
employees to for purposes of this comparison? Some argue that it should be
an average of all employers, large and small. Others argue that it should only
be the larger employers because these are the ones the Federal government com-
petes with for many types of employees and whose work environment is more near-
ly similar. It is assumed, although even this question has not been analyzed
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
conclusively, that pension benefits of some small employers are smaller than
those of large private employers. If there is such a difference, difficult
judgments will be required to settle the question.
If Congress were to peg the cost of the new system on prevailing pension
practice in the private sector, the cost of the new system,"when combined with
social security would be substantially less than the current CSRS. (There would
be many ways to reduce cost's. Since the single most expensive feature of the
current system is its full and automatic COLAs, the new system would be less ex-
pensive if it promised less than full and automatic COLAs.) Several State gov-
ernments, facing funding problems in their pensions, have done this very thing.
They have begun new pensions for future workers with lower costs than the exist-
ing system.
A serious concern, if this were to be done, is that, while the current CSRS
is more expensive than private practice, Federal employees still have less "to-
tal compensation" than private workers, according to most studies. (OPM disagrees.)
In his Brookings study, Robert Hartman argues that the pension sytem should be cut
back to be more nearly comparable to private practice; that Federal salaries, par-
ticularly at the top levels be substantially increased to be more competitive. Mr.
Hartman does not say what he would propose to do with the pension system in the ab-
sence of corresponding increases in salaries. The studies indicate that not only
have Federal employees fallen behind in total compensation, they are also falling
behind in the nonretirement aspect of total noncash compensation. Recent studies
by the Congressional budget Office and Hay Associates have concluded that although
Federal pensions are more costly than their private sector counterparts, recent cut-
backs in health insurance benefits and other noncash benefits accompanied by a low-
er relative salary base, which directly affects most benefits, have pulled Federal
workers about even with private sector employees in total noncash benefits even
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
I
though the Federal pensions, as a part of this-total, are clearly more valuable.
Thus, before settling on a new pension plan for Federal workers covered by
social security, its effect on total compensation should be examined, considering
the effect that a cut in total compensation would have on the type and quality
of employees.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
is 0
EBRI
SOURCES OF ECONOMIC SECURITY
-.--TABLES---
A Presentation for the United
States Senate Committee on
Governmental Affairs'Subcamittee
on Civil Service, Post Office and
General Services
POLICY FORUM
ECONCNIIC SECURITY PROGRAMS
December 13, 1983
Prepared by: Fzgloyee Benefit Research Institute
Presented by: Dallas L. Salisbury,. President
EMPLOYEE BENEFIT RESEARCH INSTITUTE
1920 N Street, NW/Suite 520/Wuhington, DC 20036/Telephone [phone]
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
The Six Pillars of Economic Security
Economic Security
Personal Savings
Needs-
in
om
Social
and
Related
c
e
Fr
Family
Retirement
Security
om
Transfers
Plans
Investments Public Plans Employment
I 1 Savings
Medicaid
r I Full time
I I Family
! Home equity
SSI
! Part-time
support
Profit-
i Business
In-kind
Seasonal
Bequests
sharing
equity
benefits
Gifts
Thrift
IRAs
savings
ILife
Stock
insurance
ownership
i : Stocks.
bonds
I 'l Annuities
I I (Health
insurance)
(Life
insurance)
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
is
Percent
Receiving
Total Income.
from Each Source=
(billions)
Average
Income3
Social Security
92%
S 66
S4,300
Employer Pensions
34
23
4,100
Savings/Wealth
71
36
3,000
Employment Earnings
31
49
9,500
Public Assistance/In-Kind
Programs
95
43
2,200
Total
N/A
5217
N/A
Source_ ICF analysis for EBRI of the March 1980 Current Population Survey data.
'Elderly households consist of married couples with a bead age 65 and over and single
persons age 65 and over.
=Total income provided to all elderly from each source listed.
3Average amount from each source provided to households, which receive income from the
particular sources noted.
'ICF estimates derived from data presented in U.S. Social Security Administration. Demo
graphic and Economic Characteristics of the Aged (1975).
TABLE 1-2
OASI Beneficiaries and Benefit Payments-1980
Beneficiaries
(millions)
Benefit Payments
(billions)
Average
Benefit
Retired Workers
19.1
S 74.3
54,100
Wives and Husbands
3.0
5.8
2,100
Children
3.1
8.2
2,600
Widowed Mothers
0.6
1.6
2,900
Widows and Widowers
4.4
15.2
3,700
Other
0.1
0.2
1.500
Total
30.3
5105.3
53.700
Sour= U.S. Social Security Administration. Social Security Bulletin. vol. 44, no. 4
(198 1). pp. 48-49.
TABLE 1-3
Age 65 Social Security
Pretax Replacement Rates-1979
Single
Worker
Worker with
Dependent Spouse
Maximum Covered Wage Earner
25%
38%
Average Covered Wage Earner
44
65
Full-Time Minimum Wage Earner
57
85
Source: U.S. Department of Health and Human Services, Advisory Council on Social
Security, Social Security Financing and Benefits (1979), p. 59.
'The Social Security minimum benefit has been elitttitwted for all persons retiring after
l:utuary I. 1982.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
is
i
Relation Between Preretirement Earnings and Postretirement Social Security Benefits
65-Year Old Workers at Different Earnings Levels-1982-
FuU-time Minimum
Neairnum FAmer,~-,_n~e F~rner' R'age'
Half-lane Minimum
Wae1
With
Depiu dent
Spusi
Sin
l
With
Dependant
With
Dependent
With
Dependent
g
e
Spnuw
tingle Spwne
Single
Spurn,e
Sintle
Prerettrement 1981
Gross monthly earnings
Less taxes and ex
enses
$2.475
52,475
$1.129
51,129 5581
5581
$290
$290
p
859
1.048
276
0
Disposable earned income
616
1
427
1
%
133
37
39
Postretirtment 1982
.
.
853
0
79
44
$
3 53
25
251
AIME
January PIA
51.258
51.258
S 847
S 847 $459
5459
5229
Mn
Average monthly benefit
er
ear
606
606
476
476 319
319
226
226
p
y
Plus: SSl
940
627
739
493 495
330
351
234
Replacement Rates
0
0
0
0 0
0
41
34
OASI benefit/gross earnings
OASI and SS1 benefits/net earnin
38?.
25::.
65':.
447C 85%
M.
121.'
91%
gs
58'c
44%
87%
621% 102
74%
155%
107'3f
Source: Social Security Financing and Benefits Report of the 1979 Advisory Council on Social Security (Washington. D.C.. 1980). p. 49.
'Worker has always earned maximum level taxable under Social Security.
3Worker has always earned average earnings in employment covered by Social Security.
Worker has always worked full-time at the Federal minimum wage.
*Worker has worked 50 pcn:cni of the time at the FL-dead minimum wage.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
?
L) I1e11Ls-1YZU 10 1y1y'
?
Year
Beneficiaries
(millions)
Total Benefits
(millions)
Average
Benefit
Real
Average Bcncfit
(1979 dollars)
1950
3.5
$ 961
$ 276
S 832
1955
8.0
4,968
624
1,691
1960
14.2
10,677
754
1,848
1965
19.1
16,737
875
2,013
1970
23.6
28,796
1,222
2,284
1975
27.7
58,509
2,110
2,846
1979
30.3
90,556
2,984
2,984
Source: U.S. Social Security Administration, Social Security Bulletin Annual Statistical
Supplement, 1977-79, pp. 66-67.
'These estimates are different from those in Table 1.1 because they include beneficiaries of
all ages.
TABLE I.5
Benefits and Beneficiaries in
Employer Pension Programs-19791
Employer Plans
Beneficiaries
(millions)
Percentage
of Elderly
Households
Receiving
Benefit
Payments
(billions)
Average
Benefits
Private
8.7
23%
$23.6
52,700
State/Local
2.3
8
10.8
4,700
Civil Service
1.6
4
12.5
7,700
Military
1.3
1
10.3
8.000
Subtotal
13.9
34%
557.2
54,100
OAS1
30.3
92%
S90.6
53,000
Source: See tables in Appendix A.
'L'iiintates of beneficiaries, benefit payments and average benefits presented in this table
include benetciaries of all ages. They. therefore, differ from those in Table 1.1.
TABLE 1-6
Private Sector Employer Pension
Programs-1950 to 1979
Year
Beneficiaries
(thousands)
Total Benefits
(millions)
Average
Benefit
Real
Average Benefit
(1979 dollars)
1950
450
S 370
S 822
52,479
1955
980
850
867
2,350
1960
1,780
1,720
966
2,368
1965
2,750
3,520
1,280
2,945
1970
4,740
7,360
1,553
2,899
1975
7,115
14,850
2,087
2,815
1979
8,700
23,600
2,713
2,700
Sources: Alfred M. Skolnik, "Private Pension Plans. 1950-1974," Social Security Bullr-
tin. vol. 39. no. 6. June 1976, pp. 3-17. American Council of Life Insurance.
Pension Facts (Washington, D.C.. 1977). pp. 30-31. 36. Private pension values
for 1979 were derived from ICF Incorporated. A Private Pension Forecasting
Sfinlrl, 1979.
?1'ritfit Sharing Research Foundation. "Cumulative Growth in Number of Qualified De-
Icrrcd Profit Sharing Plans and Pensions in the U.S. 1939 Through 19811 if:. ........ It.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 State and f.nrttl rnva.?n..,.,..?
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Year
??
Beneficiaries
(thousands)
- ?.. ??.a-ai//
Total Benefits
(millions)
IV 17/7
Average
Benefit
?
Real
Average Benefit
(1979 dollars)
1955
427
$ 595
51,393
$4,200
1960
660
1,078
1,633
4,002
1965
886
1,775
2,003
4,608
1970
1,291
3,280
2,541
4,750
1975
1,730
7,025
4,061
5,477
1979
2,300
10,770
4,683
4,683
Sources: U.S. Social Security Administration, Social Security Bulletin Annual Statistical
Supplement. /977-79, pp. 66-67; American Council of Life Insurance, Pension
Facts (Washington, D.C., 1980), table 15.
TABLE I-8
Benefits under the Federal Civil Service
Retirement System-1955 to 1979
Year
Beneficiaries
(thousands)
Total Benefits
(millions)
Average
Benefit
Real
Average Benefit
(1979 dollars)
1955
297
$ 380
$1,279
S3,857
1960
515
814
1,581
3,875
1965
729
1,385
1,900
4,371
1970
959
2,838
2,959
5,531
1975
1.372
7,056
5,143
6,936
1979
1,617
12,380
7,656
7,656
Swuces: U.S. Social Security Administration, Social Securirv Bulletin Annual Statistical
Supplement, 1977-79, table 16. pp. 66-67; American Council of Life Insurance,
Pension Facts (Washington, D.C., 1980), table 13, pp. 26=27-
TABLE 1.9
Benefits under the Military
Retirement System-1955 to 1979
Year
Beneficiaries
(thousands)
Total Benefits
(millions).
Average
Benefit
Real
Average Benefit
(1979 dollars)
1955
179
$ 419
52,341
57,059
1960
256
694
2,711
6,645
1965
4&4
1,384
2,860
6,580
1970
773
2,849
3,686
6,890
1975
1,073
6,242
5,811
7,845
1979
1,286
10,279
7,993
7,993
Source: U.S. Department of Defense, Defense Manpower Data Center, Office of Actuary.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
rta:ntLIgC or ramifies wttn family d
Wage 64 to 69 with Assets-1975
Percentage
[folding
Median Value of
This Asset for
Those with Asset
(1979 dollars)
Liquid Assets
81%
S 7,300
Life Insurance, Annuities
75
5.100'
Home Equity
69
27,000
Illiquid Assets
24
13,500
Any of Above Assets
89
33,100
Sources: Joseph Friedman and Jane Sjogren. "The Assets of the Elderly As They Retire"
(Cambridge, Mass.: Abt Associates. Inc., I980). pp. 15. 36, 46, 49 and 66.
Median value of assets in 1979 dollars estimated by ICF Incorporated.
'This estimate reflects the cash value rather than the face value of these policies.
TABLE 1.11
Labor Force Participation Rates
for Men-1950 to 1980
1960
- 1970
19[30
55-59
87%
88%
90%
83%
60-64
79
78
75
61
65 and over
39
29
27
20
Source: U.S. Department of Labor, Bureau of Labor Statistics.
"EBRI, Coverage and Benefit Entitlement, p. 57.
teICF analysis of the May 1979 Current Population Survey data.
TABLE 1.12
Percentage of Workers Who Work Part-Time-1977
25_S4
SS-59
60-64
65 and Over
Men
4%
4%
10%
48%
Women
28
27
35
62
Total
14%
14%
20%
54%
Source: U.S. Department of Labor. Bureau of Labor Statistics, "Work Experience of the
Population in 1977." Special Labor Force Report no. 2r4, p. A-7.
TABLE I.13
Public Assistance and In-Kind
Benefit Program for the Elderly- 1980
Programs
Elderly
Individuals
Participating
(millions)
Level of
Benefits
(millions)
Average
Benefits per
Participant
Medicare
24.5
528,300
S1,200
Medicaid
5.1
4,300
800
SSI
1.9
2,400
1,300
Subsidized
Housing
1.3
1,600
1.200
Energy
Assistance
l: r,Aff Crft 1 ft.
1.9
i A
700
fnn
400
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 TA*.,LF 1.1.5
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Married Couples
Single
Persons
All Eldt
Households
Less than 52,500
2.0%
( 2.0)
9.1%( 9.1)
5.8%( 5.8)
52,500-4,999
10.7
( 12.7)
39.4 ( 48.5)
26.1 ( 31.9)
S5,000-9,999
31.9
( 44.6)
29.8 ( 78.3)
30.8 ( 62.7)
510,000-14,999
23.5
( 68.1)
10.2 ( 88.5)
16.4 ( 79.1)
S 15,000-24,999
18.8
( 86.9)
7.4 ( 95.9)
12.6 ( 91.7)
525,000 and over
13.2
(100.0)
4.0 (100.0)
8.2 (100.0)
100.0%
100.0%
Source: ICF analysis for EBRI of the March 1980 Current Population Survey data.
'Cumulative distribution in parentheses.
TABLE I-1S
Cumulative Cash Income Distribution of the Elderly,
By Marital Status and Age-1979
Married Couples Age
Single
Persons Age
65-72
Over 72 ,
65-72
Over 72
Less than $2,500
1.9%
2.1%
9.1%
9.1%
Less than S5,000
10.8,
15.5
43.8
52.1
Less than S 10,000
37.4
55.6
74.9
81.0
Less than S15,000
61.9
77.4
86.7
90.0
Less than 525,000
84.1
90.9
95.5
96.3
Total
100.0%
100.0%
100.0%
100.0%
Source: lCF analysis for EBRI of the March, 1980 Current Population Survey data.
age 72. Employment income earned by those age 65-72 explains most of
this difference.
Much of the income variation is explained by income sources. High-
income elderly families generally receive income from different sources,
and from more sources than.low-income families. Social Security income
was received by over 90 percent of all households. As shown in. Table 1-16,
except for those with the lowest and highest incomes, more than 90 percent
TABLE I-16
Percentage of the Elderly Receiving Cash Income, By Source-1979'
Employment
Earnings
Social
Security
Employer
Pensions
income from
Assets .
Government
Assistance
Other
Government
Programs
Other
Less than S2.500
6%-
69%
4%
32%
23%
3%
2%
52.500-4.999
9
94
10
49
24 .
8
2
55.000-9.999
24
96
37
75
6
10
3
510,000-14,999
42
95
55
87
4
10
4
515,000-24,999
60
91
52
88
4
10
5
55,000 and over
75
80
52
94
2
10
5
All Elderly
31%
92%
34%
71%
11%
9%
3%
Source: ICF analysis for EBRI of the March 1980 Current Population Survey data.
'Percentages are based on the number of married couples with a head-of.household over age 65 and single persons over age 65.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 TABLE I-17
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Elderly Htl
holds, By Earnings Status-197
With
Employment
Earnings
Without
Employment
Earnings
Total
Less than 52,500
1.2%
7.8%
5.8%
Less than $5,000
8.8
42.1
31.9
Less than 510,000
32.9
75.9
62.7
Less than S 15,000
55.0
89.7
79.1
Less than S25,000
79.8
96.9
91.7
Total
100.0%
100.0%
100.0%
Source: !CF analysis for EBRI of the March 1980 Current Population Survey data.
TABLE I.18
Working Status of the Elderly,
By Age and Marital Status-1979
Full Time
14%
6%
3%
7%
4%
1%
Part-Time
26
23
12
21
13
6
No Work -
60
71
85
72
83
93
TABLE I-19
Cumulative Distribution of Cash Income, By
Marital Status, Work Status and Age-1979
Not Working
Part.Time Work
Full-Time Work
65-69
70-72
Over 72
65-69
70-72
Over 72
65-69
Over 69
Married Couples
Less than S2.500
2.7%.
1.8%
2.0%
0.8%
0.5%
3.1%
1.9%
0.5%
Less than 55.000
14.0
14.9
17.0
5.5
4.1
9.0
3.8
1.1
Less than S10.000
43.8
50.8
59.8
25.9
27.2
36.6
11.5
14.0
Less than S15.000
71.3
74.5
81.7
50.7
58.1
61.1
21.8
29.1
Less than 525,000
90.1
93.1
93.6
78.5
77.9
84.1
55.1
55.5
Total
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Single Persons
Less than S2.500
12.0%
11.2%
9.5%
4.1%
3.6%
4.1%
3.5%
N/A
Less than 55,000
50.3
54.4
54.1
26.7
22.1
27.8
8.4
N/A
Less than $10,000
79.1
82.9
82.2
67.1
71.9
67.8
31.7
N/A
Less than $15,000
88.4
92.9
90.9
83.8
87.9
80.1
59.7
N/A
Less than $25,000
96.2
97.0
96.9
94.2
95.8
91.1
87.0
N/A
Total
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
N/A
Source !CF analysis for EBR! of the March 1980 Current Population Survey data.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Relative Contribution Setirement Income Sources for thOderly-1979'
Cash Income
in 1979
Employment
Earnings
Social
Security
Employer
Pensions
Income from
Assets
Government
Assistance
and Other
Total
Less than 55.000
3%
76%
3%
6%
12%
100%
55.000- 9,999
10
59
12
15
4
100
510,000-14,999
18
43
17
19
3
100
S 1 5 ,000- 24,999
34
26
16
22
2
100
525.000 and over
48
11
13
27
1
100
All Elderly
27%
37%
13%
20%
3%
100%
Source: ICF analysis for EBRI of the March 1980 Current Population Survev data.
'Percentages are based on the number of married couples headed by an individual age 65 and over and single persons age 65 and over.
TABLE 1-21
Percentage-of Elderly Households Receiving
Certain In-Kind Benefits-1979
Medicare
Medicaid
Food
Stamps
Housing
Assistance
Less than $2,500
88%
30%
16%
12%
S2,500-4,999
95
29
16
13
55,000-7,499
95
14
4
5
57.500-9,999
95
12
2
3
S10,000-14,999
94
10
2
2
S 15,000 and over
89
9
1
-
All Elderly
93%
16%
6%
5%
Source: U.S. Department of Commerce, Bureau of the Census. Characteristics of House,-
holds and Persons Receiving Nun-Cash Benefits: 1979, Current Population Re.
ports. series P-23, no. 110, pp. 12. 15. 16 and 18.
TABLE 1.22
Cumulative Income Distribution of Elderly Households,
Including Cash Income and In-Kind Benefits-1979
Cash
Cash Income and In-Kind Benefits
Income
Only
Excluding
Medical Benefits
Including
Medical Benefits
Less than S2.500
5.8%
4.9%
3.6%
Less than 55,000
31.9
29.7
25.9
Less than S10,000
62.7
62.5
59.4
Less than S 15,000
79.1
79.0
77.0
Less than 525,000
91.7
91.8
91.1
Total
100.0%
100.0%
100.
Source: ICF analysis for EBRI of the March 198U Current Population Survey data. See
Appendix B (cc details.
"if these in-kind benefits were valued at cost, the cumulative percentage of families with
incomes:
less than 52.500 would total 1.4 percent;
less than 55,000 would total 15.6 percent;
less than S 10.000 would total 54.7 percent:
less than S 15.000 would total 75.3 percent;
less than 525,000 would total 90.8 percent.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 ?
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
to 0
TABLE 1-23
Cumulative Income Distribution of Elderly
Households, Including Cash Income,
In-Kind Benefits and Annuitized Wealth Income-1979
Cash
Cash Income, In-Kind Benefits and
Annuitlzed Wealth
Income
Only
Excluding
Medical Renef is
Including
Medical Benefits
Less than S2,500
5.8%
2.9%
2.3%
Less than S5,000
31.9
19.9
17.2
Less than S 10,000
62.7
50.8
47.8
Lets than $15,000
79.1
71.0
68.7
Less than 525,000
91.7
88.6
87.6
Total
100.0%
100.0%
100.0%
Source: ICF analysis for EBRI of the March 1980 Current Population Survey data. See
Appendix B for details.
"This analysis attempts to provide a perspective of the cash income potential of personal
assets. In evaluating this information, it should be understood that some elderly are unable
or unwilling to convert personal assets into cash.
s'This methodology is also described in Appendix B.
EBRI and the LTS Department of Health and Htmlan Services have jointly
funded a 1983 CPS update of the 1979 data reported on in these tables.
By mid-1984 all of this analysis will have been updated to reflect
economic security status in May, 1983.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Defining Retirement Income Objectives
by Vincent M. Tobin, F.S.A.
Group Executive
Buck Consultants, Inc.
Retirement in good health with sufficient resources to meet the needs of the
retiree and family can be a most rewarding time of life. Retirement in poor
health, or with inadequate resources, can be traumatic and devastating. As
leaders, as legislators, as employers, we are limited in the actions we can
take to assure good health for retirees. We can be more effective in our
efforts to help provide adequate financial resources to our employees if we
first adopt proper retirement income objectives.
In analyzing the financial resources available to meet the needs of our
retirees from an employer's point of view, we cannot only look at the
traditional three-legged stool of :personal savings, Social Security, and
private pension; we must also look at medical insurance, including medicare.
Furthermore, since our concern is not only with our retirees, but also with
their dependents, we must consider life insurance and other.available death
and survivor benefits. As an employer deciding upon and implementing a
retirement income policy, we must understand all aspects of the finances
available to retirees in order to best meet our and their retirement income
objectives.
SOURCES OF RETIREMENT INCOME
Of the sources of retirement income, personal savings undoubtedly has the
greatest variation among individuals - even among those working for a single
employer. In recent years, the availability to employees of Individual
Retirement Accounts will. greatly increase the portion of retirement income
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Page 2
from this source. (Some analysts consider IRAs as part of the pension leg of
the retirement stool. In form it may be that, but in substance I believe it
properly is included in the personal savings leg). IRAs for employees are
too recent to have a significant impact on the retirement income of current
retirees. However, each year their impact will increase. Certainly, by the
year 2000, IRAs will provide a meaningful portion of many retirees' financial
resources. Nevertheless, many, perhaps most, retiring employees - even in the
year 2000 - may not have .. signif icant personal savings for retirement. They
will be dependent on Social Security and employer-provided retirement income.
Employees in the private sector, employees of most state and legal
governments, and employees of many non-profit organizations have long been
covered by and eligible for Social. Security benefits. New federal employees
will be joining this group. The range in Social Security benefits among
covered employees is less than the range in other sources of retirement
income. Nevertheless, Social Security reflects a significant percentage of
salary for the lower paid employee and even provides significant amounts of
after-tax income for the higher paid employee since at least 5O% of the
Social Security benefit is tax-free.
To the extent that expenses after retirement are lower, the retiree's income
needs to maintain a given standard of living are eased. Put another way:
the absence of commuting costs and other work-related expenses, lower taxes,
and the existence of such conditions as medical coverage under Medicare
.enable retirees to maintain their pre-retirement standard of living at a
level of income below what they were earning when they retired. On the other
hand, necessities formerly paid for by the employer must now be provided by
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Page 3
the retiree out of retirement income. This can be particularly significant
for medical expenses for early retirees - before they are eligible for
Medicare. To the extent that the employers provide medical benefits after
retirement to former employees, the financial strain on their retirement
income can be eased.
Medical insurance should not be considered a budgetable expense since through
the combination of Medicare and employer-provided medical care, we have
relieved the retiree of the major burden for this item. Nevertheless, some
employees who retire early may temporarily have to incur significant medical
expenses if they lose the employer-provided benefit at retirement yet are not
eligible for Medicare coverage until age 65.
Earlier I referred to the retiree and family. This implies that the needs of
the retirees do not end with their death. The financial needs of surviving
spouses and other dependent family members continue after the retiree's death
- in some cases with a substantial reduction in income. A sound retirement
income policy takes into account the post-retirement death benefits that are
provided. This may. include either a single sum death benefit or a pension
payable to the survivor for life or over an extended period.
From the employer's viewpoint, however,, the pension, together with Social
Security, is the key starting point in determining the adequacy of retirement
income for employees. Furthermore, the pension is the only source of
retirement income provided by the employer, subject to its control, and
determinable in amount by the employer. A sound pension policy is the key to
establishing appropriate pension goals to best meet the retirement income
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Page 4
needs of the workforce at a cost level most affordable to the employers. How
these retirement income objectives are defined, and how they are achieved is
the subject of this paper.
RETIREMENT INCOME
The general objective of a retirement income policy is to satisfy the
economic needs of employees after their working career ends and they retire.
Unfortunately, there are no universally accepted objective standards to
determine when the employees' needs are satisfied. However, as I will
discuss, there are techniques that may be helpful.
The retirement income policy of an employer may be described in a number of
ways ranging from general to specific. A policy is typically found in
the provisions of the current plan, itself,
the historical development of those provisions,
the granting of pension payments which are supplemental to those
payable from the plan, and
the granting of increases in pension after retirement to. compensate
for a deficiency created by inflation.
The sum total of the foregoing provides the extent to which the employee's
pre-retirement income will be replaced by Social Security and the pension for
various ages at retirement and the degree to which that ratio will not be
eroded by inflation.
Pension policy indicates the relative importance of various items that are,
or might be considered, in the calculation of a pension benefit.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? ?
Page 5
In designing a pension benefit policy, the employer must address the
following questions:
- Is retirement income going to be provided entirely by a defined
benefit plan, entirely by a defined contribution plan, or by some
combination of both?
Are all
eriods of servi
i
d
l
d
d
h
p
ce
nc
u
e
or
oes t
e pension ignore
certain periods such as service before age 25, after age 65, after
30 years of service, etc?
Is the employee's total compensation taken into account or are
items such as overtime and bonuses excluded?
Are benefits based on career earnings or average final pay?
Does the employer pay the full cost of the retirement benefit or is
the employee expected to set aside a part of current income in
order to be eligible for part or all of the retirement plan's
benefits?
Is the plan "integrated with Social Security" or does it provide
the same percentage of pay to all employees with equal service?
Are pensions "indexed" for inflation during retirement?
Is the pension calculation oriented toward providing substantial
benefits upon early retirement, or are "full" benefits only payable
upon normal retirement?
Would the amount of normal retirement benefit be considered modest,
average or high for a long-service employee (30 or more years)? a
medium-service employee (20 to 30 years)? a short-service employee
(less than 20 years)?
What ancillary benefits are provided?
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Page 6
The answers to these questions affect the size of the retirement income, its
balance between short-service and long-service employees, between normal and
early retirements, and between higher and lower paid employees.
Type of plan
Defined benefit plans should be the foundation of a sound retirement income
policy. They are the only plans that can assure that the retiree receives
the level of periodic income which the employer's retirement income policy
attempts to provide. They are also the only plans that can relate the
retirees' pension to their income in the years closest to retirement.
Defined contribution plans can be a valuable supplement to defined benefit
plans but cannot replace them. Furthermore, it is not clear that the
retiring employee requires a capital accumulation plan to be available at
retirement. Since the topic of defined benefit and defined contribution
plans is' being addressed by other speakers, I will not discuss it further
here. The rest of the paper, except where noted, basically concerns defined
benefit plans.
Service
Crediting all service from employment to termination is generally perceived
as the most even-handed approach. Limiting service by not crediting service
before or after a certain age, or limiting the total number of years included
in the pension formula, will favor or penalize certain groups - short-service
employees, long-service employees, employees who work after normal retirement
age, employees who retire early, etc. For the long service employee it
doesn't matter if the formula is 2% a year for a maximum of 25 years or
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
Page 7
1-2/3% a year for 30 years. For the short and medium service employee it
does make a difference in their retirement income.
Earnings Definition
If the ratio of the employees' pay for pension purposes to their total pay
varies very little from employee to employee, the definition of pay for
pension purposes does not have too much of an effect provided the formula is
adjusted to produce the, appropriate level of retirement income. The
distinction is more important where variations among employees or from year
to year exist.
The pension should, of course, replace earnings that are earned during the
period. Unearned items, such as sick pay or vacation pay that are carried
over from a previous period should not inflate the earnings base and the
resulting retirement income.
Earnings Base
Retirement benefits based on final pay can be related very closely to.income
in the years immediately preceding retirement. Benefits based on career
average earnings cannot be closely related to final earnings without frequent
"updates" through plan amendments. However, career earnings plans are
updated periodically in order to reestablish the desired level of retirement
income for currently retiring employees. Final average earnings are usually
computed over a period of three or five years. In some-governmental plans,
final one or two year periods are used. In private plans periods fewer than
three years normally are not used; they can be deemed to be discriminatory by
the Internal Revenue Service.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Page 8
As just indicated, some employers with career average plans update accrued
benefits from time to time to approximately track final average plans. While
this technique can partly overcome the deficiency of this type of plan, it
can also produce dichotomies between the pensions of someone who retires just
before an update and someone who retires right after one.
Employee contributions
With the rapid increase in employee Social Security contributions
(effectively 6.7% of salary' up to $37,800 in 1984 and 7% of some higher
amount in 1985), the number of retirement plans with mandatory retirement
contributions is declining, a decline which has been occurring for at least
two decades. The recent increase in the number of employers that provide
both defined benefit and defined contribution plans (many of which have
employee contributions) has accelerated the decline in contributory defined
benefit plans.
Initially, defined benefit plans were mostly contributory, partly due to the
view that the employee should share the responsibility for saving for
retirement. The growth in retirement plan coverage after World War II,
however, resulted in the establishment of many noncontributory plans and a
decline in the popularity of contributory plans. With the increase in the
availability of retirement savings plans (e.g., IRAs) and defined
contribution plans, contributory defined benefit plans have continued to
decline -- especially in private industry. The practice is still fairly
widespread in state and local governmental plans. Now that new federal
employees will, for the first time, be contributing to Social Security, the
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Page 9
question of the level of their contributions, if any, to their own retirement
plan requires careful study and consideration.
Integration with Social Security
The integration of the retirement pension with Social Security is an issue
not only in determining the benefit formula, but also in determining the
level of contributions. One level for earnings subject to Social Security
can be adopted while a second level (presumably higher) can apply to earnings
above the FICA base. Many benefit professionals, however, believe that the
most desirable approach from a benefit design viewpoint is to directly offset
a percentage of the Social Security benefit. against the pension. While more
equitable than the "step-up" approach, the offset approach is more
administratively complex.
Adjustment for Inflation
Inflation during the employees' working career can generally be managed by
means of a pension based on average final pay. After retirement, adjustments
for inflation can be made periodically on an ad hoc basis or automatically
through an indexing formula. The ad hoc approach is common in private plans;
the automatic indexing approach is common in public plans. Although most
adjustments have been calculated based on the Consumer Price Index (CPI-U or
CPI-W), it is questionable whether the CPI accurately reflects the effect of
inflation on retirees. To date, no index for retirees has been developed.
Early Retirement
The level of early retirement benefits, as compared to the normal retirement
benefit, can act as a strong force encouraging or discouraging early
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? Page 10
retirement. If the early retirement benefit is the-actuarial equivalent of
the accrued normal retirement benefit, the formula does little to encourage
early retirement. If the benefit is greater than the actuarial equivalent,
especially when the early retirement benefit is equal to the full accrued
pension, it can encourage employees to take early retirement. The age (or
age and service combination) at which the full accrued benefit can first be
paid has a significant effect on early retirement. It is the earliest age at
which employees perceive that they are not being penalized if they retire
early.
A pension benefit policy should recognize the concept of a "proper"
retirement age such as 62, 65 or 70 and that most employees who "retire" at
age 50 or 55 do not leave the work force but seek other full- or part-time
employment. It costs at least twice as. much to provide full retirement
income starting at age 55 as starting at age 65.
Level of Benefit
The issue of the level of retirement income as compared to income just before
retirement is twofold. First, the desirable level of normal retirement
income must be defined. Second, it must be defined in terms of the
long-service employee, the medium-service employee and the short-service
employee..
Most corporate plans determine the "proper" retirement income level for the
long-service employee and give pro rata benefits to shorter service
employees. Through the use of front-loaded formulas, other plans give
somewhat more than proportionate benefits to the short-service employee. The
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? ?
Page 11
present Civil Service plan gives less than a proportionate benefit through
the use of a back-loaded formula.
Ancillary Benefits
In addition to providing retirement income, a sound retirement income policy
will address the question of providing other benefits to meet the needs of
the retiree. This can include disability income, death benefits after
retirement, and medical ,insurance in conjunction with Medicare. These
benefits can be. provided as part of the retirement program. Death benefits,
particularly joint and survivor benefits to beneficiaries of retirees, are
almost always provided as part of the retirement plan. These benefits are
often highly subsidized by the employer, as is done in the Civil Service
System, but may be paid for entirely by retirees who elect the benefit
through an actuarial reduction in their pension. Medical benefits and other
forms of death benefits are generally provided as separate benefits.
Replacement Ratio and Benefit Ratio
The various issues discussed above are reflected in the retirement benefit
formula. Together with the benefit accrual rate, they define the amount of
retirement income. However, this definition is not sufficient to determine
whether the retirement income objectives have been attached. For this,
certain ratios are useful. They are the replacement ratio and, the benefit
ratio.
The benefit ratio may be defined as the ratio of the plan benefit to gross
income in the year preceding retirement. It is most useful for comparing
benefits under the pension plans of various employers to each other at
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Page 12
various salary or service levels. However, it is not too useful for
determining whether retirement income objectives are met because it does. not
take Social Security and other retirement factors into account.
A better analysis is obtained through the use of replacement ratios, which
may be defined as the ratio of the pension plus Social Security to the
employee's spendable income immediately prior to retirement--all on an after
tax basis. This concept is discussed more fully in the next section.
REPLACEMENT RATIOS
Traditionally retirement income objectives for total plan benefits plus
primary Social Security benefits have ranged from 70% to 80% of
pre-retirement salary for lower paid employees and from 50% to 70% of pre-
retirement salary for higher paid employees. These benefit objectives would
normally.apply for an employee who spends the major part of his career (i.e.,.
30 years or more) with an employer. Lesser benefit objectives are
appropriate for shorter service employees, particularly in view of the fact
that retirement plans now have relatively liberal eligibility conditions for
vesting. Thus, many employees who join an employer later in their careers
will have'vested benefits from their previous employer or employers.
Personal savings are not directly taken into account in determining benefit
objectives since employees accumulate varying amounts of personal savings at
retirement and a large percentage of personal savings' in this country is
represented by home purchases.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
40 1 P, Page 13
In any event, the generally accepted basic goal for a retirement program is
to provide retirement income from all sources that allows the retired
employee to maintain a standard of living that is reasonably consistent with
his standard of living before retirement. In determining the amount of
retirement income that this goal necessitates, it should be recognized that
when employees retire:
they no longer have work-related expenses such as clothing,
commuting costs and meals away from home,
they probably will be eligible for Medicare upon attaining age 65,
tax exemptions increase upon attaining age 65 and a larger portion.
of their income is taxed at lower rates (at least 50% and in many
cases 100% of the Social Security benefit is tax free)
the cost of supporting children has probably ceased,
mortgage payments have probably stopped, and
life insurance premiums may have been reduced or eliminated.
The cessation of the latter three items, while significant, obviously does
not coincide with retirement. If these costs exist immediately before
retirement, in most cases (other than premium paid up at 65) they will
continue for a time after retirement. On the other hand, if they have ceased
before retirement, the employees will have adjusted their standard of living
accordingly.
After retirement employees have three sources of income:
- their retirement benefit under the employer's retirement program,
- Social Security, and
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
is 10 Page 14
the income which can be earned on accumulated personal savings or
through the liquidation of a portion of personal savings.
A useful measure of the adequacy of this income is the replacement ratio.
The term replacement ratio applies to the percentage that retirement income
provided by a retirement plan and by Social Security is of an employee's
income immediately prior to retirement on an after-tax spendable income
basis.
Pre-retirement.spendable income
Pre-retirement spendable income may be defined as gross earnings less the
following:
Federal and state income taxes
Social Security tax
Work-related expenses
Life insurance costs
Medical insurance costs.
Post-retirement spendable income
Post-retirement spendable income may be defined as the
Pension from the employer's plan, plus
Social Security benefits, less
Federal and state income taxes, less
Medicare premiums.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Page 15
As an example -
Employee earns - $25,000
Company pension - $ 8,000
However - Income taxes on the pension might be $100 and as much as $2,500 on
the $25,000 in wages. Since total retirement income is less than $25,000,
there would be no income tax on the Social Security benefit.
Adding - Social Security Benefit - $ 8,000
Social Security Taxes - $ 1,675
Commuting Costs - $ 725
We get the following:
Take home pay
Spendable Retirement Income
Replacement Ratio - $15,900
$20,100 - 79%
$25,000 (wages)
-2,500 (taxes)
-1,675 (Soc. Sec.)
- 725 (commuting)
$20,100
$ 8,000 (company pension)
+8,000 (Soc. Sec.)
- 100 (taxes)
$15,900
A pretty good plant
Approved For Release 2008/08/21 : CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
9 1P
Page 16
In this example we have used only the primary Social Security benefit. If
the retiree has a spouse, the Social Security pension will be 50% bigger, and
the replacement ratio would be over 100%.
The question of whether to include the spouse's Social Security in the
calculation is a philosophical one which has to be addressed in the
employer's retirement policy. Certainly the spouse's Social Security benefit
is part of the total retirement income of the couple. To ignore it is to
understate the total retirement income of the couple.
On the other hand, employees receive the same pension from the employer
whether they are single or married. If the spouse's Social Security benefit
is taken into account, the pension plan that is adequate for married
employees will be inadequate for single employees. The employer's retirement
income objectives should try to balance the needs of single and married
employees on this issue.
CONCLUSION
The goal of a retirement income policy is to meet the retirement income needs
of the workforce at a cost level most affordable to the employer. There are
many issues to be decided regarding service, compensation, age at retirement,
integration with Social Security, etc. The calculation of the replacement
ratio based on the retirement benefit as defined by the answers to the stated
issues is the best measure of how well the employer's retirement income
objectives have been met.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0
DEFINED BENEFIT PLANS
S
Paper Presented by Dan M. McGill
before the Policy Forum
of the Senate Governmental Affairs Committee
December 13, 1983
Other participants in this educational forum are address-
ing the setting of retirement income objectives and the sources
of income that may be available to meet those objectives. It
is generally agreed that retirement income objectives for any
group of employees should be set at a level sufficient to per-
mit a retired person and spouse to enjoy a standard of living
throughout retirement roughly equivalent to that which they
enjoyed during the years immediately prior to retirement. Gen-
erally'speaking, retirement income from all sources in the
range of 50 to 75 percent of gross compensation at time of re-
tirement, dependent upon the level of earnings, will enable a
retired individual to enjoy a post-retirement standard of living
reasonably comparable to his or her preretirement standard of
living. The percentage of preretirement income needed declines
as gross earnings increase, primarily because of tax effects.
It would appear that throughout the foreseeable future,
the primary sources of retirement income for the bulk of the
populace will be Social Security and employer-sponsored pension
plans. In undertaking to meet some of its employees' old-age
financial security needs,'an employer has the broad choice of
*Portions of this paper have been adapted from materials that
I prepared for Winklevoss and McGill, Public Pension Plans, pub-
lished i Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? ?
promising a set of retirement benefits, subject to specified
conditions, or promising to set aside funds for the employees
on a stipulated basis, with the intent that the balance in an
employee's account at time of retirement be sufficient to
provide the targeted level of income. The latter approach is
described as a defined contribution plan and the former as a
defined benefit plan. Defined contribution plans predominate
in number but defined benefit plans account for the bulk of
coverage and pension assets.
The remainder of this paper is concerned with the benefit
design and funding of defined benefit plans.
A. Benefit Design
The core of a defined benefit plan is a benefit formula
that over time and under a given set of circumstances determines
the amount of benefits that a plan participant will receive at
retirement. Broadly conceived, a benefit formula is composed
of three elements: (1) a description of the basis for determin-
ing annual benefit accruals, (2) a statement of the conditions
under which the benefits will be payable, and (3) a stipulation
of the form in which the benefits will be payable. Stated dif-
ferently, the benefit accrual for a given year of credited
service must be defined in terms of the dollar amount
of the annual benefit accrual, the age of the
participant at which the aggregated benefits will become pay-
able in full, and the annuity form in which the benefits will
be paid unless an alternate form of payment is elected by the
participant.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? ?
Amount of Benefits
All defined benefit plans explicitly or implicitly assign
a unit of benefit to each year of recognized service with the
employer. The unit of benefit may be expressed as a percentage
of compensation, as defined -- the usual procedure under a plan
for salaried employees -- or as a specific dollar amount, a
type of formula associated with a collectively bargained plan
for hourly employees.
Under an earnings-related formula, the unit of benefit
credited for any particular year of employment may be based
upon the participant's compensation for that year or upon the
participant's average compensation during .a specified period,
such as three, five, or ten years prior to retirement. The
first type of formula is called a career average formula, while
the second is referred to as a final average formula. A common
modification of the final average approach is to base the ben-
efit on a specified period of consecutive years of highest aver-
age compensation, whether or not the period fell immediately be-
fore retirement.
The principal appeal of the final average formula is that
it automatically provides benefits appropriately related to the
participant's compensation during the years close to retirement.
In other words, it protects the accruing benefits against loss
of purchasing power because of inflation. This is a result
normally desired by both the plan sponsor and the plan partici-
pant, but it creates uncertainty for the sponsor as to the
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
financial magnitude of its undertaking. The cost is reasonably
predictable as a percentage of covered payroll, a more logical
measure of cost than dollar outlay.
The dollar value of a career average benefit accrual is
known at all times but its future purchasing power is not known
nor predictable with any precision. In a period of steady and
substantial inflation, the purchasing power of career average
benefits can be seriously eroded. For example, if there was
a constant 6 percent inflation rate during the entire period
of service of a participant who entered the plan at age 25,
the benefit produced at age 65 by a career average formula would
have only 29 percent of the purchasing power of the benefit pro-
duced by a 5-year final average formula. For a 30-year partic-
ipant, the career average benefit would still be only 43 percent
of the final five benefit, assuming salaries reflect inflation
in full. To counteract this erosion, the sponsor of a career
average plan may amend the plan from time to time to restate
the accrued benefits in terms of current compensation. If done
,frequently enough, this practice will protect accrued benefits
reasonably well -- without an advance commitment from the spon-
sor -- but it can generate substantial unfuhded liabilities.
As noted above, many collectively bargained plans express
the benefit accrual as X dollars of monthly income per year of
crodited service. This is an acceptable approach in industries
where the range of hourly wage rates is relatively narrow. The
current monthly benefit in many industries is $15 or more for each
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
year of credited service. The benefit is generally renegotiated
every three years and made retroactive to all prior years of
credited service to preserve the purchasing power of the accrued
benefits.
Most defined benefit plans grant benefit credit for some
or all years of service prior to establishment of the plan. This
is necessary in order to provide an adequate retirement benefit
to those with many years of service and possibly even.approach-
ing retirement. Indeed, this is the genius of the defined ben-
efit plan. It is a way to provide adequate pensions to all
employees irrespective of their ages or years of service at in-
ception of the plan. It results in an effective allocation of
employer pension contributions that favors older and longer
service employees. In the process, substantial supplemental
liabilities are created, the amortization of which may extend
.over 30 years. This resource allocation capability'of a defined
benefit is not needed, it may be observed, for a group of newly
hired employees.
In determining the benefit structure of a defined benefit
plan, the plan sponsor (and collective bargaining representatives,
if any) will usually take cognizance of the benefits available
under Social Security. This recognition is described as integra-
tion of the plan with Social Security. If provision is made
in the plan for a modification of the plan benefits to recog-.
nize Social Security benefits, the plan is said to be explicitly
integrated. If the plan benefits are initially set at a lower
level than they would be in the absence of Social Security, the
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
plan is said to be implicitly integrated. Implicit integration
is feasible only for plans covering groups with relatively flat
wage rates. Integration is carried out in recognition of the
fact that Social Security benefits favor lower income workers
and operate to reduce the need for plan benefits, as a percent-
age of compensation. Full integration is a significant.source
of cost-saving for the plan sponsor.
Retirement Age
The age at which a participant is permitted to retire with
full, unreduced benefits is an essential component of the ben-
efit formulation and is a prime determinant of the cost of the
plan. This is identified as the normal retirement age. Most
plans permit participants to retire over a range of ages, such
as from age 55 to 70, usually with some adjustments to the normal
retirement benefit. Thus, there may be early, normal, and de-
ferred retirement.
Normal Retirement Acme. At one time a simple notion, normal
retirement age has become an elusive concept in many pension
plans. In its most elemental form, the normal retirement age
is the earliest age at which eligible participants are permitted
to retire with full benefits. A more precise definition is
provided in Section 2.08 of Revenue Ruling 71-446: "Normal re-
tirement age is the lowest age specified in a plan at which the
employee has the right to retire without the consent of the
employer and receive retirement benefits based on service to
date of retirement at the full rate for such service set forth
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
. .
in the plan" (i.e., without actuarial or similar reduction be-
cause of retirement before some later specified date).
In the private sector, the most common single age specified
as the normal retirement age is 65. Under ERISA, the normal re-
tirement age cannot be higher than 65 or, if later, the age at
which the employee completes ten years of service. it is cus-
tomary to require a minimum of five or ten years of credited
service for entitlement to benefits at normal retirement age.
Because of'a general desire on the part of employees to enter
on their pensions at ever younger ages, pressure from younger
employees to remove older workers from the labor force, and the
desire of management to ease employees who have lost their
effectiveness from the payroll, the simple concept of a single,
designated normal retirement age has been supplanted in many
plans and many sectors of the economy by provisions that permit
employees under defined circumstances to qualify for full accru-
ed benefits at ages lower than that specified as the normal re-
tirement age. For example, full unreduced benefits for service
accrued to date may be made available at age 60 with 20 years
of service or at age 55 with 30 years of service. Under these
arrangements, it would have to be concluded that there are multi-
ple normal retirement ages, depending upon the related service
required and the employee's age of entry into the plan.
In practice the choice of the normal retirement age must
be made in the light of cost, personnel policy, and public wel-
fare considerations. In terms of pension costs the normal re-
tirement age is one of the most critical features of plan design.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? .
As the plan's normal retirement age is decreased, there is an
associated increase not only in the proportion of employees who
will survive in active service to reach retirement but also in
the average number of years over which each pensioner will re-
ceive benefits. The effect of these changes on plan costs is
indicated in Table 1. The table shows the multiple of the level
annual percentage of salary needed to fund members' retirement
at age 65 that would be required to fund retirement at various
other ages.' For a member entering the plan at age 35, the
annual percentage of the employee's salary required to fund re-
tirement age age 55 is more than twice that required to fund
retirement at age 65. This is true even though the member will
have accumulated fewer benefit accruals at the earlier age.
Table 1 also shows that the relative cost of retiring employees
at different ages is not overly sensitive to the entry ages of
the employees.
Table I
Sensitivity of the Cost of Retirement Benefits to the
Normal Retirement Age
Level annual percentage of salary re.
quired to fund retirement benefits
Normal
commencing at various ages, as a
Retirement
multiple of the age 65 value!
Age
Entry Age
25
35
45
50
2-8
2.6
2.5
55
2.2
2.1
2.0
60
1.6
1.5
1.5
62
1.3
1.3
1.3
65
1.0
1.0
1.0
70
0.5
0.5
0.6
a Values based on the implicitly integrated prototype plan.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
46 46
Another way to view the cost impact of the normal retire-
ment age is to compute benefit accrual equivalencies among var-
ious retirement ages. This has been done in Table 2. For the
purpose of constructing the second column of the table, it is
assumed that a plan initially provides for an annual benefit
accrual of 1.5 percent of final three-year average salary pay-
able at age 65 in the form of a single life annuity. The column
shows the benefit accruals that would generate an equal cost
to the employer if the benefit were to commence at any age from
64 down to 50. The third column shows the benefit accrual for
normal retirement age age 65 which would cost the employer the
same amount as an accrual rate of 1.5 percent of final three-
year average salary if the benefit were to commence at the
earlier ages shown.
Table ,2
Comparative Benefit Accrual Rates for Various Male Normal Retirement Ages*
Accrual rates up to various Accrual rates up to age 65 that
normal retirement ages that generate the same costs as
Normal generate the same costs as an accrual rate of 1.5 percent
Retirement an accrual rate of 1.5 percent up to various other normal
Age up to age 65 retirement ages
65 . . . . . . . .
1.50%
1.50%
64 . . . . . . . .
1.46
1.54
63 . . . . . .
1.42
1.58
62 . . . . . . .
1.39-
1.62
61 . . . . . . . .
1.36
1.66
60 . . . . . . . .
1.32
1.70
59 . . . . . . . .
1.30
1.74
58 . . . . . . . .
1.27
1.77
57 . . . . . . . .
1.25
1.81
56 . . . . . . . .
1.22
1.84
55 . . . . . . . .
1.20
1.87
54 . . . . . . . .
1.18
1.90
53 . . . . . . . .
1.16
1.93
52 . . . . . . . .
1.15
1.96
51 . . . . . . . .
1.13
1.99
50 . . . . . . . .
1.12
2.02
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
The table indicates that a benefit of 1.5 percent per year
of service payable from age 65 costs the same amount as an
accrual of 1.32 percent payable at age 60 and 1.20 percent pay-
able at 55. Conversely, a benefit of 1.5 percent per year of
service payable from age 55 costs as much as an accrual of 1.87
percent payable from age 65. A benefit of 1.5 percent payable
at age 60 would be equivalent to an accrual of 1.70 percent
payable from 65. Thus, it may be observed that benefit accrual.
rates that seem perfectly defensible when payable at a particu-
lar retirement age are seen to be excessive from a cost stand-
point when viewed in the context of a normal retirement age that
has tended to become the norm for plans in the private sector and may have to become the norm for public plans.
With respect to the pension plan's function as an instrument
of personnel policy, the normal retirement age is a statement of
the employer's desires concerning the timing of employee retire-
ment. If the employer wants employees to retire at relatively
young ages, then this should be encouraged by incorporating a
low normal retirement age in the pension plan or by providing
very generous retirement privileges prior to normal retirement.
If the employer wants the employees to remain on the job as
long as they are effective, then the normal retirement age should
be pushed back and attractive deferred retirement provisions
should be provided. The normal retirement age need not be view-
ed as the mandatory retirement age.
The normal retirement age should not be set in the absence
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
w ?
of economic and societal. considerations. From a purely economic
standpoint employees in both the public and private sectors
should continue in active employment as long as their mental
and physical powers permit. Society as a whole would benefit
from the increased flow of economic goods and services from pro-
longed employment, and the physical and psychological well-being
of many individuals would be enhanced by a continuation of mean-
ingful economic activity. Of course, if the economy is not ex-
pected to generate.-enough jobs to provide employment for all
present and future members of the labor force, a lower retire-
ment age might be needed to remove the older workers from the
labor force and permit their places to be taken by persons just
entering the labor market. it could also be argued that society
might prefer more leisure time to pursue recreational, cultural,
and other noneconomic interests, even if it means a lowering
of their financial standard of living. In other words, society
might logically sanction an institutional arrangement under
which members of the labor force would be afforded the opportun-
ity to spend many years in full-time noneconomic pursuits at
the end of a moderately long service career. Such a societal
judgment would obviously involve a balancing of economic and
noneconomic considerations.
Unfortunately, there is no body of economic, political, or
social theory that points unequivocally to the optimum normal
retirement age from a societal standpoint. Most business firms
chose age 65 because of the precedent set by the Social Security
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 0
Act in 1935. Today, life expectancy is much longer, general
health is improved through medical advances, and technology
has reduced the physical demands of most jobs. These factors
alone argue for a higher retirement age than that which has
been institutionalized in this country.
An even stronger argument for a later retirement age is
found in the changing demography of.this country. The drastic
decline in fertility rates that has occurred over the last
decade and'seems likely to be sustained throughout the immediate
future will ultimately produce a situation in which a shrink-
ing labor force must support an expanding retired population.
it has been predicted that if the fertility rates remain at
their present level until the year 2010, when the individuals
born in the "baby boom" of the late 1940s begin to reach re-
tirement, there will be one Social Security beneficiary (includ-
ing children and other secondary benefit recipients as well as
retired persons) for every two persons in the active labor
force. This would put a tremendous strain on the available
goods and services unless an unprecedented increase in produc-
tivity were to occur in the meantime. While these economic and
demographic forces do not argue for an immediate raising of
the normal retirement age in all existing pension plans, they
do argue for a halt to the trend toward ever lower retirement
ages and a move towards higher retirement ages.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
S ?
Early Retirement. In order to further utilize the pension
plan as an instrument of personnel policy, an employer may
choose to permit employees. to retire with reduced benefits at
ages prior to the normal retirement age. A plan permitting
normal retirement at, age 60, for example, might provide for
retirement with reduced benefits at age 55 if the employee has
accumulated ten years of creditable service.
The early retirement benefit is generally calculated as a
percentage'of the.accrued portion of the benefit that would
have been paid at the normal retirement date. Some plans pro-
vide for a stipulated percentage discount for each month by
which actual retirement precedes the normal retirement date.
Typical discounts are one half of one percent per month (6 per-
cent per year) or 1/180 per month (6 2/3 percent per year). On
the other hand the discount factors may depend on the interest
and mortality assumptions that the actuary uses to value the
plan costs and liabilities, in which case the reduction is said
to be actuarial.
An arbitrary scale of reduction factors, particularly the
one half of one percent reduction per_.month, is easier to ex-
plain to participants than the full actuarial reduction. It
is frequently designed to encourage early retirement, being
coordinated with the overall personnel policy of the employer.
..The general practice of using early retirement factors more
favorable to the employee than the full actuarial reduction
is referred to as subsidized early retirement.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
?
The cost to the employer of an early retirement program is
not easily ascertainable since it depends on several factors
which have offsetting characteristics. Moreover, some of
these factors are dependent on the age and sex mix of the plan
and the other provisions of the plan. In the first place there
is the age at which benefits commence. The cost of a specified
monthly income beginning at age 65 is significantly less than
the cost of the same benefit payable from age 60. On the other
hand an employee who retires early gives up future creditable
service as well as future salary increases; therefore, the
employee's full benefit (i.e., before reduction on account of
early retirement) will be appreciably less than if taken at
the normal retirement date. Then there is the consideration
of the precise reduction factors that are applied to the accrued
benefits and the amount of retirement income subsidy which the
employer is willing to support. Also there is the question of
possible benefit increases based on cost-of-living indices that
may enhance the value of the earlier income stream. There is
one final factor, the impact of which cannot be ascertained
from the pension plan balance sheet. If employees are encour-
aged to retire at earlier ages, there is an associated cost of
payroll for new entrants to the plan. While it is true that
the employer saves the cost of the payroll for the persons who
retired, it may be that younger employees must be promoted at
a faster rate, thereby minimizing and possibly negating poten-
tial payroll savings.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0
0
Deferred Retirement. Participants in s pension plan may
want to continue in service beyond the normal retirement
age in'order to earn additional benefit credits, to enlarge
the salary base to which the benefit formula will apply, to
spread the liquidation of the accumulated assets of a money
purchase plan over. a shorter period of years (and thus increase
the amount of the periodic payments), or to enjoy the contin-
uation of their salary. There may also be nonfinancial reasons
for wanting to continue on the job.
Prior to the 1978 Amendments to the Age Discrimination in
Employment Act, corporate pension plans generally stipulated
that a participant could continue in service beyond normal
retirement date only.with the consent of the employer and then
only to the mandatory retirement age; which might be age 68 or
70. Practice'varied as to whether an individual permitted to
continue in employment would receive benefit credits for the
additional years of service and would have the benefit base
adjusted for salary increases. The retirement systems of pub-
lic bodies tended to permit participants to remain in service
beyond.the normal retirement age without the employer's con-
sent, up to a specified mandatory retirement age, such as 70.
Benefits for the additional years of service accrued in the
:normal manner, and salary increases were recognized for bene-
fit compensation purposes.
The 1978 Amendments to the Age Discrimination in Employment
Act (ADEA) made it unlawful for an employee benefit plan or
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0
0)
seniority system of a private sector employer or a state or
local government entity to require or permit the involuntary
retirement of an employee prior to age 70. The law does not
require the plan sponsor to credit benefits for service beyond
normal retirement age, to recognize salary increases for ben-
efit computation purposes, or to make contributions under a
defined benefit or defined contribution plan on behalf of
employees who continue in service beyond normal retirement
age. Nor does the Act require an actuarial adjustment.in the
employee's benefits to reflect his actual age at retirement.
There is a wide range of opinion as to what benefit an
employee who delays retirement beyond normal retirement age
should receive upon actual retirement, and practices current-
ly reflect this diversity of opinion. At one extreme is the
view that the employee should receive precisely the same dollar
benefit that the employee would have been entitled to had he
retired on the normal date. The basic argument in favor of
this position is that the plan promised a benefit payable from
a specified date, which the employee could have received by
retiring on that date without the consent of the employer. By
continuing to work, the employee enjoyed the economic benefits
of a full salary, which would typically be much larger than a
pension. Having enjoyed that advantage with the approval of
the employer, the employee has no right to claim a higher pen-
sion upon eventual retirement. Under this approach, delayed
retirement reduces the employer's cost by the benefits that
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0
?
would have been paid had the participant retired at the normal
retirement age, and the employer incurs no cost for pension
accruals in respect of the participant after the normal re-
tirement age. This practice is permissible under ERISA and
the amended ADEA.
At the other extreme is the view that the employee should
upon actual retirement receive benefits that are the actuarial
equivalent of those that would have been payable at the normal
retirement.date, adjusted for any increase in the salary base
and augmented by additional benefit credits at the regular
rate. This approach offers a strong inducement for employees
to continue in service beyond the normal retirement date.
Between these extremes a number of intermediate positions
may be found. Table 3 displays the impact of several of these
intermediate deferred retirement policies on the amount of
monthly benefit receivable at ages after the normal retirement
age (i.e., age 65). The first column of the table shows the
impact of allowing the deferred benefit to be based on the up-
dated final average salary of the retiring employee. The in-
crease in the retirement benefit is seen to be 5 percent per
year, which is the assumed rate of increase in salaries at the
older ages. The next column indicates the impact of adjusting
.the retiring employee's benefits for only the benefit accruals
earned subsequent to the normal retirement date. Since it is
assumed that the retiring employee entered the plan at age 30,
by retiring at age 68, for example, he would have accrued 38
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
a .
Table 3
Effect of Deferred Retirement Provisions on Retirement Income
Receivable
Retirement benefit payable as a percentage of benefit
payable at age 65'
Deferred
Allow
Adjust
Retirement
Update
Additional
Provide
for Salary,
Age
Final
Accrued
Actuarial
Service,
Average
Benefits
Equivalence
and
Salary Only
Only
Only
Equivalence
65
100%
100%
100%
100%
66
105
103
114
123
67
110
106
129
151
68
116
109
148
186
69
122
111
169
229
70
128
114
195
285
? Table is based on implicitly integrated prototype plan.
Assumptions: Age 30 entrant.
Salaries increase at S percent per year plus merit.
benefit is straight life annuity increasing by 4 percent per annum.
instead of 35 years of service, but this additional accrual
would be applied to the final- average salary as of the normal
retirement date. The third column exhibits the effect of pro-.
viding the actuarial equivalent of a benefit which could have
been received at age 65. Under this policy, the cost to the
employer is independent of the retirement age. The benefit
amount is increased primarily because of the shortened payout
period but is also dependent on the interest and mortality
assumptions used by the actuary to value the plan liabilities.
Note that the benefit amount is not affected by an increase
in the final average salary, nor by.additional earned credit-
able service of the employee. The table shows that this defer-
red retirement policy provides a strong inducement to the employ-
ee to remain in employment after the normal retirement date.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 0
Finally, the last column reveals the impact on retire-
ment benefits of an extreme deferred retirement policy pro-
viding credit to the employee for additional service, an
increased salary base, and the reduced payout period. The
impact is to great that the. proportion of preretirement earn-
ings provided in deferred retirement would probably be sub-
stantially in excess of any logical targets. In any event
the employer's attitude toward the delayed retirement pro-
vision maybe influenced not only by philosophical and actu-
arial considerations, but also by personnel policy; that is,
whether the employer wants to encourage or discourage the
employees to continue in employment beyond the normal retire-
ment date.
Normal Annuity Form. As noted earlier, the benefits
under a pension plan are established and their associated costs
calculated on the premise that the benefit payments will con-
form to a particular pattern. This pattern is known as the
normal annuity form, even though the plan may be funded through
a trust and the benefits paid directly from the trust fund
rather than in the form of insurance company annuities. The
normal annuity form specified in most noncontributory plans
is the straight life annuity, which provides monthly payments
to a single, designated individual (the employee in a pension
plan), for as long as the employee lives, with no payments to
the estate of the deceased or any other person. Contributory
plans usually adopt a modified cash refund annuity. This form
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0
promises that should the employee die before receiving retire-
ment benefits equal to the accumulated value at retirement of
contributions, the difference between the benefits paid and
the accumulated contributions will be refunded in a lump sum
to the estate or to a designated beneficiary. Some contribu-
tory plans prescribe a life annuity with payments guaranteed
for five or ten years, either form of which will generally
ensure the return of the employee's accumulated contributions.
Pension plans have traditionally given the participant the
option of electing, either before or at retirement and at
his own expense, an annuity form different from that prescrib-
ed in the plan document. The range of options has differed,
some plans offering a wide choice and others being rather
restrictive, but it has been customary to offer some form of
joint and survivor annuity in order that the participant might
assure his spouse of a life income in some amount. One type
of joint and survivor annuity provides for an initial benefit
amount to be paid while the pensioner and spouse are both
living and an ultimate amount (usually less than the initial
amount) to be paid after either recipient dies. A second type
of annuity provides that the, initial amount will be paid as
long as the plan member is living and that the ultimate amount
will be paid only if the spouse of the member is the only
living recipient. This form is called a joint and contingent
survivor annuity.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? is
In response to increasing concern over the general fail-
ure of plan participants to elect voluntarily a joint and
survivor annuity for the protection of their spouses, Congress
ordained through ERISA that all qualified pension plans must
provide that retirement benefits payable to an employee married
to a current spouse for at least one year will be automatical-
ly paid in the form of a qualified joint and survivor annuity
form unless the participant elects otherwise. A qualified form
is a joint' and survivor contingent annuity that provides in-
come to the surviving spouse in an amount equal to at least
one half of the income payable during the time that the employ-
ee and spouse are both alive.
In the absence of a provision to the contrary, the spouse
would lose her interest in the joint and survivor annuity if
the participant were to die before retirement. This is con-
sidered to be a particularly inequitable consequence if the
participant dies :.after becoming eligible for early retirement
but before entering on the joint and survivor annuity. ERISA
now requires that a participant be permitted to make an election
that would have the effect of providing joint and survivor
annuity benefits to the spouse if the participant dies after
eligibility for early retirement (within ten years of normal
retirement) but before actual retirement. Legislation current-
ly before Congress would extend this protection to all vested
benefits and provide a surviving spouse with one-half of the
participant's vested benefits if he should die before retirement.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
The financial significance of providing normal annuity
forms other than a straight life annuity can be appreciated
by reference to Table 4. For various ages at retirement, the
table displays the cost of the alternative forms as a per-
centage of the cost of the straight life annuity. The least
costly alternative annuity form is seen to be the five-year
certain annuity which guarantees payment for five years. This
is because about 90 percent of the annuitants who retire at
age 65 will survive the five-year period, thereby rendering
the guarantee of little value. The additional cost of the ten-
year certain annuity, on the other hand, is seen to be sensi-
tive to the retirement age. It is not unlikely that the guar-
antee will serve to extend the payment period for the older
retirees. The cost of the modified cash refund annuity is
based on the assumption that employee contributions of 5 percent
of salary have been paid from age 30 and have been credited
with 7 percent interest per annum. Under these assumptions the
accumulated contributions are fully paid out after a period of
Table 4
Cost of Various Annuity Forms as a Percentage of the Cost of a Starigftt We
Annuity
Annuit
Retirement Age
y
Form
55
60
62 - .
70
Straight Life
100%
100%
100%
1~%
100%
Five-Year certain
101
101
101
102
104
Ten Year certain
102
104
106
108
116
Modified Cash Refund
101
102
102
103
108
Joint and One-Hail Survivor
112
114
115
117
119
Joint and One-Half Contingent
Survivor
116
119
121
123
127
Joint and Full Survivor
133
139
142
146
153
Aasumptions: Employee contributions made from age 30 at S percent of aalary, eaeditad with
7 pcrveat into at per annum.
Femioov is male with spotue 3 years younger.
Annual payment inctaso by 4 pa=t per annum.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
between five and ten years, so that the cost of this form lies
between the costs of the 5-year and 10-year certain annuities.
The additional costs of three types of joint and survivor
annuity forms are displayed in the table. As should be expect-
ed, the cost of reducing the benefit to one half the initial
amount when either recipient dies (joint and one-half survivor)
is less than the cost of reducing the benefit amount only when
the spouse of the plan member dies (joint and one-half con-
tingent survivor). The most costly of the annuity forms shown
in Table 4 is the joint and full survivor which provides for
the initial benefit amount to be continued as long as either
spouse is living.
A perennial issue in plan design is whether a participant
upon reaching retirement should be permitted to take the actu-
arial value of retirement benefit in the form of a lump sum
rather than in monthly payments spread over remaining lifetime.
Justification for the cash option is generally couched in
terms of flexibility of financial. planning. It is argued that
some employees have a more urgent need for a lump sum rather
than for a life income. They may need.the money for medical
treatment or to buy a retirement home. Some may want to in-
vest in a business of their own. Others may. feel that they can
invest their share of the plan assets more profitably than the"
investment manager or in a way that will provide more protec-
tion against inflation. In some cases, the pension benefit
may be too small to justify installment payments, whilein others
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? ?
it may be so large that the participant should be permitted
to draw down some of it in a lump sum. Under some plans, the
cash option may have been the only way that an employee in
poor health could preserve his pension for the protection of
a spouse or other dependents. In the final analysis the
employer must decide whether the pension plan is to be re-
garded as a general savings program with all the flexibility
that one would want in such a program or'as an instrument of
business and social policy designed to ensure a dependable
source of income throughout the remaining lifetime of retired
workers.
The cash option and all annuity options involve the con-
cept of actuarial equivalence. That is, any optional form of
benefit payment must have the same actuarial value as the nor-
mal retirement benefit and any other optional benefit. In-
asmuch as the concept of actuarial equivalence can be rather
imprecise or "flexible" in application, the Internal Revenue
Service has decreed that beginning January 1, 1984 all plans
must either specify the specific adjustment factors that will
be applied in optional benefit settlements or the actuarial
assumptions (mainly mortality and interest) that will be used.
This is in implementation of the requirement that benefits be
"definitely determinable." The ruling was primarily designed
to prevent abuses in the use of "market" rates of interest in
calculating the cash out or lump sum value of accrued benefits.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? 0
The IRS ruling does not bar the use of current interest rates
but if they are to be used, they must be determined in accord-
ance with an objective standard articulated in the plan.
Cost-of-Living Adjustments
If one subscribes to the notion that at point of retire-
ment an individual should have a flow of income sufficient to
sustain his or her preretirement standard of living, it is
difficult to argue against the logic of protecting the purchas-
ing power of that.income during the individual's retirement
years. The erosive impact of inflation on the purchasing
power of any flow of income is all too familiar to this gener-
ation but Table 5 provides perspective in terms of pension
benefits.
Table- 5
Purchasing Power of Benefits under Selected Rates of Inflation and after Selectb
me Periods as a Percentage of Purchasing Power In a Nonlnflatlor y
Ti
Environment
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? 0
The table shows, for example, that with a steady 6 percent
rate of inflation, a stream of pension income will lose a fourth
of its purchasing power within five years and almost half by
the end of ten years. Within fifteen years after retirement
-- the approximate life expectancy of a male aged sixty-five --
the purchasing power of an individual's pension benefits will
be only 42 percent of its initial value. Naturally, the ero-
sion of purchasing power is less with lower levels of inflation.
The same type of erosion occurs with respect to the vested
benefits of terminated participants, since typically the de-
ferred benefits of such persons are frozen at their preterm-
ination level. The loss of purchasing power is a function of
the elapsed time between termination and retirement and the
rate of inflation. This may be seen in Table 6.
Table 6
Loss of Income Attributable to Nonindexed Vested
Deferred Benefits
Number of Vested
Annual Benefit
Accruals Earned
Initial Monthly Income
Received as a Percent-
il
f I
A
bl
First
Second
a
ncome
va
age o
e
from Sole Pension Plan*
Plan
Plan
0
35
100%
5
30
88
10
25
78
15
20
70
20
15
68
25
10
70
30
5
81
? FEOploytx Ii aasumed to ester awed plan immediately upon
withdrawal from first plan. Salary is awamed to incease by S per.
amt plus merit per year. aotb plans are assumed to have id=bad
bandit structure, including full vatiag of acwih after five yeah
of creditable eervicc.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? ?
This table assumes that a particular individual earns
benefit credits in two successive pension plans having iden-
tical.benefit structures, the benefits in the earlier plan
being vested. if an individual accumulated ten years of pen-
sion credits in the first plan and twenty-five in the second,
the combined benefits of the two plans at retirement will
amount to only 78 percent of the benefits payable if all the
service had been in the same plan, assuming that salaries in-
crease by 5 percent plus merit each year. If the person had
had twenty years of service in the first plan and fifteen in
the successor plan, the combined benefit would be only 68 per-
cent of the amount that would have been paid if all service
had been in one plan.
Employers are not psychologically attuned to indexing or
adjusting the benefits of individuals who have left their employ,
presumably in search of greater economic opportunities. Never-
theless, it is interesting to note that on the basis of typical
experience and 5 percent sustained inflation, the vested ben-
efits of a plan could be fully indexed at an additional plan
cost of 10.8 percent.
It is expensive to index or otherwise adjust the benefits
of retired persons to protect their purchasing power. At a
5 percent level of inflation, full indexing of benefits of re-
tired persons would increase the level percentage of payroll
cost of a plan by about 54 percent. Annual adjustments of 10
percent would more than double the long-run cost of the plan.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Decisions concerning indexing depend importantly on judgments
about the allocation of economic resources between the work-
ing and non-working elements of the adult population. This
issue has both economic and ethical implications, not to
mention the political. If indexing is to be undertaken, a
choice must be made as between wage and price indexing, as
well as among the various indexes that are available or could
be constructed.
B. Funding
If a pension plan promises a set of determinable benefits
-- as a defined benefit plan does -- the sponsor assumes re-
sponsibility for accumulating enough plan assets to pay the
promised benefits. The cost: of a set of benefits can be de-
termined only in retrospect, so the funding policy of the plan
sponsor must be based on the best estimates of future costs
that can be provided through actuarial assumptions and tech-
niques.
The plan participants may be asked to bear a portion of
the plan costs through mandatory contributions, especially in
the public sector, but the sponsor -- i.e., the business enter-
prise or employing agency -- has the ultimate responsibility
of assuring asset adequacy. Consequently, the plan sponsor
usually sets the investment policy of the plan and is credited
with any investment earnings in excess of those assumed to be
earned in the actuarial cost projections. By the same token,
the sponsor must make up any deficiency in investment earnings
as compared to projections. It is relevant to note that if a
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
49 ?
fully funded plan consistently earns a 6 percent rate of re-
turn, approximately 70 percent of plan costs will be met out
of investment earnings. Stated differently, 70 percent of
the benefits will be paid out of investment income rather
than plan contributions.
Plans subject to ERISA must set aside enough money to fund
the normal cost of the plan currently and to amortize the
supplemental liability, sometimes referred to as the "unfunded
actuarial liability," over a maximum period of thirty years.
Actuarial losses must be funded over a maximum of fifteen years.
Thus far, plans in the public sector are not subject to any
mandatory funding standards.
Designed to assure the actuarial soundness of the pension
undertaking, these statutory funding standards and their legal
underpinnings constitute the principal objection to a defined
benefit plan from the standpoint of existing and potential
plan sponsors. Under existing law, if a single employer defin-
ed benefit pension plan terminates and the plan, assets are in-
sufficient to discharge all benefit obligations under Title
IV of ERISA, the plan sponsor may be required to contribute
additional sums to the plan, or to the Pension Benefit Guaranty
Corporation, up to 30 percent of its net worth as computed by
the PBGC. Under proposed legislation, the 30 percent of net
..worth limitation would be removed and the plan sponsor would
have to continue contributing to the plan until all vested
benefits are fully funded. Employer participants in multi-
employer pension plans have similar obligations. A signatory
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
! 0
firm that withdraws from a multiemployer plan must assume and
ultimately fund its attributable share of the unfunded vested
liabilities of the plan.
The Financial Accounting Standards Board (FASB) is consid-
ering an accounting principle, called a Statement, that would
not only prescribe a new basis for computing annual pension
cost accruals but would require that unfunded pension liabil-
ities be shown on the plan sponsor's balance sheet. There
would be an offsetting intangible asset to avoid an initial
impact on net worth, but the intangible asset would have to be
written off over a relatively short period of years. These
proposals, which are likely to be adopted in some form, are
based upon the assumption that a defined benefit plan will
continue to function throughout the indefinite future and that
all benefits accrued as of any given time will eventually have
to be paid. There is a supporting assumption that the plan
sponsor has a legal obligation to pay all promised benefits,
the obligation running to the participants rather than to the
plan. The FASB proposes that these new accounting require-
ments apply to pension plans of state and local government
agencies.
These disadvantages of a defined benefit plan from the
sponsor's vantage point must, of course, be weighed against
the advantages of the arrangement that have nurtured and sus-
tained it over the years.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
Characteristics of Defined Contribution Pension Plans
by
Steven N. Weisbart, Ph.D.
Senior Research Officer
Teachers Insurance and Annuity Association
College Retirement Equities Fund
All pension arrangements -- defined contribution and defined ben-
efit alike -- have as their fundamental purpose the systematic accumu-
lation, during a participant's working years, of sufficient funds for
retirement income to assure that the transition from work to retire-
ment will be feasible and orderly, not only for the employee, but for
the workforce and the workplace as well. The essence of the defined
cbntributibn type of pension plan is that each plan participant has an
individual account, and the amount accumulated in that account when
the participant retires depends on the amount of contributions made to
the plan on the participant's behalf plus investment earnings these
contributions produce. Rules for determining contribution amounts for.
plan participants are defined in the plan document.
Defined contribution pension plans are sometimes confused with
thrift or profit-sharing plans. Both the conceptual and the opera-
tional differences between a defined contribution j;ensibn plan and a
thrift plan for personal savings are significant, but not always obvi-
ous. Thrift plans have little or no influence on an employer's
retirement or personnel policies. By contrast, pension plans do
affect employers' personnel policies -- staffing, turnover, career
paths, retirement ages, and of course, actually retiring. A pension
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
plan is a shared concern, with both employers as well as employees
having a stake in the provisions and end results of the plan.
We'll concern ourselves in this paper solely with defined con-
tribution pehsibh plans. We specifically exclude profit-sharing,
thrift, tax-deferred annuity, and similar plans -- not because of
their unimportance but because they don't have the specific objective
or function (orderly and feasible retirement) that pension plans do.
And for the most part, my comments about defined contribution pension
plans will be drawn from the experience of the TIAA-CRLF system, a
defined contribution pension system that has worked very successfully
for higher education for 65 years.
Statement' bf " the" Employer's. Commitment
In most defined contribution plans, the employer promises to con-
tribute a stated percentage of each participant's pay per month (or
similar period of service) into individual accounts under the plan.
As each contribution is made, the employer's obligation with respect
to accruals for the period of service is satisfied. Thus, the plan is
always fully funded. No additional employer payments are required to
make up for funding deficits caused by investment performance that
falls short of what the plan's actuary assumed. Instead, in a defined
contribution pension plan, the participants take the investment risk;
if investment results are less than had been hoped, for example, the
pension income is also. Of course, the opposite is also true -- where
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
investment results are greater than had been expected, the partic-
ipant's income is also greater.
The amount contributed under the plan is usually a percentage of
salary. A common starting point is to set the contribution percentage
so that the plan produces a retirement income for the career employee
that, when combined with Social Security, achieves a desired replace-
ment ratio (pre-tax retirement incase as a percent of pre-tax prere-
tirement earnings). Commonly, for example, the plan objective is a
replacement ratio of about 70%, with about half coming from Social
Security and half from the pension plan. If we assume that wage
increases and the pension plan's investment earnings each year match
the rate of inflation, then for a 40-year career (say, from age 25 to
65), a contribution rate of about 14% of pay would fit the plan's
income objective.*
The contribution rate can be split between the employer and the
employee.-- perhaps 10% employer and 4% employee, or 7% each for
employer and employee -- or all can be made by the employer. The rate
can be the same percentage for all employees, or it can differ, per-
haps, by years of service. However, since the Supreme Court's t?:anhart
decision in 1978, contributions may not differ by sex of the employee,
*This assumes that the accumulation is converted to income at the par-
ticipant's age 65, that pay-out is in the form of a single life annui-
ty, that the assumed investment return (AIR) included in the annuity
factor is 4%, that the annuity mortality table used is TIAA's
Merged-Gender Mod I, with ayes set back one year, which produces annu-
ity rates about halfway between the male and female rates of the 1971
IAt?: Table (1, 2.5), and that an expense charge of 1/4 of 1% of assets
is deducted from investment earnings each year.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
?
and since the Norris decision this summer, benefits derived from con-
tributions made on and after August 1, 1983 also may not differ by sex
of the recipient.
The. defined contribution approach to pensions imposes a valuable
funding discipline on the employer-sponsor. Unlike the situation with
a defined benefit plan, any failure to make full and timely contrib-
utions'to a defined contribution plan directly affects plan partic-
ipants and is readily apparent to them. In addition, the defined
contribution approach also imposes what I'll call a cost-benefit dis-
cipline. by this I mean that defined contribution plans avoid the
problem that defined benefit plans sometimes fall into, by which bene-
fit increases are promised without concanmitant fundiny increases.
Defined contribution plans don't create for one generation of partic-
ipants benefits whose funding is deferred to another generation of
participants.
Integration with Social Security
Since the overall objective of the pension plan includes benefits
to be received from Social Security, it is logical that the contrib-
ution rate of a defined contribution plan be designed to coordinate
the plan's benefits with Social Security benefits. As Dr. NcCill has
written,
To design plan benefits in recognition of Social Securi-
ty benefits is known as integration'. If only the gener-
al level of the PIA is anticipated, the benefits are
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
said to be implicitly integrated. If both the level and
shape of Social Security replacement ratios are taken
into consideration, then integration of,the benefit formula is
said to be explicit.*
Many defined contribution plans have a level, or flat, contrib-
ution rate -- an implicitly integrated design. For explicit inte-
gration, a two-tier, or step, contribution rate is often used to bring
persons of high, low, and middle salary levels to retirement with com-
bined Social Security and pension benefits that represent a more uni-
form percentage, or replacement ratio, of their final salaries than
can be achieved with a level contribution plan. In recognition of the
Social Security benefit formula's weighting toward
lower-career-earnings workers, a step-rate design calls for pension
contributions set at one rate on the portion of a person's salary up
to a dollar limit, and at a higher rate for salary over that limit.
Among typical defined contribution pension plans with step-rate pro-
visions, the rate on the lower part of salary is between 10% and
12.5%., and the rate on the rest of salary is between 15% and 18%. The,
step-rate pattern has been used primarily by employers having a range
of salary levels extending well beyond the Social Security wage base.
Previously the most commonly-used amount above which the higher
contribution rate applied was the Social Security wage base. but the
wage base has moved up so fast in recent years that if step-rate plans
had not changed, the higher contribution rate would. have applied to
but few participants and but little of their salaries, making
*Iio%jard E. Winklevoss and Dan I;cCil1, Public' Pension Plans: Standards
Of' Design; - Fi nding; ' and" Reporting , (Ilomewoo , IL: Dow-Jones Irwin,
1979), p. 57.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Is ?
step-rate plans, in effect, level rate plans. Instead, many plans now
use the current "second formula bracket point" in the Social Security
PIA formula, which increases each year as the Social Security brackets
change. For 1984 the second bracket point translates as an annual
salary of $19,344, above which the higher step, contribution rate is
applied. Therefore, a plan using the bracket point approach may call
for a 10% contribution on salary up to $19,344 and 15% on the excess.
Defined Contribution Pension
Plains' and Inflation
Earlier, in connection with the discussion of setting the plan's
overall contribution rate, I assumed that employees' wage increases and
the pension plan's investment earnings each year exactly match the rate
of inflation. At this point I'd like to discuss the implications of
this assumption, and then tackle the matter of the behavior of defined
contribution pension plans in an inflationery environment.
First, consider the assumption that the plan's investment earnings
match inflation. From the multi-decade historical studies of Ibbotson
and Sinquefield, we know that C.S. Treasury Bill yields have shown this
characteristic, with but small deviations from the CPI year by year.
So the assumption is a realistic -- and maybe even a conservative --
one. To it, let's couple the assumption that an individual's wages keep
even with inflation. This, too, is probably a conservative assumption,
since even when the general level of wages loses ground to the price
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
?
?
level, merit increases and promotions for individuals moderate that
effect over a working lifetime.
The following table shows the effect of these two assumptions on the
contributions in selected years.on behalf.of an individual covered by a
defined contribution pension plan. The table assumes a roughly 7% annual
inflation rate, but the rate itself doesn't matter; the relationships
shown hold for any interest rate.
(1)
Year
(2)
Age
(3)
Nominal
Salary
(4)
Contribution
Rate
(5)
Contribution
Amount
-(6)
Accumulated
Value at 65
1
25
$20,000
14%
$2,800
$42,000
20
45
$80,000
14%
11,200
42,000
40
65
$300,000
14%
42,000
42,000
The key column, for our purposes, is column (6). It shows that,
looking backward from the point of retirement, each year's contrib-
ution was "worth" as much toward the retirement income generated as
every other year's contribution, and -- what is perhaps more important
-- that they are all the same as the final year's contribution. It's
as though contributions were made each year Lased on the final year's
salary. The point appears simple and obvious here, but it is often
missed by those who ignore the effect of investment earnings in a
defined contribution system, and who disparage it as a "career
average" -based method of determining each participant's pension
income.
Let's finish the point by comparing the following two equations:
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
(1) 15% x "final salary" x 40 years = pension accumulation
(2) 1.5% x final salary x 40 years = pension income
The first equation describes a defined contribution plan, the second a
defined benefit plan. If $10 of pension accumulation produces $1 of
yearly pension income, then the two formulations are equivalent.
But does $10 of pension accumulation. produce $1 of pension
income? There is no simple answer to this seemingly-simple question.
The complexity lies principally in the choice of an interest rate for
the calculation that converts an accumulation into a stream of life-
time income -- what for simplicity I'll call the Assumed Interest
Rate, or AIR. The higher the AIR, the higher the income that can be
paid out for life from a given accumulation. In round figures, an
interest rate in the 8% range would produce a 10:1.ratio of accumu-
lation to income.
An actual 8% earned rate would have to be sustained for the
entire pay-out period, or else the level of income paid out would have
to drop -- an unattractive occurrence for a retiree. It would be wis-
er, perhaps, to use a lower AIR than 8% -- one that would likely be
sustainable over the lifetimes of all retirees. With a 4% AIR and
actual earnings over 4% used to provide additional income in retire-
ment, the pension stream would increase each year that total
investment earnings were over 4%. host important, it would increase
retirees' incomes without any added funds from the plan sponsor,
either at the time of the increase or at any later time. These
increases come, of course, at a cost. With a 4% AIR, to get $1 of
initial yearly income takes about $13.60. Stated differently, the
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
initial income produced by a 4% AIR would be roughly three quarters of
the initial income that could be provided by an 8% AIR from a given
accumulation.
Timing of Employee's Retirement:
"Early,"' "Nbrinal", "Late,"' or- "Phased"
Philosophically, defined contribution pension plans place more of
the decision about when to retire in the hands of each individual
employee than do most defined benefit plans. Defined contribution
plans have a "normal" retirement age, which means the age toward which
the plan funding objective is oriented. Some defined contribution
plans discontinue making contributions on behalf of active employees
who attain the "normal retirement age."
If an employee wants to retire before attaining the plan's normal
retirement age, the incase benefit payable is whatever the accumu-
lation to that point will produce when divided by the annuity factor
for the individual's attained age. Defined contribution plans don't
subsidize early retirement by applying a less-than-full actuarial
reduction, nor do they penalize employees who retire after the plan's
normal retirement age with a less-than-full actuarial boost.
6niat's more, defined contribution pension plans easily accommo-
date a variety of styles of retirement, because starting to receive
annuity income need not coincide with termination of employment from a
particular employer. Under many defined contribution plans, partic-
ipants may retire from an employer without beginning annuity income
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
until they reach their 70's -- living in the meantime on part-time-
employment, Social Security, interest income from personal savings, or
whatever other sources they may have, timing the pattern of their
income to what best suits their financial needs. Until income is
started, the accumulation continues to be credited with interest, and
of. course the later starting aye spreads the accumulation over a
shorter life expectancy, making each paynent larger than otherwise.
moreover, participants in defined contribution plans can be
allowed to "split the accumulation" into parts, starting each part on
a different. date, so that they phase into receiving their full annuity
income. The chief advantage of this multiple-starting-date feature is
the phased retirement environment it nurtures. The suddenness and
trauma of being a worker one day and a retiree the next need not
occur. Phased retirement offers a chance for employers and their
workers with declining physical or mental abilities to tailor the job
duties and pay to the diminished capacity of an experienced worker,
while part of the pension income makes up for the lower salary. Where
mandatory retirement has been abolished, voluntary phased retirement
may prove to be a good personnel policy with which to replace it.
Vesting" and' Portability
Particularly over the last decade, the importance of early vest-
ing to achieving meaningful levels of employer-pension income has
become accepted. American workers are Lecaniny more mobile, pulled Ly
the demands on one worker of a spouse's emp.loyrient, pushed by the high
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
rate of divorce, among many other forces. The problem of workers
reaching retirement with a trail of nonvested forfeitures and and/or
income benefits frozen at some inflation-depleted level has now become
widely appreciated. As a. result, vesting provisions in many plans are
being shortened, though not perhaps as short as the immediate vesting
that is characteristic of most defined contribution plans.
Vesting may be particularly important to workers whose skills are
in especially heavy demand, since they are likely to be the most
mobile. As an illustration, here's a snapshot from the pages of Busi-
he'ss Week:
When Robert B. Young, Jr. took over in 1980 as president
of Lockheed Engineering & Management Services Co. (LEM-
SCO), he wanted to cut the cost of the company's pension
plan, high by U.S. industry standards. He also needed a
recruiting tool for engineers, computer programmers, and
other technically skilled workers. Young's solution to
both problems is a new retirement plan that guarantees
immediate vesting and portability of pension contrib-
utions in place of a plan that required a worker to stay
at LEMSCO at least 10 years to earn a pension.... Now,
early vesting is helping attract and keel. employees, who
rate pensions high on the list of Lockheed benefits .
instead of at the bottom. "Even though the old plan was
very costly, it was valued low by workers because they
didn't expect to see any advantage from it,' says Ray-
mond H. Kann, a partner with Hewitt Associates, which
helped developed the plan.*
Don Grubbs among others has argued that rapid, even immediate
vesting, isn't enough, by itself, for pensions to. achieve their
intended purposes. Portability is also essential, he claims, for two
reasons. First, immediate. vesting of the contributions or benefits of
*"Pension Plans Get More Flexible," Business Wee),, I7ovemLer 8, 1982,
p. 82.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
highly mobile workers creates an. administrative burden for an
employer, requiring recordkeeping for numerous small inactive
accounts. Second, he says that when pension plans allow terminating
participants to cash out small benefit amounts, to ease the burden of
the plan's administering these small sums, experience shows that those
receiving cash spend the money rather than preserve it in a rollover
IRA. There can be no disagreement with his observation that "the
worker who changes jobs every few years has just as great a need for a
retirement income as the one who works 10 or more years for the same
employer."* And this need persists, we would add, even when the work-
er would dissipate the funds originally set aside for to meet that
need.
.he importance of immediate vesting goes beyond the realm of
workers who change employers frequently but remain continuously in the
workforce. It applies with equal, or greater, force to those who move
in and out of the workforce -- disproportionately more women and
minorities than others. Crubbs advocates a central clearinghouse to
accept and administer small pension amounts, contending that a clear-
inghouse
might make earlier vesting feasible from a cost-benefit
standpoint for additional employers. Mobile workers
would have a greater opportunity to receive retirement
income reflective of most of their years of employment
rather than just their longest job.*
*Donald S. Crubbs, Jr., "Vesting and a Federal Portable Pension
System," Jourhal"of'Pension'Plannin and'Com liance, Vol 9:5 (Oct.
1983), pp. 391-397.
*ILid.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3 .
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
A central clearinghouse, of sorts, is what TIAA-CREF' has Leer,
operating for defined contribution pension plans in higher education
for 65 years. An employee at any participating institution has con-
tributions made to a fully and immediately vested individual annuity.
When a participant moves to another institution with a TIAA-CREF
defined contribution retirement plan, the new employer's contributions
under its plan are made to the same annuities that were issued under
the plan at the previous employer. The annuity accounts of people who
move to employers that are not part of the TIAA-CREF system continue
to participate in the investment experience of TIAA and/or CRLF, so
that small accounts -- especially those established early in a work-
er's career -- grow and do not constitute an administrative burden.
Further, TIAA and CREF annuities designed for use in a defined
contribution pension plan have no cash or loan values. In all of our
communications on this point, with both participants and participating
employers, we stress that the absense of cash or loan values is neces-
sary to provide employers with the assurance they seek that the con-
tributions will be preserved for retirement purposes, and in return
for which the plan offers full and immediate vesting.
In discussing the pros and cons of his portability proposal,
Crubbs notes that implementing portability may create inequities and
undermine pension benefit security. lie writes,
for many defined benefit pension plans, the assets of
the plan are less- than the value of vested benefits. In
such a case, a transfer for one participant of assets
equal to 100% of the value of his vested benefits auto-
matically reduces the ratio of the remaining assets to
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0
9
the value of the vested benefits of the remaininy par-
ticipants. Not only could this this reduce the benefit
security of the remaining participants, in some circumstances
it could increase the potential liability of PBCC ... .*
A defined contribution pension plan needn't have this problem, and it
won't if the assets in which the individual accounts are invested are
"marked to market," since by definition the assets of the account are
equal to the value of vested benefits. In that case, funds removed
from the plan have no adverse effect on the benefit security of the
remaining participants.
Communicating with Plan Participants:
Clarity"," Frogliency', * Content
It is often said of defined contribution pension plans that they
do not lend themselves "to simple calculation or expression of bene-
fits." Of course it is true that a formula-defined benefit sounds
simple and clear, especially in contrast with a statement that your
benefit will be whatever income the accumulation provides at the time
of retirement. To forecast a defined contribution plan's benefit
requires forecasts of an individual's salary levels and interest
rates, among other elements. But an equally daunting set of assump-
tions must be made to project any particular individual's defined
benefit income amount -- including assuming that the worker stays with
the employer for the rest of his or her working career.
Most defined contribution pension plans can be designed to mail
each participant -- those on whose behalf contributions are currently
*Ibid, p. 387.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 .
being made as well as those with no current contriLutions -- a yearly
report on the amount of the annuity accumulation in his or her account
as of.year-end. The report could also project the amount of annuity
income that. would be paid at the individual's retirement age under one
or more preselected contribution assumptions and one or more 1reselec-
tedinvestment earnings, retirement age, and lifetime income option
assumptions. With appropriate software on a microcomputer, a partic-
ipant could consider the effect of a variety of assumptions on the
amount of retirement income the defined contribution account would
produce.
In TIAA-CRLF's experience, participants give these reports close
attention. From the roughly three-quarters of a million reports that
are mailed each year, some 50,000 people request additional income
illustrations. In addition, as you'd expect, we get numerous calls
and letters with a variety of questions, more or less touched of f by
receipt of this report. Many people keep successive years' reports,
comparing each new arrival to prior years' income illustrations. Whi-
le I can't claim that this is a perfect communications activity, I can
report that most 84% of those responding to a 1982 survey we took of
new TIAA-CREF retireees say that they were kept well informed about
their annuities over the years, and most of those who didn't feel this
way were people who had been in the TIAA-CREF system only a short
time.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Investment Media' and' Investment Objectives
Since in a defined contribution pension plan the participant, not
the employer, takes the investment risk, the selection of media in
which to invest the contributions constitutes, in effect, part of the
plan's design. ZJhat are appropriate investment media and investment
objectives for a defined contribution pension plan? Media that satis-
fy the following two characteristics would seem to be suitable:
o Media that are likely to provide a positive real
investment return, on average, over the long run -- or
failing that, that are likely to provide a zero real
return, on average, over the long run; and
o Nedia that have readily-obtainable market values, so
that the account values may be easily determined, and
so that if funds are withdrawn for rollover, neither
the terminating worker nor the remaining participants
subsidize the other.
A number of major studies indicate that short-.term debt invest-
ments have these characteristics, as do long-term publicly-traded
bonds and common stocks. Should defined contribution pension funds be,
invested in just one of these media, or spread among several? Should
there be several common stock funds with different risk levels, or
perhaps just one, indexed to the market as a whole? Should plan par-
ticipants be free to transfer accumulations back-and-forth among funds
as they prefer, or should transferability be limited? There is a
diversity of viewpoints on these questions.
The main issue in the question of multiple investment alterna-
tives arises from two different views of defined contribution pension
plans. On the one hand, both employers and staff members want their
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
PALL 17
pension funds committed to achieving the plan's objective of a life-
time income sufficient to make retiring financially acceptable for the
career employee by normal retirement age. Cn the other hand, some
plan participants want greater freedom to invest the funds that play
such an important role in shaping their financial security, whether or,
not this freedom brings with it risks of undermining the pension
plan's main objective for them and the employer. After all, they ask,
whose money is it, anyway?
Given the freedom to switch the funds underlying their future
pensions freely 'among an assortment of investment managers, many peo-
ple might invest the monthly contributions successfully and reach
retirement age with sufficient funds intact to retire. but those who
don't may have little choice except to stay at work as long as they
can, even though their performance may be impaired or their desire to
work diminished. Moreover, staying on the payroll is easier for them
-- and its consequences tougher on employers and the rest of the work-
force --where there is no mandatory retirement age. The freedom to
write one's own pension investment ticket may be in step with the
times, but whether it will prove to be good pension practice for the
long pull remains to be seen.
Disability and' Death Benef its
Defined contribution pension plans can easily make supplementary
death and total disability benefits available at virtually no extra
cost. In the event of the participant's death before retirement, the
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
value of the participant's account can simply be paid to a named Lene-
ficiary, either as a single sum or as a lifetime income.
In the event of total disability, two types of disability insur-
ance benefits are available. If the individual pension accounts are
funded with annuities, it is easy to buy a waiver of premium benefit
that continues contributions to the annuity as long as the partic-
ipant's total disability continues until he or she reaches the plan's
normal retirement age. Under TIAA-CREF annuities, this benefit deter-
mines contributions by reference to the contribution schedule under
the pension plan, and increases contributions by 3% per year to
account for salary increases that the participant could be presumed to
have received had he or she continued working. The other type of dis-
ability benefit is, of course, the possibility of starting to receive
income from the pension accumulation.
In - Siunmary
Defined contribution pension plans have a number of especially
attractive characteristics. Perhaps the three most outstanding are
one, the simplicity of design and administration, two, the ease of
communication of the plan's design and its benefits to participants,
and three, the security that an always-fully-funded plan offers par-
ticipants and its corollary, cost control for plan sponsors. Defined
contribution pension plans can offer participants a wide variety of
flexibilites, particularly in controlling the timing and amount of
pension income and coordinating it with other sources of retirement
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0
?
income, all at no cost to the plan. By an appropriately designed con-
tribution rate structure, they can be integrated with Social Security.
With immediate vesting provisions, they can preserve for retirement
the contributions of each employer that an individual works for,
increasing the likelihood that workers who change jobs, or who are in
and out of the workforce during their careers, will reach retirement
with an income sufficient to make retirement's challenges affordable
and attractive.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0
?
DEFINED BENEFIT/DEFINED CONTRIBUTION: A COMPARISON
by
Robert D. Krinsky, President
Martin E. Segal Company
December 13, 1983
A comparison involves an analysis of two or more items to determine
similarities and differences. Unfortunately, comparisons of defined
benefit and defined contribution plans usually focus on the differ-
ences and treat the two types of plans as competitive rather than
complementary.
In the past few years, the press and trade journals have devoted
considerable attention to the relatively small number of corporations
that have terminated defined benefit plans and in some cases substituted
a defined contribution plan. This focus implies that the two types of
plans are competitive and that a thorough evaluation will indicate which
one is best for a particular situation. In fact, however, a large and
growing number of private and public employees participate in both
defined benefit and defined contribution plans - and the plans comple-
ment each other in meeting the retirement income needs of covered
employees.
I hope that we all leave this policy forum with a "complementary
perspective". It seems to us that both the public and private sectors
should view defined benefit and defined contribution plans as comple-
mentary rather than competitive.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Prevalence of Defined Benefit Plans
Defined benefit plans predominate in the public sector. Almost all
full-time public employees are covered by retirement systems that
provide pensions based, in whole or in part, on the member's final
average salary. With the exception of public higher education
employees who participate in TIAA-CREF.plans, a very small proportion
of public employees are covered by a "basic" retirement system that
provides money purchase retirement benefits, i.e., benefits determined
solely by the actuarial conversion of an accumulated account.
Historically, a_relatively large number of public employees were
covered by combination plans (plans that used a money purchase
approach with respect to employee contributions and a final average
salary formula to determine the benefits provided from employer funds)
but most of these so-called combination plans have now been converted
to full defined benefit plans.
Among public employees covered by retirement systems, practically all
federal employees and more than nine out of ten state and local employees
who qualify for retirement benefits will have their benefits determined
on the basis of some measure of final average salary. Given the recent
trends in inflation and public employee salaries, the advantages to
covered employees of final average salary benefit formulas are obvious.
While other methods have been used to adjust the benefit structure of
retirement plans to recognize economic changes up to the point of
retirement, basing benefits on final average salary appears to be
the most systematic and equitable method of automatically protecting
the real value of benefits in relation to rising salaries. Under this
type of benefit formula the basic purpose of the retirement system -
to.replace some portion of earnings, depending on length of service,
in the event of old age, disability and death - is directly related to
existing economic conditions. Moreover, final average salary benefit
formulas reward employees who realize a steep progression in salary and
may thereby encourage long-term service and provide an incentive for
improved performance.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? ?
In addition, federal retirement systems and a majority of state and
local systems have dealt with the issue of protecting the purchasing
power of retirement benefits during the lengthening retirement period.
In general, the benefits provided by Social Security and federal
retirement systems have been fully adjusted for rising prices as
measured by the CPI; state and local retirement systems, on the other
hand, have used a wide variety of approaches to adjust benefits after
retirement and partially protect the erosion in purchasing power caused
by rising prices.
Supplemental Defined Contribution Plans
Given the history of public employee retirement systems and the preva-
lence of final average salary defined benefit plans, we suspect that
the "basic retirement benefits" for a new federal employee are likely
to be provided through a defined benefit retirement system. Your
Committee and the Congress may also want to evaluate the implications
of creating one or more "supplemental plans" which provide benefits
based on the amount accumulated in an individual's account.
Recent changes in federal law have expanded the number of supplemental
retirement plans which are or could be made available to public employees.
The expanded availability of these plans may alter how employees perceive
and plan for retirement. In.the future, we expect that the retirement
decisions of an increasing number of public employees will be influenced
not only by the level of benefits provided by the "basic" defined
benefit retirement system, but also by the dollar amount accumulated
by the employee under various "supplemental" defined-contribution-type
plans.
Supplemental plans which are or could be made available to public
employees include the following:
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
?
Section 457 plans (public employee deferred
compensation)
Section 403(b) plans (tax sheltered annuities
for school employees)
Section 401(k) plans (cash or deferred plans)
Thrift plans
Individual retirement accounts
Deductible voluntary employee contributions
Employer pick-up plans (per Section.414(h)(2))
Although the technical elements of these supplemental plans vary signifi-
cantly,,they share at least three common characteristics: (1) all are
defined contribution plans; (2) the contributions are tax deferred - at
least for federal income tax-purposes; and (3) participation in these
plans is voluntary - with the exception of most 414(h) employer pick-up
plans.
The characteristics of supplemental plans complement some emerging trends
affecting retirement, and. offer a tax-sheltered means for public employees
to save for retirement. A series of events have combined to increase the
popularity of defined contribution plans. The growth of these plans in
the private sector invites a "me too" approach in the public sector.
While many public employees prefer the few decisions required by defined
benefit retirement systems (typically limited to payment option), a grow-
ing number recognize the desirability of tailoring retirement income to
meet their own specific income and tax planning requirements. Voluntary
supplemental defined contribution plans offer this flexibility.
Our experience indicates that a growing number of public employees are
now prepared to make the investment decisions required in many supple-
mental plans. Inflation and fluctuating interest rates during the
past decade encouraged many people to forsake the passive security
of their savings accounts for the less familiar territory of money
market funds and special time deposits. Economic changes required
new investment decisions, expanded the availability of new investment
products, and resulted in a substantial increase in the investment
awareness of many employees.
- 4 -
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 ?
Supplemental defined contribution plans offer some current income tax
relief, especially to that enlarged and growing class of public employees
who are members of two wage-earner families. The assumption is, of
course, that Congress will not materially alter the tax-favored treatment
which employees now expect to enjoy when they reach retirement age and
begin to receive the "taxable" benefits of these supplemental plans.
The Social Security Amendments of 1983 provide an opportunity - and a
challenge - to develop creatively-designed retirement programs for new
federal employees. We suggest that consideration be given to the
feasibility of developing defined benefit and defined contribution
plans designed in a complementary fashion to meet specified retirement
income objectives. As a basic structure for a new program, your
Committee may want to consider the implications of establishing a
"basic" defined benefit plan and at the same time offering new federal
employees the opportunity to participate in one or more "supplemental"
defined contribution plans.
Comparison of Major Features
In view of the topic assigned to me, I feel compelled to make some brief,
comments on the major features of defined contribution and defined bene-
fit plans.
The EBRI volume - Economic Survival in Retirement: Which Pension Is
For You? (1982) - comprises articles presented at a May 20, 1982 policy
forum on understanding the differences between defined benefit and
defined contribution plans. The "Forward" by Dallas Salisbury. suggests
that the relative attractiveness of defined contribution and defined
benefit plans is changing, but that the recent emphasis on defined
contribution plans has not been accompanied by a thorough examination
of the potential effects of such plans on future retirement income
security.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
?
In his "Introduction" to the EBRI volume, my colleague Robert Paul
points out that:
"Today, workers who are entering the labor force for the
first time are exposed to a different type of environment.
Private defined benefit plans and defined contribution
plans are now taken for granted. Past service is not as
important in the minds of today's young workers, because
they have a full career in which to earn a retirement
benefit. Thus, people are asking themselves: Which
pension is the better alternative? Which retirement
income components offer an effective blend? Do we need
both defined benefit and defined contribution plans?
Moreover, there are more and more two-worker families.
Both members of such families may consider individual
savings opportunities as well as the questions surround-
ing the options of defined contribution and/or defined
benefit plans..
Finally, there is the general question of whose respon-
sibility it is to provide retirement income. Is it the
government's responsibility? Is it the private employer's
responsibility? Is it the individual's responsibility?
Or is it the responsibility of all three parties?
There is increasing emphasis today on the notion that
it is up to individuals to provide a greater portion
of their own retirement income security. This is also
contributing to our reexamination of the issues
surrounding the question of Economic Survival in
Retirement: Which Pension Is for You?"
To a large degree, Bob Paul's introductory comments also apply to new
federal employees. As compared with almost all current federal employees,
they will definitely be exposed to a different retirement program
structure. Because new federal employees will be covered under Social
Security, the extent to which the various types of Social Security
benefits meet the economic needs of a typical new federal employee
needs to be recognized - and the structure of the new retirement program
needs to take account of Social Security coverage.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 0
You are all familiar with the major differences between defined benefit
and defined contribution plans. (See the EBRI volume for a discussion
of differences from various perspectives: a labor perspective, a
corporate perspective, and a participant's perspective.)
The difference in employer commitment is basic: the employer undertakes
to provide a specified level of retirement income in the defined benefit
plan, and to make a specified contribution to individually-allocated
investment accounts in a defined contribution plan. In terms of
individual equity, the principle underlying a defined benefit plan
is one of "equal benefits" - whereas a defined contribution plan
generally defines individual equity in terms of "equal contributions"
and accepts the necessarily unequal benefits that result from equal
employer contributions.
A defined contribution plan that is strictly a money-purchase arrangement
is necessarily prospective. Benefits derive only from contributions and
investment income after the plan begins. The employee's length of service
or salary history prior to the creation of the plan is ignored. A defined
benefit plan, on the other hand, is usually retrospective as well as
prospective. In the early years of the plan a significant part of the
employer contribution is usually allocated to financing benefits based
on service rendered before the plan became effective.
Another major difference between the plans - one that is frequently
emphasized - is the party who bears the risk of future investment perform-
ance. A defined benefit plan deals directly with the level of benefits to
be provided for covered employees, whereas the benefits produced by
defined contribution plans vary depending on future investment performance
(and possibly certain other factors such as terminations and future profits).
Because investment risk is borne by the employee in a defined contribution
plan, some people feel that employees "own" the assets in such a plan to a
greater degree than in a defined benefit plan. However, if pensions are
perceived essentially as a form of deferred compensation, differences
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
0 0
between the plans can be considered primarily in terms of individual
equity in relation to a given employer contribution - rather than in
terms of who "owns" the assets. We understand that this broad "ownership
issue" (which encompasses the issues of benefit security and benefit
guarantees) is important to certain federal employee unions, because
they believe that a defined contribution plan can be structured in a
manner that results in "ownership" and control of the assets by covered
employees. On the other hand, Congress apparently retains the authority
to amend the provisions of defined benefit plans covering most federal
employees.
In considering the design of retirement programs for new federal employees,
your Committee will undoubtedly want to evaluate this "ownership-benefit
guarantee issue", and consider the differences, if any, in Congress'
authority to modify the provisions of defined benefit or defined contribu-
tion retirement programs for new federal employees.
Allocation of Contributions
Contributions required to finance the benefits promised by a defined
benefit plan can vary. Variability depends on numerous factors, includ-
ing changes in the characteristics of participants, the investment return
on.plan assets, and changes in the benefit design itself. For most
defined benefit plans in the public sector, contributions are determined
on the basis of a specific actuarial funding method which seeks to smooth
contribution requirements as a percentage of payroll over a long period.
Contributions are determined on an aggregate basis for all participants
rather than on an individual basis. This last point is sometimes over-
looked.by those who do not realize that employer contributions to a
defined benefit plan may have little relationship to the value of benefits
earned by a specific employee during that year. Rather, the employer
contributions reflect the overall funding requirements of the plan and
are not allocated to individual accounts.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
A substantial majority of public employee retirement systems are con-
tributory defined benefit plans. Such plans maintain individual accounts
for each participant and generally provide that the amount in the
employee's contribution account will be refunded upon termination of
service. Some employees covered by contributory defined benefit plans
incorrectly perceive the employer contribution as also "belonging" to'
the participant, and this gives rise to some of the confusion surround-
ing the "ownership issue". A participant who terminates before vesting
and elects to receive a refund of his contributions to a defined benefit
plan does not forfeit employer contributions which "belong" to him.
Employer contributions to a defined benefit plan are not allocated to
individual employee accounts, but rather reflect the amount needed, in
aggregate, to fund the benefits promised in accordance with the established
funding method.
The contribution-oriented character of a defined contribution plan tends
to direct attention to the level of the contribution and the dollar amount
accumulated in an individual's account. Unfortunately, this focus on the
total dollar amount of the account is often not accompanied by even a
very rough notion of the level of, lifetime pension benefits that can be
provided by the accumulated amount. Moreover, individuals participating
in defined contribution plans may not realize that the amount of retire-
ment income produced'by the plan depends on market conditions prevailing
at retirement; identical account balances will produce monthly annuities
that vary considerably depending on. interest and annuity rates in effect
at the time of retirement.
Summary of Differences
A schedule summarizing differences between defined benefit and defined
contribution plans is attached. Some brief comments on a few of the
differences follow:
Early vesting. Defined contribution plans generally
have very early vesting. This feature is attractive
to new employees who do not expect to remain in
service long enough to vest. under a defined benefit
- 9 -
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
i ?
plan. The "longer-staying" employees - including
those new employees who do not expect to remain in
covered employment but actually do! - may see it
differently: (a) early vesting uses some of the
financial resources of the plan for a short-term
employee, and (b) it requires the plan to maintain
many relatively small individual accounts for a
long period of time.
Early distribution and loan provisions. A defined
contribution plan usually provides for payouts from
the employee's account in the event of termination
of covered employment (even if termination occurs
years before actual retirement), and it may also
have provisions for emergency loans. Both of these
features are attractive for some employees.
Disability pensions and death benefits. A defined
benefit plan makes whatever provision it considers
desirable for these contingencies. Under a defined
contribution plan, on the other hand, all that is
immediately available for disability pensions and
death benefits is the balance in an individual's
account - and in the early years of the plan that
balance will not be substantial. While this problem
can be solved in a defined contribution plan, it
requires separating a piece of the contribution and
using it to buy disability insurance or death bene-
fits, or both.
Post-retirement benefit adjustments. As noted,
public employee retirement 'systems have used a
variety of methods to adjust benefits after-retire-
ment. The ability of a defined benefit plan to
partially compensate retirees for the erosion in
purchasing power caused.by rising prices is largely
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
a function of the employer's willingness to contribute
more to the plan. Under the defined contribution
rationale, however, all employer contributions for a
given employee have already been made prior to retire-
ment. When viewed in this perspective, defined contri-
bution plans are more limited in coping with the
post-retirement adjustment problem than defined benefit
plans. In another sense, however, the development of
variable annuities and automatic "investment-sharing
provisions" indicate that defined contribution plans
can incorporate mechanisms providing for benefit
adjustments related to investment performance during
the retirement period.
Concerns Regarding Defined Contribution Plans
A comprehensive public policy regarding defined contribution plans
apparently has not been formulated. However, recent federal legislation
has had the effect of encouraging the development of such plans. A
variety of tax provisions enacted in recent years have tended to "favor"
defined contribution plans, and others have had the effect of impeding
the development of defined benefit plans.
Several individuals and organizations have expressed concerns regarding
the growing popularity of defined contribution plans. Our concerns
include such issues as lump sum payouts, plan costs, cost/benefit
comparisons, and investment results.
Lump sum payouts. As noted, defined contribution plans usually provide
for lump sum distributions (which now receive more favorable tax treat-
ment). If an employee's "basic" retirement benefits are to be provided
through a defined contribution plan, should public policy favor distri-
butions in the form of lifetime pension payments rather than lump sum
distributions? How will the increasing utilization of lump sum payouts
affect future retirement benefit levels?
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
? ?
Plan costs. For a given contribution, a defined benefit plan can
generally provide more in the way of retirement benefits than a defined
contribution plan. This results from the greater flexibility in financ-
ing defined benefit plans (which do not have to be fully funded), the
relatively larger forfeitures in such plans, and the ability to take
forfeitures into account in advance in a defined benefit plan. These
cost considerations are often overlooked by proponents of defined
contribution plans, who tend to emphasize only that certain categories
of employees are likely to receive larger retirement benefits from a
defined contribution plan than from a defined benefit plan.
Cost/benefit comparisons. Studies comparing the costs and benefits of
the two types of plans may be misleading (sometimes inadvertently and
sometimes deliberately). In some instances comparisons are made on the
basis of investment return assumptions that vary substantially - as if
it is reasonable to assume that the assets of the defined benefit plan
will earn an average return of 7% or 8%, but that the assets of the
defined contribution plan will produce a much higher rate of return.
In addition, some cost/benefit comparisons totally ignore future salary
increases and the differences in projected replacement ratios resulting
from changes in assumed rates of investment return and salary increases.
The table which follows shows the replacement ratios resulting from a
1%-of-salary annual contribution based on various combinations of in-
vestment return rates and salary increases. To take one example, consider
an employee who enters service at age 35 and retires at 65 under a defined
contribution plan with a 5%-of-pay annual contribution: the replacement
rate for this employee will be 18.5% of final salary if the investment
return/salary increase combination is 7%/6%, and 24.0% based on a 9%/7%
combination. (In this example the replacement rate produced by a 9%/7%
combination is almost 30% higher than the replacement rate produced by
a 7%/6% combination!) In our opinion, meaningful cost/benefit comparisons
should include full consideration of the replacement ratios that will be
produced by various economic environments.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
Ratios of (a) retirement income produced by a 1%-of-salary annual
contribution,
to
(b) final salary during the year prior to retirement at
e 65__ _
R
t
f i
t
t
-
Entry Age
es o
men
a
nves
return/salary increases
25
30
35
40
45
50
55
60
7/5
6.3%
5.2%
4.3%
3.4%
2.6%
1.8%
1.2%.
0.6%
7/5/
5.7
4.8
4.0
3.2
2.4
1.8
1.1
0.5
7/6
5.2
4.4
3.7
3.0
2.3
1.7
1.1
0.5
8/6
6.7
5.6
4.5
3.6
2.7
2.0
1.2
0.6 ?
8/6/
6.1
5.1
4.2
3.4
2.6
1.9
1.2
0.6
8/7
5.5
4.7
3.9
3.2
2.5
1.8
1.2
0.6
9/7
7.1
5.9
4.8
3.8
2.9
2.1
1.3
0.6
9/7'
6.5
5.4
4.5
3.6
2.8
2.0
1.3
0.6
9/8
5.9
5.0
4.2
3.4
2.7
1.9
1.3
0.6
Notes:
(1)
Annuity conversion based on 1971 Group Annuity Mortality Table.
(2)
To convert above ratios to percent of final average salary (rather
than percent of final salary), multiply by the following:
Assumed Annual
Salary Increase
3-Year Average
5-Year Average
0
5
1.049
1.100
5'
1.054
1.110
6
1.059
1.120
6!
1.064
1.130
7
1.068
1.140
7!
1.073
.1.150
8
1.078
1.160
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
4P 4P
Investment results. The increasing emphasis on individual responsibility
for retirement savings - and the trend toward giving individuals more
flexibility in allocating retirement savings to various investment media
- raise broad questions regarding the adequacy of future retirement
benefits. A recent examination of IRA investment patterns (see March
1983 EBRI Issue Brief, "Individual Savings For Retirement: A Closer
Look") indicates that only about 17% of all IRA assets were in accounts
that achieved an 8.7% to 9% annual real rate of return over the 5 years
ended December, 1982. The average real rate of return for the remaining
83% of IRA holders was.less than 1% over this 5-year period. Based on
its analysis of IRA investment returns, EBRI concluded that: "The
choices made by IRA investors and the real rates of return they have
achieved to date do not support the arguments or assumptions of those
advocating proposals for greater reliance on IRAs." We agree that pro-
posals placing greater emphasis on individual responsibility for retire-
ment savings need to be thoroughly evaluated.
Defined benefit and defined contribution plans offer different approaches
to retirement savings. Neither is perfect. Consequently, we expect that
an increasing number of employees will be covered by both types of plans.
In our opinion the two types of plans should be viewed as complementary
rather than competitive. Both defined benefit and defined contribution
plans should continue to have a role in our aggregate retirement income
system.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3
 Approved For Release 2008/08/21 : CIA-RDP86B00338R000400510001-3
?Defined Benefit Plan and
Defined Contribution Plan
Feature
Benefit for
future service
Benefit for
past service
Increase amendment
for prior service
Disability pension
Death benefits
Vesting
Early retirement
Pension payments
Early cash-outs
on termination
Loan from the plan
Investment experience
Investment options
9
Defined Benefit Defined Contribution
Defined formula, What the accumulation in
considered supportable the employee's account
by overall financing of can buy or support.
the plan.
Formula defined by the None.
plan.
Defined formula. ? Not possible.
Account balance, unless
part of contribution is
used to buy disability
insurance.
10 years covered service,
or less, as fixed by plan.
As provided by plan.
Life annuity or joint-
and-survivor or other
options offered by plan.
Possible, but not simple
to administer.
Possible, but not simple
to administer.
Gains and losses - for plan
as a whole, benefits not
directly affected.
No options offered to
employee.
Maintenance of "regular
data base" - periodic
actuarial valuations,
at least once every
three years.
Account balance, unless
part of contribution is
used to buy life insurance.
Early vesting, such as
immediate or 1-5 years
covered service.
What the employee's
accumulation can buy
or support.
Lump sum or similar range
of options offered by plan.
Possible and relatively
simple.
Relatively simple.
Gains and losses - individual
employee accounts affected
directly.
Options usually offered to
employee.
Maintenance of individual
accounts - periodic valuation
of account balances, usually
several times a year.
Approved For Release 2008/08/21: CIA-RDP86B00338R000400510001-3