Social Security: Individual Accounts as an Element of Long-Term Financing
Reform (Testimony, 03/16/99, GAO/T-HEHS-99-86).

Pursuant to a congressional request, GAO discussed how best to ensure
the long-term viability of the nation's social security program.

GAO noted that: (1) social security forms the foundation of the nation's
retirement income structure, and in so doing, provides critical benefits
to millions of Americans; (2) yet, problems facing this program pose
significant policy challenges that need to be addressed soon in order to
lessen the need for more dramatic reforms in the future and to
demonstrate the federal government's ability to deal with a known major
problem before it reaches crisis proportions; (3) some social security
proposals include adding individual accounts similar to defined
contribution plans, to the current defined benefit program; (4) these
individual accounts offer the potential for increased investment returns
but they cannot by themselves restore social security's solvency without
additional changes to the current system; (5) in assessing the
proposals, policymakers must consider the extent to which the proposals
offer sustainable financing for the system; (6) also, they must consider
how to balance improvements in individual equity while maintaining
adequacy of retirement income for those individuals who rely on social
security as their primary or sole source of income; and (7) choosing
whether to incorporate individual accounts into the social security
system will require careful consideration of a number of design and
implementation issues if such a system is to function effectively at a
reasonable cost.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  T-HEHS-99-86
     TITLE:  Social Security: Individual Accounts as an Element of 
             Long-Term Financing Reform
      DATE:  03/16/99
   SUBJECT:  Social security benefits
             Economic analysis
             Future budget projections
             Financial management
             Retirement benefits
             Retirement pensions
             Investment planning
             Federal social security programs
IDENTIFIER:  Social Security Program
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Before the Committee on Finance, U.S.  Senate

For Release on Delivery
Expected at 10:00 a.m.
Tuesday, March 16, 1999

SOCIAL SECURITY - INDIVIDUAL
ACCOUNTS AS AN ELEMENT OF
LONG-TERM FINANCING REFORM

Statement of David M.  Walker
Comptroller General of the United States

GAO/T-HEHS-99-86

GAO/HEHS-99-86T


(207060)


Abbreviations
=============================================================== ABBREV


SOCIAL SECURITY:  INDIVIDUAL
ACCOUNTS AS AN ELEMENT OF
LONG-TERM FINANCING REFORM
============================================================ Chapter 0

Mr.  Chairman and Members of the Committee: 

Thank you for inviting me here today to continue the ongoing
discussion on how best to ensure the long-term viability of our
nation's Social Security program.  Demographic trends threaten the
program's solvency such that assets could be depleted by 2032. 
Numerous proposals to restore the Social Security program's solvency
have been put forth; as one element of reform, many of these include
individual accounts which could provide greater individual choice in
retirement investment and increased rates of return. 

In my remarks today, I will discuss several different approaches to
restoring the Social Security program's solvency and sustainability
and the various factors that must be considered in determining
whether individual accounts should play a role as an element of
Social Security reform.  My comments are based on several recent GAO
reports and testimonies, as well as our ongoing work.\1

In summary, Social Security forms the foundation for our retirement
income structure and, in so doing, provides critical benefits to
millions of Americans.  Yet, problems facing this program pose
significant policy challenges that we need to address soon in order
to lessen the need for more dramatic reforms in the future and to
demonstrate the federal government's ability to deal with a known
major problem before it reaches crisis proportions.  Some proposals
suggest adding individual accounts--which are similar to defined
contribution plans--to the current defined benefit program.  These
individual accounts offer the potential for increased investment
returns, but they cannot by themselves restore Social Security's
solvency without additional changes to the current system.  In
assessing these proposals, policymakers must consider the extent to
which the proposals offer sustainable financing for the system. 
Also, they must consider how to balance improvements in individual
equity (i.e., rates of return on individual contributions) while
maintaining adequacy (i.e., benefit levels, certainty) of retirement
income for those individuals who rely on Social Security as their
primary or sole source of income.  And finally, choosing whether to
incorporate individual accounts into our Social Security system will
require careful consideration of a number of design and
implementation issues to determine if such a system would function
effectively at a reasonable cost. 


--------------------
\1 In particular, see Social Security:  Different Approaches for
Addressing Program Solvency (GAO/HEHS-98-33, July 22, 1998). 


   MANY PROPOSALS TO RESTORE
   SOLVENCY INCLUDE INDIVIDUAL
   ACCOUNTS
---------------------------------------------------------- Chapter 0:1

A wide array of reform proposals have introduced the concept of
personal or individual retirement accounts into the debate over
Social Security's future solvency.  In evaluating these proposals we
must understand

  -- Social Security's fundamental role in ensuring the income
     security of our nation's elderly;

  -- the nature, extent, and timing of Social Security's financing
     problem; and

  -- the differences between the current program and a program that
     might include individual accounts. 


      SOCIAL SECURITY IS THE
      FOUNDATION OF OUR NATION'S
      RETIREMENT INCOME SYSTEM
-------------------------------------------------------- Chapter 0:1.1

Social Security\2 has long served as the foundation of our nation's
retirement income system.  That system has traditionally comprised
three parts:  Social Security, employer-sponsored pensions (both
public and private), and personal savings in the form of real and
financial assets.\3 Social Security is viewed as providing a floor of
income protection that the voluntary forms of employer pensions and
individual savings should build upon to provide a secure retirement. 
However, private pension plans cover only about 50 percent of the
full-time work force, and a significant portion of the American
public does not have any other significant personal savings.  In
addition, Social Security is the sole source of retirement income for
almost a fifth of its beneficiaries.  Given Social Security's
importance as the foundation of retirement income security, it has
been a major contributor to the dramatic reduction in poverty among
the elderly population.  Since 1959, poverty rates for the elderly
have fallen from nearly 35 percent to 10.5 percent.  (See fig.  1.)

   Figure 1:  Poverty Rates for
   the Elderly Have Fallen Since
   1959

   (See figure in printed
   edition.)

Social Security represents a retirement income insurance program that
helps workers collectively pool the risks associated with loss of
earnings due to old age, disability, and death.  It is a mandatory
and almost universal program.  As a result, the vast majority of
American workers take Social Security credits with them when they
change jobs.  Social Security also provides inflation-protected
benefits for the life of the retiree.  No matter how long they live,
under the current program design retirees continue to receive Social
Security benefits uneroded by inflation.  The program, which provides
benefits not generally available as a package in the private market,
includes benefits for retired workers, their spouses and dependents,
and their survivors as well as for the disabled. 


--------------------
\2 Social Security refers here to the Old-Age, Survivors, and
Disability Insurance program, also referred to as OASDI. 

\3 For a discussion of this traditional approach to retirement
income, see Retirement Income:  Implications of Demographic Trends
for Social Security and Pension Reform (GAO/HEHS-97-81, July 11,
1997). 


      THE FINANCING PROBLEM NEEDS
      TO BE ADDRESSED NOW
-------------------------------------------------------- Chapter 0:1.2

The Social Security system has required changes in the past to ensure
future solvency.  Indeed, the Congress has always taken the actions
necessary to do this when faced with an immediate solvency crisis. 
However, the program faces demographic conditions that require action
now to avoid unfairly burdening future generations with the program's
rising costs and to give these individuals time to make the necessary
adjustments to their retirement planning.  Social Security's
financial condition is directly affected by the relative size of the
populations of covered workers and beneficiaries.  Historically, this
relationship has been favorable.  Now, however, the covered
worker-to-retiree ratio and other demographic factors, such as life
expectancy, have changed in ways that threaten the financial solvency
and sustainability of this important national program (see fig.  2). 

   Figure 2:  Ratio of Workers to
   Beneficiaries Is Declining

   (See figure in printed
   edition.)

Thus, although the program was put in 75-year actuarial balance just
15 years ago, the Trust Fund balances now are projected to be
exhausted in 2032 (as estimated in the 1998 Trustees' Report).  In
addition, the program will begin to experience a negative cash flow
in 2013, which will accelerate with the passage of time.  Absent
meaningful program reform, this will place increased pressure on the
federal budget to raise the resources necessary to meet the program's
ongoing costs.\4 To restore solvency to the program today, we would
need to immediately increase annual program revenues by 16 percent or
reduce annual benefit payments by 14 percent across the board. 

Even if such actions were taken today, attention would need to be
given to their sustainability.  We measure solvency in this program
over a 75-year period.  As each year passes, because the system is in
temporary surplus, a year of surplus is dropped from the calculation
and a year of deficit is added into the 75-year average.  Hence,
changes made today that restore solvency only for the 75-year period
will result in future actuarial imbalances nearly immediately.  For
this reason, we must consider what is needed to put the program on a
path toward sustainable solvency so we will not face these difficult
questions on a recurring basis. 

Another way to understand the magnitude of the problem is to consider
what the system will cost as a percentage of taxable payroll in the
future.  If we did nothing and let the Trust Funds run out in 2032,
resources equivalent to 18 percent of taxable payroll would be needed
simply to finance the system in the following yearmore than 37
percent higher than the revenues projected to be available under the
12.4 percent payroll tax that currently finances the system (see fig. 
3). 

   Figure 3:  Changes Are Needed
   to Maintain Solvency

   (See figure in printed
   edition.)

Note:  Percentage changes are necessary to maintain solvency for the
next year only. 

By 2075, the end of the Trustees' current long-term projection
period, resources equivalent to nearly 20 percent of payroll--or a
48-percent increase in projected revenues--will be necessary. 
Alternatively, if we were to address these gaps through benefit
reductions, changes equal to 27 percent of benefits in 2032 and 32
percent in 2075 would be required.  Clearly, these dates are far off
and projections are fallible.  For example, stronger economic growth
than currently projected would make it possible to meet the program's
commitments more easily.  Health advances that extend life expectancy
beyond current expectations, and other variables, however, could make
things worse.  In addition, these revenue or benefit changes relate
only to one year's financing gap.  The percentages would have to be
considerably higher to make the program solvent and sustainable over
an ensuing 75-year period. 

If we do not take measures to recognize and address this entire
financing gap, we will have to revisit this difficult debate time and
time again.  The program's future financial situation calls upon us
to make prudent judgments today that will affect those in the future
who will be asked to meet these benefit commitments.  Importantly,
since we can anticipate this situation, and because our economy is
strong, we can act now to avoid more painful decisions in the future. 


--------------------
\4 Social Security:  What the President's Proposal Does and Does Not
Do (GAO/T-AIMD/HEHS-99-76, Feb.  9, 1999). 


      THERE ARE MANY DIFFERENCES
      BETWEEN THE CURRENT PROGRAM
      AND ONE THAT INCLUDES
      INDIVIDUAL ACCOUNTS
-------------------------------------------------------- Chapter 0:1.3

A wide spectrum of Social Security reform proposals has surfaced in
this debate, and they reflect different perspectives and opinions
about how best to address the program's financing problem.  Let me
describe briefly the two main perspectives on the appropriate benefit
structure for Social Security, which are analogous to the distinction
between defined benefit and defined contribution pension plans. 

The current Social Security system's benefit structure is designed to
address the twin goals of individual equity and retirement income
adequacy.  Individual equity means that there should be some
relationship between contributions made and benefits received (i.e.,
rates of return on individual contributions).  Retirement income
adequacy is addressed by providing proportionately larger benefits
(redistributive transfers) to lower earners and certain household
types, such as those with dependents (i.e., benefit levels and
certainty).  The current benefit structure combines these twin
goals--and the range of benefits Social Security provides--within a
single defined benefit formula.  Under this defined benefit program,
workers' retirement benefits are based on the lifetime record of
earnings, not directly on the payroll tax contributed.  Given the
current design of the Social Security program and known demographic
trends, the rate of return individuals will receive on their
contributions is declining.  In addition, as noted previously,
current promised benefits are not adequately funded over the 75-year
projection period. 

Alternatively, those who propose individual accounts as part of the
financing solution emphasize the potential benefits of a defined
contribution structure as an element of the Social Security program
and/or financing reform.  This approach to Social Security focuses on
directly linking workers' contributions to the retirement benefits
they will receive.  Workers' contributions are invested in financial
assets and earn market returns, and the accumulations in these
accounts then provide income in retirement.  The advantage of this
approach is that individual workers have more control over their
accounts and more choice in how the accounts are invested.  This
control enables individuals to earn a higher rate of return on their
contributions than under current law.  Of course, these opportunities
for higher returns exist because investors assume some measure of
risk that the return expected may not actually be realized. 

To illustrate the differences between the current Social Security
defined benefit structure and a primarily defined contribution
structure, we recently studied the experience of three counties in
Texas that withdrew from the Social Security system in 1981 and
substituted a defined contribution plan for Social Security.\5 The
Texas plans offer retirement, survivors, and disability benefits. 
Although contributions are somewhat higher than those of Social
Security, they are roughly comparable when Social Security's
financing gap is considered.  Benefits are based on contributions and
earnings from investments.  Under the Texas plans, contributions are
invested conservatively in fixed income securities that are readily
marketable.  We simulated the benefits that typical workers could
receive under these plans and compared them with what would have been
received under Social Security.  We found that for higher income
workers the Texas plans provided higher benefits, especially
initially.  However, because of the Social Security benefit formula
"tilt" toward lower earners, many such workers could have done better
under Social Security.  Other features of Social Security, such as
adjustments for inflation, also suggest that many median-wage workers
might have done at least as well, if not better, had they stayed
under Social Security.  However, the Texas plans followed a
relatively conservative investment strategy with lower returns than
are usually assumed in most individual account proposals. 
Nonetheless, our analysis does suggest we need to be careful that
those most reliant on Social Security are adequately protected. 

Some reform proposals incorporating individual accounts address the
need for such protection by combining defined contribution and
defined benefit approaches into a "two-tiered" structure for Social
Security.  Under such a structure, individuals would receive a base
defined benefit amount with a progressive benefit formula and a
supplemental defined contribution account benefit.  Individuals could
be guaranteed a minimum monthly benefit.  This approach, however,
raises a number of risks and administrative issues which I will
discuss later in this statement. 


--------------------
\5 Social Security Reform:  Experience of the Alternate Plans in
Texas (GAO/HEHS-99-31, Feb.  26, 1999). 


   FINANCING SUSTAINABLE SOLVENCY
---------------------------------------------------------- Chapter 0:2

Financing a sustainable solution relates to how we bring long-range
program costs and revenues into balance.  Addressing the current
projected financing imbalance requires either raising revenues or
decreasing program costs.  Funding future benefit commitments in
light of changing demographics through higher investment returns can
help make the needed measures less severe, and this is one of the
reasons many reform proposals include individual accounts. 

Still, creating individual accounts does not by itself address the
solvency problem.  Although individual accounts offer the potential
to capture higher investment returns, if the accounts are adopted
without the higher returns being shared within the system or without
accompanying benefit reductions, the solvency problem will not be
alleviated. 

The extent to which individual accounts affect long-term solvency
depends in part upon whether the accounts are "added on" to the
existing system or carved out of it.  Some proposals add on
individual accounts as a type of supplementary defined contribution
tier.  This approach effectively leaves the entire 12.4 percent
payroll tax contribution available to finance the program while
dedicating additional revenues for individual accounts.  These
additional revenues might come from a payroll tax increase or from
future unified budget surpluses.  However, this approach does nothing
to help Social Security unless incremental investment income is used
to either supplement Social Security revenues or offset current
promised benefits. 

The carve out approach involves creating and funding individual
accounts with a portion of the existing payroll tax rate.  Thus, from
the current combined payroll tax rate of 12.4 percent, a portion
could be carved out and allocated to individual accounts.  The
obvious effect is that less revenue is available to finance the
current benefit structure, so the system's solvency is further
eroded. 

Thus, individual accounts represent a way of using higher rates of
return to raise more revenues in the future than does the existing
Social Security program.  At the same time, including such accounts
as an element of reform requires that we consider ways to share the
increased returns with Social Security or revise the existing defined
benefit structure for future beneficiaries.  In other words, to
improve Social Security solvency, individual accounts and Social
Security reform must be considered together. 

In addition, finding the appropriate balance between the defined
contribution and defined benefit approaches also has implications for
the near-term financing of the Social Security program and its
payments to current retirees and those in the near future.  If
individual accounts reduce existing program revenues to finance
higher returns over the long term, we must still be able to continue
to finance ongoing benefits to retirees in the short term.  This
problem of "transition costs" means that we may have to devote
additional resources to the program in the near term.  The trade-off
is that in the long run individual accounts may, if structured
properly, help finance the program in a more sustainable way. 


   BALANCING EQUITY AND ADEQUACY
   IN THE BENEFIT STRUCTURE
---------------------------------------------------------- Chapter 0:3

Because individual accounts cannot contribute to restoring solvency
without combining with Social Security in some way, it is useful to
focus on the implications of individual accounts for Social
Security's defined benefit program.  The existing program includes a
mix of benefits covering disability, spouses and dependents, and
survivors.  It also includes transfers to lower earners and families. 
Some proposals that include individual accounts have been criticized
for not fully considering these other benefits when touting the
advantages of higher returns on defined contribution accounts.  But
most proposals address the defined benefit portion by making a number
of changes and adjustments to the existing program, and some
proposals incorporate a guarantee of current law benefits.  I will
discuss some elements of these proposals briefly and also address the
issue of whether to make the individual accounts mandatory or
voluntary. 

Decisions about the appropriate balance between the defined
contribution and defined benefit portions will need to consider the
purpose of the original Social Security program.  The altered defined
benefit portion will still be relied upon to provide a foundation
that ensures an adequate and certain retirement income level. 
Existing proposals attempt to revise this part of the program in a
variety of ways, including revising the benefit formula (usually to
make it more progressive), changing features of the program (such as
lowering the cost-of-living adjustment), raising the age of
eligibility for normal and early retirement, or revising ancillary
benefits (such as those for spouses).  Most of these proposed changes
are structured so as to leave current and near-term retirees
unaffected.  In addition, many would include an individual account
element only for workers under a stated age, often around 50. 

There are also ways to determine offsets to the individual accounts
that would raise revenue for the defined benefit program.  For
example, Social Security could reclaim a portion of the individual
account accumulations.  This reclaiming, or so-called "claw-back,"
could raise significant "expectation gap" issues with individuals. 
These expectation gaps might be addressed by pooling the investment
accounts and other measures. 

Another feature of some proposals involves a guarantee of a certain
benefit level.  This guarantee could be provided in tandem with other
benefit structure changes such that the worker would be guaranteed a
minimum benefit.  One approach would guarantee the current defined
benefit.  If the individual account provided less than the current
benefit, then the system would ensure that benefits were provided to
fill the gap.  Such an arrangement might be desirable from a benefit
adequacy perspective but would require safeguards against the
government becoming an insurer of excessive risk-taking by
individuals. 

Clearly, the number of proposals and features make it difficult to
sort out exactly what should be done.  We need to study carefully
what impacts any given proposal would have, not only on the overall
cost of the system but also, very importantly, on individuals and
families. 

One basic feature in this regard concerns whether to make investment
in individual accounts mandatory or voluntary.  Insofar as individual
accounts are intended to substitute for a portion of benefits
provided under current law to make it easier to finance the program,
most discussion has involved accounts that are mandatory.  This is
consistent with the stated goal of Social Security to ensure a
measure of income protection in old age. 

The notion of making the accounts voluntary has entered the debate
through proposals that seek to maintain the existing benefit
structure of the program.  A voluntary account is an add-on approach
that would supplement Social Security benefits and provide a measure
of individual choice.  But under such an approach the overall
implications for retirement income would be uncertain.  If the
voluntary account was supplementary, then it might be difficult to
determine whether a voluntary account added to total retirement
income; it might merely substitute for other forms of saving. 

Another potential result of creating a system of individual accounts
would be the development of an infrastructure that would allow
workers to build up additional savings to meet both income and health
care cost needs in retirement.  For example, workers not covered by a
private pension could choose to contribute more to their individual
accounts to augment their retirement savings.  Workers could also
contribute more to their accounts as part of any possible premium
support plan to help pay health care costs after they retire.  The
accounts could thereby contribute to overall retirement security, not
just retirement income security. 


   OPTIONS ARE AVAILABLE FOR
   INDIVIDUAL ACCOUNT DESIGN AND
   ADMINISTRATION
---------------------------------------------------------- Chapter 0:4

Not all proposals for individual accounts clearly delineate how these
accounts would be administered, but those that do vary in three key
areas:  (1) who would manage the information and money flow needed to
maintain a system of individual accounts, (2) how much choice and
flexibility would individuals have over investment options and access
to their accounts, and (3) what mechanisms would be used to pay out
benefits upon retirement.  Decisions in these areas would have a
direct effect on system complexity and who would bear the costs and
additional responsibilities of an individual account system as well
as on the adequacy and certainty of retirement income for future
retirees.  Essentially, most of the decisions about the design of a
system of individual accounts amount to trade-offs between individual
choice and flexibility on the one hand and simplicity and
standardization on the other.  A full assessment of the implications
of these trade-offs will be essential to the debate on whether and
how to implement individual accounts.  Table 1 summarizes some of the
administrative functions that would accompany any system of
individual accounts, the critical decisions associated with each
function, and a partial list of the options that could be considered. 



                          Table 1
          
           Snapshot of Design and Administration
                           Issues

                Critical
Administrative  decision/
function        trade-off       Options to consider
--------------  --------------  --------------------------
Managing the    Centralize or   Build on current Social
flow of         decentralize    Security tax and payroll
information     record-         reporting structure
and money       keeping
                                Build on employer-based
                                401(k) structure

                                Build on individually
                                controlled IRA structure

Choosing        Maximize        Offer a small set of
investment      individual      indexed funds
options         choice or
                minimize risk   Offer a broad range of
                                investment options

                                Combine the two options by
                                requiring a minimum
                                account balance before a
                                broader range of options
                                is available

Paying          Maximize        Require lifetime
retirement      individual      annuities
benefits        choice or
                ensure          Make annuities voluntary
                preservation    and permit lump-sum and
                of retirement   gradual account
                benefits        withdrawals

                                Combine the two options by
                                requiring annuitization to
                                ensure at least a minimum
                                retirement income, with
                                added flexibility for
                                remainder of account
----------------------------------------------------------

      MANAGING THE FLOW OF
      INFORMATION AND
      CONTRIBUTIONS
-------------------------------------------------------- Chapter 0:4.1

When considering the design of a system of individual accounts, the
first important decision involves account administration and
management--that is, where and how the information on individuals'
contributions and the accompanying money flow would be recorded and
managed.  There are several ways in which this could be done, and the
options span a continuum ranging from a centralized record-keeping
system managed by the government to a completely decentralized system
managed by various entities in the private sector.  Each option
offers advantages and challenges. 

For example, a new system of individual accounts could build on the
current tax collection and payroll reporting system of the
government, with an agency such as the Social Security Administration
assuming record-keeping responsibilities for individual accounts. 
Alternatively, some new centralized government clearinghouse could
assume this responsibility.  Managing this information centrally
could help keep costs down by taking advantage of economies of scale. 
For example, administrative costs for the federal Thrift Savings
Plan, which centralizes both the record-keeping and investment
functions, are low--averaging about $17.00 per account in 1998. 
Centralizing these functions by building on the current system would
not be without challenges, however.  Under the current system,
employers report earnings and contributions on an individual basis
only once per year; it would take at least 7 to 22 months from the
date an individual made a contribution to the date this information
could be attributed to an individual's record.  This time lag would
likely make it necessary to pursue interim investment alternatives
and to educate individuals on the nature and impact of the lag. 
Options to change the system to enable more timely recording and
investing of contributions do exist, but they would require
significant changes in the record-keeping systems of the government
agencies, additional costs and reporting burdens for employers, or
both. 

If individual accounts were not centralized, they could be built upon
a model similar to either the current 401(k) or Individual Retirement
Account (IRA) systems.\6 While providing a wider range of
alternatives for individuals, this approach would be accompanied by
additional responsibilities and costs for employers, workers, or
both.  For example, under a 401(k) model, employers would bear the
responsibility for creating an infrastructure to quickly deposit
contributions and provide employees with links to and choices among
investment managers.  Building on an existing employer structure such
as this would pose challenges and could prove costly to employers,
however, because about 50 percent of the private sector workforce is
not covered by an employer-provided retirement plan.  Under an IRA
approach, individual employees would bear the responsibility on their
own to select an investment manager or managers and deposit their
contributions.  Under both of these decentralized options, the
appropriate government oversight role would have to be weighed and
considered. 


--------------------
\6 A 401(k) pension plan is an employer-sponsored
defined-contribution plan that allows participants to contribute,
before taxes, a portion of their salary to a qualified retirement
account.  An IRA is a personal, tax-deferred retirement account; IRA
assets can be invested in almost any kind of financial instrument. 


      PROVIDING FLEXIBILITY IN
      CHOOSING INVESTMENT OPTIONS
-------------------------------------------------------- Chapter 0:4.2

The next critical decision centers around how much choice or
discretion individuals would have in selecting who would invest their
funds and what the range of their investment options would be.  Some
proposals would allow unlimited investment choices, while others
would offer a more limited range of choices.  The primary
consideration in deciding among the proposals would be finding the
right balance between individual choice and the related risks and
costs to the individual and the government.  These inherent
trade-offs should be considered carefully. 

Proposals that build upon a centralized system often assume that the
government or some independent oversight entity would select a fund
manager (or managers) through a competitive bidding process. 
Individuals would then select from among the investment options
offered by a designated party.  Some propose that these options be
limited to a small set of passive or indexed funds similar to those
offered under the federal Thrift Savings Plan, thus minimizing risk
to the individual while providing some degree of choice.  Such an
approach would also serve to minimize administrative costs and
program complexity.  However, a centralized system of individual
accounts also raises the risk that investment decisions could become
politicized, depending on the extent of government's role in
selecting the funds and fund managers and in other investment or fund
allocation decisions.  There are, however, ways in which these risks
could be mitigated (e.g., employing master trust concepts or creating
individual participation pools). 

Other proposals would permit individuals more discretion in selecting
their fund manager or managers, either through their employers or
directly in the private market.  Under this model, individuals would
be able to select from among a much broader range of investment
options, thus providing individuals with wider latitude to maximize
their returns and enhance their retirement incomes.  However, with
that wider range of choices would come the attendant risk to
individuals that their retirement income would not be adequate, as
well as risk to the government that individuals with inadequate
retirement income would turn to the government for support from other
programs.  In addition, a wider range of choices could also lead to
added administrative complexity and higher administrative costs,
which, if not offset by significantly higher returns, would further
undermine individuals' retirement income. 

Regardless of whether individuals were offered a wide or limited
range of investment choices, there would likely be a need for
enhanced public education, especially if participation in individual
accounts was mandatory.  Some educational effort or mechanism would
be needed to provide individuals with information they could use to
make informed investment decisions and to understand the consequences
of these decisions.  For example, individuals would need information
on basic investment principles, the risks associated with available
choices, and the effect of choosing among alternatives offered for
annuitizing or otherwise withdrawing or borrowing accumulations from
the accounts.  This would be especially important for individuals who
are unfamiliar with making investment choices, for example,
low-income and less well-educated individuals who may have limited
investing experience.  Moreover, the more choices offered, the more
extensive the educational effort would need to be.  If fewer
investment choices were offered, the educational effort could be less
costly.  Who would provide such information to workers or who would
bear the cost is not clear, but it might be possible to draw from
experiences in the private pension system. 


      PRESERVING ACCOUNT RESOURCES
      FOR RETIREMENT
-------------------------------------------------------- Chapter 0:4.3

The final design element centers around how the accumulated earnings
in individual accounts would be preserved for retirement.  Ensuring
that retirement income is available for the life of the retiree is a
fundamental goal of Social Security.  Two important decisions relate
to preservation.  The first is whether to allow access to the
accounts by workers before retirement (e.g., through borrowing).  For
example, most 401(k) pension plans allow participants to borrow
against their pension accounts at relatively low interest rates.  In
prior work, we reported that relatively few plan participants--less
than 8 percent--had one or more loans from their pension accounts at
a specific point in time.\7 However, those plan participants who
borrow from their pension accounts risk having substantially lower
pension balances at retirement and, on average, may be less
economically secure than nonborrowers.  While some may argue that
individuals should be allowed the freedom to optimize their lifetime
income through borrowing from their accounts before retirement, the
added complexity and potential diminution of retirement income need
to be given serious consideration. 

The second important decision is how much flexibility to permit
workers when they retire and begin to draw on their accounts. 
Annuitization of individual accounts is one way to preserve benefits
and ensure that benefits are available for the entire life of the
retiree--no matter how long he or she lives.  However, there are many
questions to address in this area. 

  -- Because these accounts would be the personal property of
     individuals, should annuities be required or should individuals
     have the option to withdraw their account balances in a lump sum
     or through gradual payments? 

  -- Could the mechanisms that are currently available for purchasing
     annuities accommodate the significant increase in demand? 

  -- Would new structures and additional oversight be needed? 

  -- How would the various annuity options compare with those of the
     current system, and would they provide for survivors' benefits? 

  -- Should annuities offer protection from inflation? 

Once again, this is not an all-or-nothing proposition.  For example,
it would be possible to require that individuals annuitize that
portion of their accounts which would ensure a minimum retirement
income and then provide more flexibility for any funds remaining. 


--------------------
\7 These issues are discussed in 401(k) Pension Plans:  Loan
Provisions Enhance Participation but May Affect Income Security for
Some (GAO/HEHS-98-5, Oct.  1, 1997). 


      LEVEL OF ADMINISTRATIVE
      COSTS DEPENDS UPON SYSTEM
      DESIGN
-------------------------------------------------------- Chapter 0:4.4

Many people have expressed concerns about the administrative costs of
individual accounts and how these costs would affect accumulations,
especially for the small account holder.  Each of the decisions
discussed above could have a significant effect on the costs of
managing and administering individual accounts, and it will be
important to consider their effect on the preservation of retirement
income.  Administrative costs would depend upon the design choices
that were made.  The more flexibility allowed, the more services
provided to the investor, and the more investment options provided,
the higher the administrative costs would be.  For example, offering
investors the option of frequently shifting assets from one
investment vehicle to another or offering a toll-free 1-800 number
for a range of customer investment and education services could
significantly increase administrative costs.  Moreover, in addition
to decisions that affect the level of administrative costs, other
factors would need to be carefully considered, such as who would bear
the costs and how they would be distributed among large and small
accounts. 

When considering whether and how to include a system of individual
accounts as a part of Social Security reform, vital decisions on the
optimal design, administrative structure, and implementation schedule
will need to be made with great care.  A system of accounts that
spans the current 148 million workers and the 6.5 million employers,
regardless of its design, would be significantly larger than any
system we have in place today.  Such a change would take time and
careful deliberation over each of the options and trade-offs
mentioned above.  In addition, any implementation of individual
accounts would need to allow for sufficient lead time to ensure
success.  The Social Security system is one of our nation's most
important and visible programs.  Therefore, we cannot afford to incur
major implementation or administration problems.  This is especially
true because individual accounts would be highly visible to
individuals and would represent "their money."


   CONCLUSIONS
---------------------------------------------------------- Chapter 0:5

The Congress faces significant challenges in restoring sustainable
solvency to Social Security.  We have a historic opportunity to meet
these challenges because of the strength of our economy and future
budget surpluses.  We also have a historic responsibility--a
fiduciary obligation, if you will--to leave our nation's future
generations a financially stable system.  I believe it is possible to
craft a solution that will protect the Social Security benefits of
the nation's current retirees while ensuring that the system will be
there for future generations; and perhaps the answer does not lie
solely in one approach or the other--defined benefit or defined
contribution.  Bridging the gap between these approaches is not
beyond our ability.  GAO and I stand ready to provide the information
and analysis that can help the Congress meet this challenge in a way
that can exceed the expectations of all generations of Americans. 


*** End of document. ***