Social Security Reform: Information on the Archer-Shaw Proposal (Letter
Report, 01/19/2000, GAO/AIMD/HEHS-00-56).

Pursuant to a congressional request, GAO provided information on the
Archer-Shaw Social Security reform proposal, focusing on: (1) the extent
to which the proposal achieves sustainable solvency and how it would
affect the U.S. economy and the federal budget; (2) the balance struck
between the twin goals of income adequacy (level and certainty of
benefits) and individual equity (rates of return on individual
contributions); and (3) how readily such changes could be implemented,
administered, and explained to the public.

GAO noted that: (1) the Archer-Shaw proposal would not reduce
current-law benefits, and some workers could receive higher benefits
under the proposal; (2) the proposal would establish an individual
account for each eligible worker, use the balances in these accounts to
finance benefits, and remove the limits on how much retirees can earn
without having their benefits reduced; (3) the Archer-Shaw proposal
would establish mandatory individual accounts for each eligible worker
by providing workers covered by Social Security with refundable tax
credits from general revenues equivalent to 2 percent of their Old Age
Survivors and Disability Insurance (OASDI) taxable earnings for each
calendar year; (4) accounts would be managed by mutual funds, qualified
and supervised by a board of six individuals appointed by the Social
Security Trustees; (5) all account balances would be required to be
invested in qualified mutual funds maintained with a portfolio
allocation of 60 percent stock index funds and 40 percent corporate
bonds; (6) annual administration expenses would be limited to 25 basis
points, and withdrawals prior to reaching retirement (or disability)
would not be permitted; (7) upon an individual's entitlement for
retirement or disability benefits, the Social Security Administration
(SSA) would compute the monthly payment that could be provided from a
life annuity purchased with the holdings in the account; (8) the annuity
calculation would reflect the account's anticipated yield, the indexing
of payment for price inflation, and the expected payment of spouse and
survivor benefits; (9) if the computed monthly annuity amount exceeded
the level of scheduled OASDI benefits under current law, then SSA would
guarantee payment from the trust funds of the computed annuity amount
for life; (10) each month after benefit entitlement, the computed
annuity amount would be transferred from the account to the OASDI trust
funds; (11) under the Archer-Shaw proposal, the disposition of a
worker's account balance at death depends on several factors; (12) if
there are no survivors eligible for benefits and the worker dies before
receiving benefits, the account balance is transferred to the worker's
estate tax free; (13) if there are no eligible survivors and the worker
has begun receiving benefits, the account balance is transferred to the
OASDI trust funds; (14) if there are eligible survivors, the account
balance is transferred to the survivor's account; and (15) the proposal
assumes a reduction in the OASDI payroll tax rate from 12.4 percent to
9.9 percent in 2050 and to 8.9 percent in 2060.

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

 REPORTNUM:  AIMD/HEHS-00-56
     TITLE:  Social Security Reform: Information on the Archer-Shaw
	     Proposal
      DATE:  01/19/2000
   SUBJECT:  Social security benefits
	     Future budget projections
	     Financial management
	     Retirement benefits
	     Federal social security programs
	     Investment planning
	     Mutual funds
	     Tax credit
	     Securities
IDENTIFIER:  Old Age Survivors and Disability Insurance Program
	     Social Security Program
	     Gross Domestic Product
	     Old Age and Survivors Insurance Trust Fund
	     Medicare 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.                         **
******************************************************************

Report to the Ranking Minority Member, Subcommittee on Social Security,
Committee on Ways and Means, House of Representatives

January 2000

SOCIAL SECURITY REFORM

Information on the Archer-Shaw Proposal
*****************

*****************

GAO/AIMD/HEHS-00-56

Letter                                                                     3

Appendixes

Appendix I:  Criteria for Evaluating Social Security Reform Proposals

                                                                         24

Appendix II:  The Economic Model and Assumptions

                                                                         26

Appendix III:  Comments From the Social Security Administration

                                                                         30

Related GAO Products

                                                                         32

Figure 1:  Archer-Shaw: Unified Deficits/Surpluses as a Share of GDP9

Figure 2:  Archer-Shaw: Debt Held by the Public as a Share of GDP10

Figure 3:  Composition of Spending as a Share of GDP in 1998 and
Under No Action and Archer-Shaw                 11

Figure 4:  Social Security Spending as a Share of Total Federal
Revenue in 1998 and Under No Action and Archer-Shaw12

Figure 5:  No Action Model Assumptions          28

CBO     Congressional Budget Office

OASDI   Old Age Survivors and Disability Insurance

SSA     Social Security Administration

Social Security: Criteria for Evaluating Social Security Reform Proposals
Social Security Reform: Implementation Issues for Individual Accounts
See Social Security: Individual Accounts as an Element of Long-Term
Financing Reform (GAO/T-HEHS-99-86, March 16, 1999); Social Security
Reform: Implications of Private Annuities for Individual Accounts
(GAO/HEHS-99-160, July 30, 1999); Social Security: Issues in Comparing
Rates of Return With Market Investments (GAO/HEHS-99-110, August 5, 1999);
Social Security Reform: Implications of Raising the Retirement Age
                                                       Comptroller General 
                                                       of the United States

B-284323 

January 19, 2000

The Honorable Robert T. Matsui
Ranking Minority Member
Subcommittee on Social Security 
Committee on Ways and Means
House of Representatives

Dear Mr. Matsui:

This report responds to your request that we apply our criteria for
assessing Social Security reform proposals to the proposal outlined by
Representative Archer, Chairman of the Committee on Ways and Means, and
Representative Shaw, Chairman of the Subcommittee on Social Security,
Committee on Ways and Means. In November 1999, we issued a report
presenting the results of our assessments of several Social Security
reform proposals, including Archer-Shaw./Footnote1/ Our November report
used a briefing slide format, thereby documenting congressional briefings
we conducted prior to report issuance, including one to your office. This
report responds to your request that we provide you with information from
our briefing and some additional information on the Archer-Shaw proposal
in the form of a report. Since legislation on the proposal had not yet
been introduced at the time we did our work, we based our interpretation
of the proposal in large part on the memorandum provided by the Office of
the Chief Actuary at the Social Security Administration (SSA) in April
1999 when it estimated the proposal's long-range financial effect on the
program./Footnote2/

As agreed with your office, our report is based on the analytic framework
we provided to the Congress last March./Footnote3/ That framework consists
of three basic criteria: 

o   the extent to which the proposal achieves sustainable solvency and
  how it would affect the U.S. economy and the federal budget,

o   the balance struck between the twin goals of income adequacy (level
  and certainty of benefits) and individual equity (rates of return on
  individual contributions), and

o   how readily such changes could be implemented, administered, and
  explained to the public./Footnote4/

In evaluating proposals against the three basic criteria, we used a set of
detailed questions that help describe potential effects of reform
proposals on important policy and operational aspects of public concern.
(See appendix I for the list of questions.) The analysis presented below
summarizes our results.

The Archer-Shaw Proposal
------------------------

The Archer-Shaw proposal, known as "The Social Security Guarantee Plan,"
would not reduce current-law benefits, and some workers could receive
higher benefits under the proposal. The proposal would establish an
individual account for each eligible worker and use the balances in these
accounts to help finance benefits. The proposal would also remove the
limits on how much retirees can earn without having their benefits
reduced./Footnote5/ The proposal as estimated in April 1999 by SSA's
Office of the Chief Actuary includes payroll tax reductions in the out-
years. Congressional staff working on the proposal have informed us that
the payroll tax reductions are under review.

The Archer-Shaw proposal would establish mandatory individual accounts for
each eligible worker. To do this, the proposal would provide workers
covered by Social Security with refundable tax credits from general
revenues equivalent to 2 percent of their Old Age Survivors and Disability
Insurance (OASDI) taxable earnings for each calendar year. Accounts would
be managed by mutual funds, qualified and supervised by a board of six
individuals appointed by the Social Security Trustees. All account
balances would be required to be invested in qualified mutual funds
maintained with a portfolio allocation of 60 percent stock index funds and
40 percent corporate bonds. Annual administrative expenses would be
limited to 25 basis points, and withdrawals prior to reaching retirement
(or disability) would not be permitted.

Upon an individual's entitlement for retirement or disability benefits,
SSA would compute the monthly payment that could be provided from a life
annuity purchased with the holdings in the account. Under the proposal,
the annuity calculation would reflect the account's anticipated yield, the
indexing of annuity payments for price inflation, and the expected payment
of spouse and survivor benefits. If the computed monthly annuity amount
exceeded the level of scheduled OASDI benefits under current law, then SSA
would guarantee payment from the trust funds of the computed annuity
amount for life. If the computed annuity amount was less than the OASDI
benefit level, then the OASDI benefit would be payable for life. Each
month after benefit entitlement, the computed annuity amount would be
transferred from the account to the OASDI trust funds. 

Under the Archer-Shaw proposal as specified for the Office of the Chief
Actuary in April 1999, the disposition of a worker's account balance at
death depends on several factors. If there are no survivors eligible for
benefits on the basis of the worker's earnings and the worker dies before
receiving benefits, the account balance is transferred to the worker's
estate tax free. If there are no eligible survivors and the worker has
begun receiving benefits, the account balance is transferred to the OASDI
trust funds. If there are eligible survivors, the account balance is
transferred to the survivor's (e.g., spouse's) account.

The Archer-Shaw proposal as estimated by the SSA's Office of the Chief
Actuary assumes a reduction in the OASDI payroll tax rate from 12.4
percent to 9.9 percent in 2050 and to 8.9 percent in 2060. These
reductions reflect the specified portfolio allocation and an assumed real
asset yield (after inflation) of 5.35 percent net of administrative
expense. This combined yield is based on historical yields for equities
and long-term corporate bonds./Footnote6/ 

Scope and Methodology
---------------------

As you requested, we used our long-term economic model in assessing the
Archer-Shaw proposal against the first criterion, that of financing
sustainable solvency. Since 1992, we have provided the Congress with a
long-term perspective by modeling the implications of differing fiscal
policy paths for the nation's economy./Footnote7/ Although any proposal's
ability to achieve and sustain solvency is sensitive to economic and
budgetary assumptions, using a common framework can facilitate comparisons
of alternative reform proposals. Our simulation results are presented not
as forecasts but rather as a useful way to compare the potential outcomes
of alternative policies within a common economic framework. For Social
Security and Medicare spending, we use the Trustees' intermediate
estimates; in other respects, we generally rely on the Congressional
Budget Office's (CBO) fiscal and economic assumptions.

In this report we used our long-term model to simulate the potential
fiscal and economic impacts of the Archer-Shaw proposal over a 75-year
projection period. In so doing, we relied on income and cost estimates for
the proposal prepared by SSA's Office of the Chief Actuary, which reflect
payroll tax reductions in 2050 and 2060, and adapted our model as
appropriate to reflect specific reform proposal provisions. Given our
reliance on that Office's estimates in modeling the proposal, the proposal
descriptions presented in this report are generally based on its
description./Footnote8/ As in our work on other reform proposals, we
considered the proposal in isolation. That is, we did not include any
other proposals made by reform sponsors, such as proposed non-Social
Security related tax cuts or spending increases, that would affect the
government's fiscal position.

Our analysis of the Archer-Shaw proposal included comparison with three
other fiscal policy paths developed in our earlier work: (1) "No Action,"
or "Save the Surplus," which assumes continuation of current-law fiscal
policies, i.e., preservation of the entire projected unified surplus,
(2) "Eliminate non-Social Security surpluses," and (3) "Long-term on-
budget balance."/Footnote9/ The No Action path reflects the assumptions of
the CBO's July 1999 baseline, which generally assumed the continuation of
then-current law fiscal policies. With respect to discretionary spending,
the CBO assumptions implied that there is no future emergency spending and
that total actual spending would fall within the existing discretionary
caps through 2002./Footnote10/ No Action also uses the Trustees'
intermediate or "best estimates" of the program's income and cost rates,
which implicitly assume that all promised Social Security benefits will be
paid throughout the
75-year simulation period, including after the projected exhaustion of the
OASDI trust funds in 2034. (Appendix II contains more information on the
assumptions underlying our long-term model.)

We used qualitative research to examine how well the proposal balances
adequacy and equity concerns and provides for reasonable implementation
and communication of any changes. In so doing, we relied on our issued and
ongoing body of work on Social Security reform. This work addresses
various issues raised by reform approaches, including establishing
individual accounts, raising the retirement age, and the impact of reforms
on minorities and women./Footnote11/ 

Financing Sustainable Solvency
------------------------------

This criterion assesses the extent to which the proposal, if adopted,
would achieve sustainable solvency, including how the proposal would
affect the federal budget and the economy. Elements considered in this
criterion include a proposal's effects on (1) the federal government's
unified surplus or deficit in the long term, (2) debt held by the public,
(3) solvency of the OASDI trust funds, and (4) national saving. Other
elements, such as how changes would be financed and potential effects on
future program liabilities, are also considered. 

In examining a proposal's long-term economic and budgetary effects, we
compared our simulation results for the proposal with a policy of "No
Action," in which the entire surplus would be saved. Compared to this
reference point, the Archer-Shaw proposal would increase budgetary
pressure through the middle of the next century, then reduce budgetary
pressure in the latter years of the 75-year simulation period.

Figure****Helvetica:x11****1:    Archer-Shaw: Unified Deficits/Surpluses
                                 as a Share of GDP

*****************

*****************

Source: GAO's long-term model.

As shown in figure 1, the proposal compared to the No Action alternative
reduces projected unified surpluses and increases projected unified
deficits as a share of gross domestic product (GDP) through 2046.
Thereafter, due to the estimated increasingly larger amounts being
transferred from the individual account balances to the OASDI trust funds,
the Archer-Shaw proposal would decrease unified deficits compared to No
Action. Beginning around 2030, under the account yield assumptions made by
the Office of the Chief Actuary, the recapture of funds from the
individual accounts would begin to exceed the general revenue tax credits
used to fund the accounts. As shown in our long-term model results
presented in figure 1, unified deficits under Archer-Shaw emerge in 2023
and reach 9.7 percent of GDP in 2074; under No Action, unified deficits
emerge in 2030 and reach 12.7 percent of GDP in 2074. 

Compared to No Action, the Archer-Shaw proposal would increase levels of
debt held by the public until shortly before the end of the 75-year
simulation period. The levels of debt shown in figure 2 follow from the
lower unified surpluses and higher unified deficits under Archer-Shaw than
under No Action shown in figure 1. 

Figure****Helvetica:x11****2:    Archer-Shaw: Debt Held by the Public as a
                                 Share of GDP

*****************

*****************

Source: GAO's long-term model.

The lower unified surpluses mean less debt reduction than under No Action,
and the higher unified deficits under Archer-Shaw mean more government
borrowing and hence greater associated interest expense. After 2046,
recaptured account balances more than offset the general revenue tax
credits used to fund the accounts plus this higher interest expense,
resulting in lower unified deficits compared to No Action. These lower
deficits under Archer-Shaw slow the build-up of debt so that after 2067,
debt held by the public would be lower than under No Action.

The Archer-Shaw proposal does not raise payroll taxes and does not use
payroll tax revenue to finance the individual accounts. In effect, it
draws on general revenues to finance individual accounts through the
mechanism of a refundable tax credit equal to 2 percent of taxable
payroll. In congressional testimony, the proposal sponsors stated that the
Archer-Shaw proposal would use Social Security surpluses over the next 15
years to finance the accounts./Footnote12/ Under the estimates prepared by
the Office of the Chief Actuary, the transfers from the individual
accounts would exceed the general revenue credits beginning about 2031.

Figures 3 and 4 show program cost under No Action and Archer-Shaw as a
share of the economy and as a share of the federal budget, respectively.
Figure 3 shows that compared to No Action, the Archer-Shaw proposal has
little impact on the net government cost of Social Security as a share of
GDP in 2030 but cuts the net cost of the program roughly in half by 2074. 

Figure****Helvetica:x11****3:    Composition of Spending as a Share of GDP
                                 in 1998 and Under No Action and Archer-Shaw

*****************

*****************

Source: GAO's long-term model.

Figure 4 shows that compared to No Action, the Archer-Shaw proposal would
begin to lower net Social Security spending slightly as a share of federal
revenues in 2030. By 2074, net program spending under Archer-Shaw would
consume about half as much of federal revenues as in No Action--or about
the same share of federal revenues as today. 

Figure****Helvetica:x11****4:    Social Security Spending as a Share of
                                 Total Federal Revenue in 1998 and Under
                                 No Action and Archer-Shaw

*****************

*****************

Source: GAO's long-term model.

The analyses presented in figures 3 and 4 assume that total federal
revenue under Archer-Shaw is net of the tax credit for individual
accounts, and Social Security spending is net of the offset ("recapture")
from the individual accounts. This analysis of net government cost follows
the framework implicit in the estimates provided by SSA's Office of the
Chief Actuary./Footnote13/

You asked that we also provide information in this report on alternative
approaches to analyzing the proposal's cost. Alternative approaches might
focus on gross cost--that is, on the levels of total benefits provided.
Such a measure would indicate the gross cost of the retirement benefits
funded with resources flowing through the Social Security system. In this
analysis, the amounts recaptured from the individual account balances
would be included in total federal revenues, and the cost of the benefits
financed by those amounts would be included in the gross cost. Under this
approach, Social Security spending under Archer-Shaw would appear larger
as a share of GDP than in our analysis of net government cost under the
proposal and also larger as a share of total federal revenues. Compared to
No Action, however, Social Security spending under the Archer-Shaw
proposal would appear approximately the same as a share of GDP and
somewhat smaller as a share of total federal revenues in the long
run./Footnote14/ An analysis of gross cost would not affect our simulation
results of long-term fiscal position and debt held by the public under
Archer-Shaw. 

With regard to national saving, the actual effect of the Archer-Shaw
proposal is unclear due to uncertainties in predicting the saving behavior
of private households and individuals. On a national income accounting
basis, the Archer-Shaw proposal would initially have no net effect.
Through the tax credit used to create the individual accounts, government
saving would be reallocated to private saving with no net change in
national saving. The underlying assets in the economy would be unchanged,
as would the total income generated by those assets. However, the
government would ultimately receive a greater portion of the returns
because of the recapture of the account balances invested in higher-
yielding equities. If payroll tax reductions were implemented in 2050 and
2060, the net effect would be to reduce national saving. 

In a report issued in June 1999, we analyzed the effects on national
saving of several types of individual accounts as proposed in Social
Security reform proposals./Footnote15/ We found that these effects would
depend on several variables. These would include the financing and
structure of such accounts, for example, whether funds would come from
inside or outside the Social Security system and whether withdrawals would
be permitted. Where the funds used to finance the individual accounts come
from outside the Social Security system, the effect on government savings
would depend on what would have been done with the surplus or revenue if
it had not been used to finance individual accounts. 

o   If the funds would have been used to pay down debt (as assumed in No
  Action/Save the Unified Surplus), then the direct effect of using funds
  to finance individual accounts would be to reduce government saving and
  increase private saving by the same amount. 

o   If the funds would have been used to finance additional government
  consumption spending, then any increase in private saving due to the
  individual accounts would increase national saving.

o   If the funds would have been used for a tax cut, then national saving
  will increase if the individual accounts generate more private saving
  than the tax cut.

The actual effect of any reform proposal with individual accounts on
national saving--including Archer-Shaw--would depend on behavioral
offsets. For example, if households are forgoing current consumption by
saving for their retirement, they may reduce these savings in response to
a potential increase in future retirement benefits from individual
accounts. There is, however, no expert consensus on how Social Security
reform proposals would affect the saving behavior of private households
and businesses.

According to the analysis provided by SSA's Office of the Chief Actuary,
the Archer-Shaw proposal would restore 75-year actuarial balance to the
OASDI trust funds and produce a stable trust fund ratio./Footnote16/ By
the end of the projection period, the Office of the Actuary estimated the
trust fund ratio to be over 200 percent and rising. Under the Actuary's
estimates, the proposal would result in a sustainable system with or
without reductions in the payroll tax rate. Without the payroll tax
reduction, the OASDI trust fund ratio would be expected to rise to over 10
times annual outgo by the end of the long-range period due to the
estimated magnitude of the transfers flowing from the individual accounts.

The analysis prepared by the Office of the Chief Actuary noted, however,
that the proposal's effects on trust fund solvency depend greatly on the
assumed yields of the individual account investments. SSA's analysis
assumed an expected real portfolio yield of 5.35 percent, net of
administrative expense./Footnote17/ In a sensitivity analysis that assumed
yields of
1 percentage point lower than expected and no payroll tax reductions, the
Office of the Chief Actuary estimated that the OASDI trust funds would be
exhausted by 2048 and the actuarial balance would be -0.08 percent of
taxable payroll for the 75-year projection period, compared with -2.07
under current law. In a second sensitivity analysis that assumed account
yields of 1 percentage point higher than expected, the Office of the Chief
Actuary estimated that the payroll tax rate reductions would be larger and
would start sooner and the actuarial deficit would be eliminated. The
actuarial balance would be an estimated +0.07 percent of payroll, and the
trust fund ratio would be stable at about 300 percent at the end of the
75-year period.

Should the assumptions used by the Office of the Chief Actuary about
future account yields prove overly optimistic and account balances indeed
fall below what is needed to fund promised benefits, the OASDI trust funds
would bear the resulting liability. This means that payments could only be
made from the OASDI trust funds, not the general fund./Footnote18/
Further, descriptions of the proposal do not include a safety valve to
control program growth or to limit federal liabilities that could result.
Some descriptions of the proposal, however, could be interpreted to imply
a link between future payroll tax reductions and the realization of
assumed account yields. 

Balancing Adequacy and Equity
-----------------------------

This criterion evaluates the balance struck between the twin goals of
income adequacy and individual equity. Income adequacy refers to the level
and certainty of benefits provided to retirees, the disabled, dependents,
and survivors. It is particularly important for low-income workers who are
most reliant on the program, and may be achieved, in part, through a
progressive benefit formula. Individual equity refers to rates of return
on individual contributions. That is, it concerns the relationship between
the benefit individuals receive and the contributions they have made to
the Social Security system. Individual equity would be affected by giving
workers greater choice and control over their contributions, which in turn
could give them the potential to earn higher rates of return by investing
in higher yielding assets, though generally by taking on greater
investment risk at the same time. The current Social Security system makes
certain trade-offs between the degree of income adequacy and individual
equity provided by its benefit structure. Redistributive transfers
embedded in the current system create an implicit "safety net" for workers
and their families. At the same time, linking benefits to contributions
invokes the standard of individual equity.

Because the Archer-Shaw proposal makes no changes to the benefit structure
of the current Social Security system, it would not adversely affect
income adequacy as compared to current promised benefit levels.
Specifically, the Archer-Shaw proposal maintains current-law benefits for
current and future retirees, including low-income workers and others most
reliant on Social Security./Footnote19/ It makes no changes to disabled,
dependent, or survivor benefits and thus, retains the existing safety net.
The proposal makes no changes from the current Social Security structure
in the way workers are covered, and it preserves the progressivity of the
existing system. In addition, it retains the compulsory nature of the
current payroll tax. According to SSA actuaries, higher earners,
especially two-earner couples with high earnings levels, would be more
likely than lower income workers or one-earner couples to receive higher
benefits under the proposal than under current law.

Consequently the Archer-Shaw proposal's individual account structure does
have implications for individual equity. The proposal could enhance
individual equity to the extent that it results in higher rates of return
on workers' contributions. At the same time, individuals bear no
investment risk because the proposal guarantees current law benefits. The
Archer- Shaw proposal would pay current law benefits out of the OASDI
trust funds, with the trust funds recapturing an amount from the
individual account to finance those benefits. In the short run, the effect
on individual equity would be limited, as the SSA actuaries estimate that
these transfers would almost always take the entire account balances,
leaving workers with the equivalent of their current law benefit. The
Archer-Shaw proposal also affects individual equity through its provision
to reduce payroll tax rates. /Footnote20/ According to SSA's projections,
future workers would pay lower payroll tax rates that would take effect in
2050 and 2060 with no benefit reductions. These tax reductions would
increase rates of return on contributions, enhancing individual equity. 

Individual equity applies not only to comparisons made within a generation
of workers, but across generations as well. The Archer-Shaw proposal's
stable patterns of contributions and benefits could result in fairly level
rates of return across generations. Benefit levels would be the same as
under current law over the next 75 years for virtually all workers. The
amounts credited to an individual's account are increased immediately by
2 percent, and then remain constant until proposed tax cuts take effect in
2050 and 2060./Footnote21/ Future workers benefiting from these tax cuts
would then receive higher rates of return./Footnote22/

Increasing workers' control over contributions and providing greater
investment choice also potentially enhances individual equity./Footnote23/
The Archer-Shaw proposal would not give individuals as much control over
their accounts as 401(k) accounts since individuals would not be able to
borrow or otherwise use these funds for nonretirement purposes prior to
death or retirement. However, in certain circumstances, workers could
bequeath their account balances to their heirs. The Archer-Shaw proposal
does provide investment choices to workers, subject to certain
limitations. Account investments are constrained to mutual funds with a
portfolio allocation of 60 percent equities and 40 percent fixed income
securities. Within this overall constraint, investment choices will be
limited largely to those bond and equity index funds approved by the
proposal's Social Security Guarantee Board.

Implementing and Administering Reforms
--------------------------------------

This criterion evaluates how readily proposed changes could be
implemented, administered, and explained to the public. Implementation and
administration issues are important because they have the potential to
delay--if not derail--reform if they are not considered early enough for
planning purposes. Moreover, such issues can influence policy choices--
feasibility and cost should be integral factors in ultimate decisions
regarding the Social Security program. In addition, potential transparency
and public education needs associated with various proposals should be
considered. Reforms that are not well understood could face difficulties
in achieving broad public acceptance and support. 

As reform proposals are developed, two key design elements are often
overlooked: timing and funding for implementation. The Archer-Shaw
proposal does not explicitly provide funding for implementation of its
changes to the Social Security system, nor does it provide a specific time
frame for implementing those changes. However, the proposal would
establish a Social Security Guarantee Board, which would have supervisory
control over funds investment and could be responsible for establishing a
time frame for implementation./Footnote24/ The Board's main role would be
to develop investment policies that provide prudent diversification of
investments and low administrative costs. The Board also could have a role
in implementing safeguards against politically motivated investing of the
accounts./Footnote25/

The Archer-Shaw proposal estimates that administrative costs for the
program would not exceed 25 basis points. This estimate may not be
realistic, especially in the long term. The cost of administering a system
that includes individual accounts would depend on decisions made about who
would assume new administrative and recordkeeping responsibilities, how
much choice or discretion individuals would have in selecting and changing
their investment options, and how workers would receive their benefits
when they retire./Footnote26/ While the Archer-Shaw proposal's reliance on
index funds could reduce administrative expenses to some degree, its
administrative costs would also depend, in large part, on the number of
indexed funds ultimately offered to participants. In previous work we
reported that, in general, administrative costs are higher for more
decentralized systems and for those offering broader investment choices,
more customer service options, or both./Footnote27/ In addition, the
estimate of 25 basis points does not take into account other costs related
to the establishment of individual accounts, such as the cost to SSA for
modifying agency software to incorporate changes in payout calculations,
the need for employee hiring and training, acquisition of equipment and
facilities to support customer service, and promulgation of new
regulations, policies and procedures. These categories represent primarily
one-time costs./Footnote28/ 

Public understanding of both the financing and benefit structures is
critical to any reform proposal. Public confidence is increased when the
financing mechanism is readily understood by all, and broad-based support
for the program is more easily maintained. Various elements of the Archer-
Shaw proposal may be particularly difficult to explain clearly and simply.
Participants will need to understand that the tax credit financing
structure of the proposal does not mean that individuals would actually
file for this credit on their tax returns. Rather, it would be
automatically credited to their individual accounts. Similarly, the
complex "offset" feature of the benefit structure, which reduces Social
Security benefits by the amount that is accumulated in the individual
account, must also be clearly explained. Otherwise retirees may expect a
larger return than the proposal actually provides, potentially creating an
"expectations gap." For example, some people may devote a significant
amount of time to choosing between the different investment options, only
to have their choices provide no more than an incremental increase in
their retirement benefit. Their perception could then be that they have
received a poor rate of return on their individual accounts. To avoid
these perception problems an extensive education program will be necessary. 

At the same time, investment education will be needed to give individuals
information about financial planning and general investment strategies, as
well as the specific choices made available to them and the risks
associated with each. The need to provide these education programs could
place an administrative burden, as well as extra costs, on SSA. The Archer-
Shaw proposal does not directly provide for such a program or the
resources necessary for SSA or other agencies to conduct such an effort.

Agency Comments and Our Evaluation
----------------------------------

We requested comments on a draft of this report from SSA. SSA generally
agreed with GAO's treatment of the issues. SSA's comments, which can be
found in appendix III, reemphasized the point made by the Office of the
Chief Actuary that the proposal's success in reaching solvency is
sensitive to assumptions about future account yields. SSA also provided a
number of technical comments, which we incorporated as appropriate.

Unless you release the contents of this report earlier, we will not
distribute it until 30 days from the date of this letter. At that time we
will send copies of this report to the Honorable Bill Archer, Chairman,
and the Honorable Charles B. Rangel, Ranking Member, House Ways and Means
Committee; the Honorable Clay Shaw, Chairman, Subcommittee on Social
Security, House Ways and Means Committee; other interested congressional
committees; the Honorable Kenneth S. Apfel, Commissioner of Social
Security; and the Honorable Lawrence Summers, Secretary of the Treasury.
Copies will be made available to others upon request. 

If you or your staff have any questions about this report, please contact
Cynthia M. Fagnoni, Director, Education, Workforce, and Income Security
Issues, on (202) 512-7215 or Paul L. Posner, Director, Budget Issues, on
(202) 512-9573.

Sincerely yours,

*****************

*****************

David M. Walker
Comptroller General
of the United States

--------------------------------------
/Footnote1/-^Social Security: Evaluating Reform Proposals (GAO/AIMD/HEHS-
  00-29, November 4, 1999).
/Footnote2/-^Memo from the Office of the Chief Actuary dated April 29,
  1999. SSA has informed us that congressional staff responsible for
  developing the Archer-Shaw proposal provided all specifications cited in
  the memo.
/Footnote3/-^(GAO/T-HEHS-99-94, March 25, 1999).
/Footnote4/-^(GAO/HEHS-99-122, June 18, 1999) and Social Security Reform:
  Administrative Costs for Individual Accounts Depend on System Design
  (GAO/HEHS-99-131, June 18, 1999).
/Footnote5/-^Currently, the "earnings test" requires the withholding of
  benefits where beneficiaries under age 70 have earnings in excess of
  certain amounts. In 1999, for those aged 62 to 64, benefits are reduced
  by $1 for every $2 of earnings over $9,600; for those aged 65 to 69,
  benefits are reduced by $1 for every $3 of earnings over $15,500.
/Footnote6/-^In discussing the Archer-Shaw proposal, SSA's Office of the
  Actuary noted that the assumed long-term average real yield for stocks
  (7 percent) was based on the work of the 1994-96 Advisory Council, and
  the assumed real yield on long-term corporate bonds was based on the
  long-term historical premium paid (0.5 percentage point) over long-term
  Treasury notes. 
/Footnote7/-^For more information on GAO's long-term model, see Budget
  Issues: Long-Term Fiscal Outlook (GAO/T-AIMD/OCE-98-83, February 25,
  1998).
/Footnote8/-^We supplemented the Actuary's memo with other material
  prepared by proposal sponsors and discussions with congressional staff.
/Footnote9/-^"Eliminate non-Social Security surpluses" assumes that
  unspecified actions (i.e., spending increases and/or tax cuts) eliminate
  all projected on-budget surpluses; these changes from current law are
  then projected through the end of the 75-year simulation period. "Long-
  term on-budget balance" assumes that on-budget surpluses are eliminated
  but thereafter the on-budget portion of the budget is kept in balance
  through unspecified actions (i.e., spending cuts or revenue increases)
  throughout the rest of the simulation period. See appendix II for
  further details on the assumptions underlying these paths.
/Footnote10/-^In December 1999, CBO estimated that discretionary outlays
  in fiscal year 2000 would exceed the discretionary caps by $16.6
  billion. CBO also estimated that total fiscal year 2000 spending would
  be $31.8 billion higher than its July 1999 baseline estimates. CBO noted
  that its January 2000 baseline estimates were likely to present a more
  favorable picture in light of recent economic trends. See CBO's Final
  Sequestration Report for Fiscal Year 2000 and also CBO's The Budget for
  Fiscal Year 2000: An End-of-Session Summary.
/Footnote11/-^(GAO/HEHS-99-112, August 27, 1999); Social Security Reform:
  Implications for Women (GAO/T-HEHS-99-52, February 3, 1999); and Social
  Security and Minorities: Current Benefits and Implications of Reform
  (GAO/T-HEHS-99-60, February 10, 1999). 
/Footnote12/-^Testimony of Representatives Bill Archer and Clay Shaw
  before the House Budget Committee Task Force on Social Security, June
  29, 1999.
/Footnote13/-^As of the date of this report, no CBO cost estimate for the
  Archer-Shaw proposal was available. 
/Footnote14/-^As noted above, in this analysis total federal revenue would
  include the transfer amounts.
/Footnote15/-^See Social Security: Capital Markets and Educational Issues
  Associated With Individual Accounts (GAO/GGD-99-115, June 28, 1999), pp.
  30-32.
/Footnote16/-^The trust fund ratio is a measure of the adequacy of the
  trust fund level in the short term. The ratio represents the proportion
  of a year's outgo that could be paid with the funds available at the
  beginning of the year.
/Footnote17/-^In its November 1999 report to the Social Security Advisory
  Board, the 1999 Technical Panel on Assumptions and Methods recommended
  that SSA lower its assumptions about expected yield for both equities
  and government securities. Applying the technical panel's
  recommendations to the Archer-Shaw proposal would lead to a 0.9
  percentage point lower return on individual accounts than the return
  used in the estimates made by the Office of the Chief Actuary. The
  Technical Panel is composed of actuaries, economists, and demographers
  appointed by the Social Security Advisory Board to review the
  assumptions and methodology used to project the future financial status
  of the OASDI trust funds.
/Footnote18/-^This statement is based on analysis of proposal
  descriptions, not on the specific wording of the legislation, which is
  not yet available. 
/Footnote19/-^If account balances fall below what is needed to fund
  promised benefits, the OASDI trust funds would bear the resulting
  liability. To address this liability, benefit or revenue adjustments
  could be required.
/Footnote20/-^As stated earlier, this tax cut provision is under review.
/Footnote21/-^In a technical sense, contribution rates in the form of
  higher payroll taxes are not increased, as Archer-Shaw relies on a
  refundable tax credit to finance its individual account feature. Under
  the proposal, Treasury will transfer an aggregate 2 percent of OASDI
  taxable earnings to Social Security for these accounts. However, since
  SSA then credits the individual account of each covered worker with this
  amount and provides periodic information on the performance of those
  contributions, they can be considered a de facto increase in workers'
  contribution rates.
/Footnote22/-^By comparison, raising the payroll tax under the current
  system immediately by 2.15 percentage points would result in the same
  actuarial balance as the Archer-Shaw proposal. Such a tax increase would
  also pay current law benefits and result in a constant tax rate over the
  next 75 years. Therefore, these two approaches would have very similar
  effects on generations for the next several decades, though the Archer-
  Shaw proposal would improve rates of return for generations in the
  distant future.
/Footnote23/-^Increased choice and control over investments could also
  expose individuals to greater financial risk and could lead to lower
  rates of return.
/Footnote24/-^SSA has expressed concerns about lead time for the Board to
  get established, including time for operational preparation, legislative
  interpretation, regulation development and publication, and systems
  development.
/Footnote25/-^While none of the proposals we have evaluated thus far have
  provided specific information about how they would be implemented, the
  Archer-Shaw proposal includes more detail than some others we reviewed.
  See Social Security: Evaluating Reform Proposals (GAO/AIMD/HEHS-00-29,
  November 4, 1999).
/Footnote26/-^For more detail on implementation issues, see Social
  Security Reform: Implementation Issues for Individual Accounts (GAO/HEHS-
  99-122, June 18, 1999).
/Footnote27/-^For more detail on administrative cost issues see Social
  Security Reform: Administrative Costs for Individual Accounts Depend on
  System Design (GAO/HEHS-99-131, June 18, 1999).
/Footnote28/-^The proposal also does not clearly discuss the treatment of
  fees and commissions to the certified qualified professional asset
  managers who will manage the accounts.

CRITERIA FOR EVALUATING SOCIAL SECURITY REFORM PROPOSALS
========================================================

Financing Sustainable Solvency
------------------------------

This criterion evaluates the extent to which the proposal achieves
sustainable solvency, including how the proposal would affect the economy
and the federal budget.

To what extent does the proposal:

o   Reduce future budgetary pressures?

o   Reduce debt held by the public?

o   Reduce the cost of the Social Security system as a percentage of GDP?

o   Reduce the percentage of federal revenues consumed by the Social
  Security system?

o   Increase national saving?

o   Restore 75-year actuarial balance and create a stable system?

o   Raise payroll taxes, draw on general revenues, and/or use Social
  Security trust fund surpluses to finance changes?

o   Create contingent liabilities?

o   Include "safety valves" to control future program growth?

Balancing Adequacy and Equity
-----------------------------

This criterion evaluates the balance struck between the twin goals of
income adequacy (level and certainty of benefits) and individual equity
(rates of return on individual contributions).

To what extent does the proposal:

o   Change current-law benefits for current and future retirees?

o   Maintain benefits for low-income workers who are most reliant on
  Social Security?

o   Maintain benefits for the disabled, dependents, and survivors?

o   Ensure that those who contribute receive benefits?

o   Provide higher replacement rates for lower income earners?

o   Expand individual choice and control over program contributions?

o   Increase returns on investment?

o   Improve intergenerational equity?

Implementing and Administering Reforms
--------------------------------------

This criterion evaluates how readily such changes could be implemented,
administered, and explained to the public.

To what extent does the proposal:

o   Provide reasonable timing and funds for implementation and result in
  reasonable administrative costs?

o   Allow the general public to readily understand its financing
  structure and increase public confidence?

o   Allow the general public to readily understand the benefit structure
  and avoid expectations gaps?

o   Limit the potential for politically motivated investing? 

THE ECONOMIC MODEL AND ASSUMPTIONS
==================================

GAO's long-term economic model has been used since 1992 to assess the
potential fiscal and economic impacts of alternative policy choices. Based
on an economic growth model developed by economists at the Federal Reserve
Bank of New York, the model simulates the interrelationships between the
budget and the economy over a 75-year projection period. The key
interaction between the budget and the economy in the model is the effect
of the unified federal deficit/surplus on the amount of national saving
available for investment, which influences long-term economic growth. 

The long-term simulations generated by the model illustrate the relative
fiscal and economic outcomes associated with alternative policy paths and
thus provide a useful way to compare alternatives within a common economic
framework. The simulations are not predictions of what would actually
happen in the future. For example, in reality policymakers likely would
take action before the occurrence of the negative out-year fiscal and
economic consequences reflected in some simulated fiscal policy paths. In
addition, the model reflects the interrelationships between the budget and
the economy over the long term and does not capture their interaction
during short-term business cycles. Assumptions underlying the model have
been chosen to be conservative in recognition of the uncertainties
inherent in long-term simulations,/Footnote1/ but variation in these
assumptions generally would not affect the relative outcomes of the
alternative policies being compared.

In our November 1999 report, we compared several Social Security reform
proposals to the following three fiscal policy paths developed as part of
GAO's ongoing model work.

o   No Action/Save the Surplus assumes no changes in current policies and
  thus results in saving the unified surpluses. This assumption implies
  no emergency spending and actual spending that falls within the
  existing discretionary caps. Thus, unified budget surpluses through
  2029 are used to reduce debt held by the public. Thereafter, unified
  deficits are permitted to emerge. Discretionary spending follows CBO's
  10-year projections, which assume compliance with the spending caps
  through 2002 and growth with inflation through 2008. Thereafter, we
  assume discretionary spending grows with the economy.

o   Eliminate non-Social Security surpluses assumes that permanent
  unspecified policy actions (i.e., spending increases and/or tax cuts)
  are taken through 2009 that eliminate the projected on-budget
  surpluses. Thereafter, these unspecified actions are projected through
  the end of the simulation period. On-budget deficits emerge in 2010,
  followed by unified deficits in 2017.

o   Long-term on-budget balance assumes that the on-budget surplus is
  eliminated through 2009, as in the previous path. Thereafter, the on-
  budget portion is kept in balance by actions that cut spending and/or
  raise revenue to prevent on-budget deficits from emerging. This results
  in a unified surplus/deficit equal to the OASDI trust funds' annual
  surplus/deficit through 2034 and equal to the OASDI annual cash deficit
  thereafter.

All three paths assume payment in full of all currently promised OASDI
benefits, including after the projected exhaustion of the OASDI trust
funds in 2034.

Fiscal and economic assumptions underlying the No Action simulation
generally reflect those of CBO's July 1999 baseline./Footnote2/ As shown
in figure 5, the model uses the Trustees' 1999 intermediate estimates for
Social Security and Medicare spending. 

Figure****Helvetica:x11****5:    No Action Model Assumptions

*****************

*****************

Note: In all our work, all CBO budget projections were converted from a
fiscal year to a calendar year basis. The last year of CBO's projection
period is fiscal year 2009, permitting the calculations of calendar year
values through 2008.

Some of the assumptions underlying No Action were not fully realized in
the recently enacted fiscal year 2000 budget,/Footnote3/ and other
assumptions may or may not prove realistic over time. For example,
discretionary spending follows CBO's July 1999 10-year projections, which
assume (1) no emergency spending and (2) actual spending that falls within
the existing discretionary caps through 2002 and then grows with inflation
through 2008. However, CBO reported in December that, including
adjustments to the discretionary caps for fiscal year 2000 enacted
emergency spending, discretionary spending enacted for fiscal year 2000
exceeded the caps.

In simulating the Archer-Shaw and other Social Security reform proposals,
we relied on estimates of OASDI income and cost provided by the Office of
the Chief Actuary at SSA. The OASDI cost estimates for each proposal
reflect all proposed reforms affecting benefits, and the OASDI income
estimates reflect reforms affecting financing elements. For example, for
the Archer-Shaw proposal the OASDI cost estimates reflect the estimated
amounts of transfers from the individual accounts to the OASDI trust
funds; the OASDI income estimates reflect proposed payroll tax reductions. 

As with other reform proposal simulations, on-budget revenue and spending
in the Archer-Shaw simulation reflect the assumptions underlying our base
simulation--the "No Action" or "Save the Surplus" path--adjusted for
reform proposal changes affecting on-budget totals. For example, in
modeling Archer-Shaw, a reduction was made to federal revenues to reflect
the refundable tax credit used to finance individual accounts; this
reduction was also based on data provided by the Office of the Chief
Actuary. 

We did not audit or validate the projections of CBO or the Social Security
and Medicare actuaries. We conducted this work from August through
December 1999 in accordance with generally accepted government auditing
standards.

--------------------------------------
/Footnote1/-^For example, interest rates and total factor productivity
  growth are held constant.
/Footnote2/-^For alternative fiscal policy simulations, certain
  assumptions are varied, which are noted in the discussion of the
  alternative policy paths. CBO will next update its baseline, including
  technical and economic assumptions, in January 2000.
/Footnote3/-^In December 1999, CBO published re-estimates reflecting the
  effect of enacted budget legislation. See CBO's Final Sequestration
  Report for Fiscal Year 2000 and also CBO's The Budget for Fiscal Year
  2000: An End-of-Session Summary.

COMMENTS FROM THE SOCIAL SECURITY ADMINISTRATION
================================================

*****************

*****************

*****************

*****************

RELATED GAO PRODUCTS
====================

Social Security: Evaluating Reform Proposals (GAO/AIMD/HEHS-00-29,
November 4, 1999).

Social Security Reform: Implications of Raising the Retirement Age
(GAO/HEHS-99-112, August 27, 1999).

Social Security: Issues in Comparing Rates of Return With Market
Investments (GAO/HEHS-99-110, August 5, 1999).

Social Security: Implications of Private Annuities for Individual Accounts
(GAO/HEHS-99-160, July 30, 1999).

Social Security: Capital Markets and Educational Issues Associated with
Individual Accounts (GAO/GGD-99-115, June 28, 1999).

Social Security Reform: Administrative Costs for Individual Accounts
Depend on System Design (GAO/HEHS-99-131, June 18, 1999).

Social Security Reform: Implementation Issues for Individual Accounts
(GAO/HEHS-99-122, June 18, 1999).

Social Security: Criteria for Evaluating Social Security Reform Proposals
(GAO/T-HEHS-99-94, March 25, 1999).

Social Security: Individual Accounts as an Element of Long-Term Financing
Reform (GAO/T-HEHS-99-86, March 16, 1999).

Social Security and Surpluses: GAO's Perspective on the President's
Proposals (GAO/T-AIMD/HEHS-99-96, February 23, 1999).

Social Security and Minorities: Current Benefits and Implications for
Reform (GAO/T-HEHS-99-60, February 10, 1999).

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

Social Security Reform: Implications for Women (GAO/T-HEHS-99-52, February
3, 1999).

Budget Issues: Long-Term Fiscal Outlook (GAO/T-AIMD/OCE-98-83, February
25, 1998)

(935341, 207076)

*** End of document. ***