Social Security Financing: Implications of Government Stock Investing for
the Trust Fund, the Federal Budget, and the Economy (Chapter Report,
04/22/98, GAO/AIMD/HEHS-98-74).

Pursuant to a congressional request, GAO reviewed the implications of
having the Social Security trust fund invest in the stock market,
focusing on the impact of such investing on the: (1) Social Security
trust fund; (2) the U.S. economy; and (3) the federal budget.

GAO noted that: (1) allowing the Social Security trust fund to invest in
the stock market is a complex proposal that has potential consequences
for the trust fund, the U.S. economy and federal budget policy; (2) for
the Social Security trust fund, stock investing offers the prospect of
higher returns but greater risk; (3) higher investment returns would
allow the trust fund to pay benefits longer, even without other program
changes; (4) however, if stock investing is implemented in isolation,
the trust fund would inevitably have to liquidate its stock portfolio to
pay promised benefits, and it would be vulnerable to losses in the event
of a general stock market downturn; (5) although stock investing is
unlikely to solve Social Security's long-term financial imbalance, it
could reduce the size of other reforms needed to restore the program's
solvency; (6) beyond the clear tradeoff between greater risk for the
prospect of higher returns, the government would face other
implementation issues inherent in owning and managing a stock portfolio;
(7) proposals for government stock investing typically suggest investing
passively in a broad-based stock index to reduce both costs and the risk
that the government would control individual companies; (8) for the
federal budget, stock investing would have the immediate effect of
increasing the reported unified deficit or decreasing any reported
unified surplus because, under current budget scoring rules, stock
purchases would be treated as outlays; (9) any money used to purchase
stocks would no longer be invested in Treasury securities, reducing the
Treasury's available cash and more clearly revealing the underlying
financial condition of the rest of the government; (10) without
compensating changes in fiscal policy, stock investing would not
significantly alter the impact of federal finances on national saving
and the economy; (11) any higher returns earned by the government would
otherwise have accrued to other investors; (12) in short, by itself
government stock investing would have no appreciable effect on future
national income; (13) although the immediate budgetary effect of
investing in stocks would be to increase unified deficits or decrease
unified surpluses, stock investing might indirectly lead to changes in
federal fiscal policy that could increase national saving; and (14) by
making Social Security's surplus unavailable to the Treasury, stock
investing could focus more attention on the budgetary imbalance that
exists when this temporary source of funds is excluded.

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

 REPORTNUM:  AIMD/HEHS-98-74
     TITLE:  Social Security Financing: Implications of Government Stock 
             Investing for the Trust Fund, the Federal Budget,
             and the Economy
      DATE:  04/22/98
   SUBJECT:  Future budget projections
             Unified budgets
             Stocks (securities)
             Budget deficit
             Social security benefits
             Investments
             Economic analysis
             Fiscal policies
             Financial management
             Trust funds
IDENTIFIER:  Social Security Trust Fund
             
******************************************************************
** 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


Report to the Special Committee on Aging, U.S.  Senate

April 1998

SOCIAL SECURITY FINANCING -
IMPLICATIONS OF GOVERNMENT STOCK
INVESTING FOR THE TRUST FUND, THE
FEDERAL BUDGET, AND THE ECONOMY

GAO/AIMD/HEHS-98-74

Social Security Financing

(935214)


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

  CBO - Congressional Budget Office
  DI - Disability Insurance
  ERISA - Employee Retirement Income Security Act
  GDP - gross domestic product
  OASDI - Old-Age and Survivors' and Disability Insurance
  OASI - Old-Age and Survivors' Insurance
  OMB - Office of Management and Budget
  S&P500 - Standard and Poor's 500
  SEC - Securities and Exchange Commission
  SSA - Social Security Administration
  TSP - Thrift Savings Plan

Letter
=============================================================== LETTER


B-274811

April 22, 1998

The Honorable Charles E.  Grassley
Chairman
The Honorable John B.  Breaux
Ranking Minority Member
Special Committee on Aging
United States Senate

This report responds to your request that we study the implications
of having the Social Security trust fund invest in the stock market. 
As requested, we assessed the impact of government stock investing on
(1) the Social Security trust fund, (2) the U.S.  economy, and (3)
the federal budget. 

As agreed with your office, unless you publicly announce the contents
of this report earlier, we plan no further distribution until 30 days
from the date of this letter.  At that time, we will send copies to
the Chairmen of the House and Senate Committees on the Budget, the
Senate Committee on Finance, and the House Committee on Ways and
Means; the Director of the Office of Management and Budget; the
Secretary of the Treasury; the Commissioner of the Social Security
Administration; the Chairman of the Securities and Exchange
Commission; and other interested parties.  Copies will also be made
available to others upon request. 

Please contact Mr.  Posner at (202) 512-9573 or Ms.  Bovbjerg at
(202) 512-5491 if you have any questions concerning this report. 

Paul L.  Posner
Director, Budget Issues
Accounting and Information Management Division

Barbara D.  Bovbjerg
Associate Director, Income Security Issues
Health, Education, and Human Services Division


EXECUTIVE SUMMARY
============================================================ Chapter 0


   PURPOSE
---------------------------------------------------------- Chapter 0:1

The Social Security program faces a long-term financing challenge,
primarily due to changing demographics.  In response, reform
advocates have suggested numerous options to curb benefits or
increase revenues.  One proposed option is to invest Social Security
funds in the stock market with the intention of earning higher
returns.  To better understand the potential implications of stock
investing for the federal government, the Senate Special Committee on
Aging asked GAO to address the impact of government stock investing
on (1) the investment earnings, investment risk, and financial
solvency of the Social Security trust fund, (2) national saving and
the financial markets, including implementation issues presented by
the government selecting and managing its stock portfolio, and (3)
the federal budget and federal debt. 

This report addresses only indirectly the broad issue of Social
Security's long-term financing needs and does not evaluate any
specific Social Security reform package or address proposals to
establish individually-owned accounts.  Rather, as discussed with
Committee staff, to fully identify the effects of government stock
investing, GAO studied changing trust fund investment policy in
isolation from any other program changes in Social Security.  In its
analysis, GAO generally reviewed the literature on financial markets,
Social Security, and federal budgeting and interviewed finance,
budget, and program experts in government and the private sector.  To
illustrate how alternative policies could affect the trust fund's
financial outlook, the Social Security Administration's (SSA) Office
of the Chief Actuary, at GAO's request, simulated the potential
outcomes of two stock investment scenarios using the Social Security
Trustees' 1997 intermediate actuarial assumptions. 


   BACKGROUND
---------------------------------------------------------- Chapter 0:2

Social Security's long-term financing problem is caused primarily by
the aging of the U.S.  population.  According to Social Security's
actuarial estimates, the number of workers supporting each Social
Security beneficiary is projected to drop from 3.3 to 2.0 between
1997 and 2030--a decline of nearly 40 percent.  Beyond 2030, the
proportion of the population that is elderly is expected to continue
growing due to relatively low birth rates and increasing longevity. 

Currently, Social Security's tax revenue each year exceeds benefit
payments, and the trust fund, by law, invests the resulting cash
surplus in U.S.  government obligations.  The trust fund also
receives interest income from the Treasury that is credited in the
form of additional Treasury securities.  Beginning in about 2012,
according to Social Security actuarial projections, the program's
annual tax revenue will be insufficient to pay yearly benefits.  To
cover the cash shortfall, the trust fund will begin drawing on the
Treasury, first relying on its interest income and eventually drawing
down its assets.  Regardless of whether the trust fund is drawing on
interest income or principal to make benefit payments, the Treasury
will need to raise the required cash through some combination of
borrowing from the public,\1 spending cuts in other federal programs,
or revenue increases.  In 2029, the trust fund is projected to be
exhausted.  After that, Social Security taxes are projected to cover
only about 75 percent of promised benefits. 

In response to Social Security's long-range financing problems, the
1994-1996 Advisory Council on Social Security (the Advisory Council)
suggested a variety of specific changes.  One proposal advanced by 6
of the Advisory Council's 13 members suggested that along with a
number of benefit and tax changes, policymakers consider allowing the
trust fund to invest in stocks.  Supporters of this approach have
pointed out that investing in the stock market is standard practice
for state and local government and private sector pension funds. 
Overall, pension funds held about 60 percent of their assets in
stocks and these holdings accounted for about a quarter of total U.S. 
stock holdings--valued at $12.8 trillion--at the end of the third
quarter of 1997. 

The seven other members of the Advisory Council opposed government
investment in the stock market.  They believed that there would be
tremendous political pressures to steer the Social Security trust
fund's investments to achieve other economic, social, or political
purposes rather than basing decisions solely on the expected risk and
return.  Also, there was concern that if the government exercised its
stock voting rights, it might influence individual companies or
industries. 

To assess the implications of changing Social Security's investment
policy, it is important to understand how Social Security fits within
the federal budget and its influence on overall fiscal policy. 
Within the federal budget, Social Security is a trust fund account
that authorizes the Treasury to pay Social Security benefits as long
as the account has a balance.  Social Security is the largest federal
trust fund with fiscal year 1997 outlays of $365 billion and a fund
balance of $631 billion invested in Treasury securities.  While the
trust fund's Treasury securities represent assets for Social
Security, they are future claims against the Treasury.  Today, Social
Security's surplus tax revenue is invested in Treasury securities and
is spent to finance other governmental activities, thereby reducing
the Treasury's need to borrow from the public.  Although the Social
Security trust fund is technically excluded from the budget, its
finances contribute to the government's impact on the economy. 
Therefore, Social Security is included, along with all other federal
programs, in the commonly used "unified" budget measure.  The unified
budget is the means to measure the government's current draw on
financial markets.  However, in considering the long-range
implications of federal policies, it is also useful to consider the
impact that Social Security's temporary surplus has on the
government's unified budget.  Social Security's current cash surplus
partially offsets the deficit in the rest of the government's
accounts.\2

Social Security has an important influence on the government's
overall fiscal position, which, in turn, affects national saving, a
key determinant of long-term economic growth.  The nation's saving is
composed of the private saving of individuals and businesses and the
saving or dissaving of all levels of government.  In general,
government budget deficits subtract from national saving by absorbing
funds that otherwise could be used for private investment, while
government surpluses add to saving.  Raising saving and investment
levels would improve the long-term productivity of the economy,
thereby boosting economic growth.  The most direct way for the
federal government to contribute to national saving and long-term
economic growth is to achieve and maintain a balanced budget or a
surplus.  A bigger economic pie would make it easier for future
workers to meet the dual challenges of paying for the baby boomers'
retirement while achieving a rising standard of living for
themselves. 


--------------------
\1 If the unified budget were in surplus, then financing the excess
benefits would require less debt redemption, rather than increased
borrowing. 

\2 Interest credited on the trust fund's Treasury securities has no
current effect on the unified federal deficit because it is a payment
from one part of the government to another part. 


   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3

Allowing the Social Security trust fund to invest in the stock market
is a complex proposal that has potential consequences for the trust
fund, the U.S.  economy, and federal budget policy.  For the Social
Security trust fund, stock investing offers the prospect of higher
returns but greater risk.  Higher investment returns would allow the
trust fund to pay benefits longer, even without other program
changes.  However, if stock investing is implemented in isolation,
the trust fund would inevitably have to liquidate its stock portfolio
to pay promised benefits, and it would be vulnerable to losses in the
event of a general stock market downturn.  Although stock investing
is unlikely to solve Social Security's long-term financial imbalance,
it could reduce the size of other reforms needed to restore the
program's solvency. 

Beyond the clear trade-off between greater risk for the prospect of
higher returns, the government would face other implementation issues
inherent in owning and managing a stock portfolio.  Proposals for
government stock investing typically suggest investing passively in a
broad-based stock index to reduce both costs and the risk that the
government would control individual companies.  Index investing would
help achieve these goals, and it would reduce but not eliminate the
possibility of political influence over stock selections.  However,
because index investors cannot alter their portfolio composition to
increase financial performance, the government would have a stronger
incentive to actively exercise the voting rights of its sizable stock
portfolio.  This issue would raise concerns about potential federal
involvement in corporate affairs, and could prove more difficult to
resolve since even choosing not to vote would affect corporate
decision-making by enhancing the voting power of other shareholders
or investment managers. 

For the federal budget, stock investing would have the immediate
effect of increasing the reported unified deficit or decreasing any
reported unified surplus because, under current budget scoring rules,
stock purchases would be treated as outlays.  Any money used to
purchase stocks would no longer be invested in Treasury securities,
reducing the Treasury's available cash and more clearly revealing the
underlying financial condition of the rest of the government.  If
after accounting for this effect the federal government were in a
deficit, the Treasury would need to borrow from the public to replace
cash used to buy stocks unless offsetting spending or revenue actions
were taken. 

Without compensating changes in fiscal policy, stock investing would
not significantly alter the impact of federal finances on national
saving and the economy.  To cover a deficit, the government would
issue additional Treasury securities to the public, but it would
offset this action by purchasing stocks from other investors.  This
asset shuffle could lead to higher stock prices and higher interest
rates, undercutting somewhat the potential gain from stock investing
and increasing the government's cost of borrowing.  Even with such
price effects, the government could still come out ahead by investing
in stocks.  However, any higher returns earned by the government
would otherwise have accrued to other investors.  In short, by
itself, government stock investing would have no appreciable effect
on future national income. 

Although the immediate budgetary effect of investing in stocks would
be to increase unified deficits or decrease unified surpluses, stock
investing might indirectly lead to changes in federal fiscal policy
that could increase national saving.  By making Social Security's
surplus unavailable to the Treasury, stock investing could focus more
attention on the budgetary imbalance that exists when this temporary
source of funds is excluded.  Increased attention could prompt
policymakers to address this imbalance by cutting spending or raising
revenue.  Such fiscal actions could boost national saving and
long-term economic growth.  Of course, regardless of any change in
Social Security's investment policy, policymakers could decide to
achieve a budget balance or surplus without relying on Social
Security's surplus. 


   GAO'S ANALYSIS
---------------------------------------------------------- Chapter 0:4


      POTENTIAL RETURNS AND RISKS
      FOR THE SOCIAL SECURITY
      TRUST FUND
-------------------------------------------------------- Chapter 0:4.1

The Social Security trust fund could expect to earn more by investing
in the stock market but would have to accept greater risk.  Under
current law, the trust fund invests solely in U.S.  government
obligations and, under the Social Security Trustees' intermediate
assumptions, is expected to earn 2.7 percent after inflation over the
long term.  Historically, returns on stocks have exceeded returns on
Treasury securities over the long term, averaging about 7 percent
after inflation.  However, stock returns are highly variable from
year to year, and there have been years in the past with negative
returns. 

To illustrate how much the trust fund might invest in the stock
market, GAO developed (1) an aggressive scenario of investing Social
Security's future annual cash surplus and interest, while maintaining
a contingency reserve of special Treasury securities equal to at
least a year's expenditures, and (2) a more conservative scenario of
investing only Social Security's cash surplus and selling stocks
first to finance Social Security's expected cash deficit.  Under
these scenarios, Social Security's cash surplus would not be
available to finance other government operations.  Under the
aggressive scenario, the Treasury also would have to raise additional
cash to finance interest payments to the trust fund. 

Under the aggressive scenario, assuming the historical average stock
return, the trust fund's exhaustion could possibly be delayed by
about a decade, from 2029 to 2040.  This potential delay well into
the baby boomers' retirement years would result only from the Social
Security trust fund investing aggressively in the stock market.  The
trust fund would invest more than 70 percent of its assets in the
stock market, which would be a dramatic shift from investing solely
in Treasury securities.  Under the cash surplus scenario, still
assuming the historical average return, the possible delay in the
trust fund's exhaustion would be only 3 years. 

The possible delay resulting from any stock investment scenario would
be significantly shorter if the future stock returns are lower than
the historical average of 7 percent after inflation.  As an
illustration, if the future return on stocks is 1 percentage point
lower, the delay in the trust fund's exhaustion under the aggressive
scenario would be reduced to only 6 years.  The delay under the cash
surplus scenario assuming the real return is 1 percentage point lower
would be 2 years.  The results of these simulations illustrate some
outcomes associated with two stock investment alternatives; they
should not be interpreted as forecasts and are not intended to
represent the full range of possible outcomes for the Social Security
trust fund. 

The only way for the Social Security trust fund to earn the higher
returns possible with stock investing is to take on greater risk. 
The primary risk that the trust fund would face is the possibility of
loss in the event of a general stock market downturn (market risk). 
Depending on the composition of its portfolio, the trust fund could
also face losses from a heavy investment in a group of companies or
an industry susceptible to the same economic dynamics (concentration
risk).  Diversifying the stock portfolio, for example by investing in
an index representative of the broad market, would reduce the risk of
loss associated with individual companies or an industry segment
performing poorly. 

If stock investing is implemented in isolation from other program
changes, the trust fund would face the certain need to liquidate its
assets to pay benefits with no certainty about what future stock
prices would be or whether it could recover amounts invested.  The
more the trust fund is counting on stock sales to raise cash, the
greater its vulnerability in the event of a general market downturn. 
Stock investing conceivably could be implemented in combination with
other changes that increase Social Security's funding.  As part of a
broader package, which is what the Advisory Council suggested, stock
investing could complement traditional reforms in that higher stock
returns could reduce the size of benefit cuts and/or tax increases. 


      IMPACT ON ECONOMY AND
      FINANCIAL MARKETS
-------------------------------------------------------- Chapter 0:4.2

The economic effect of government stock investing would likely be
minimal because stock investing by itself does not increase national
saving.  In the absence of further deficit reduction, the Treasury
would have to borrow from the public to offset the Social Security's
stock purchases, yielding no additional national saving.\3 The net
effect would be that private investors would end up with fewer stocks
and more Treasury securities, while the government would have stocks
and fewer of its own securities.  This asset shuffle means that any
higher returns earned by the government would otherwise have accrued
to other investors.  In short, by itself, government stock investing
would have no appreciable effect on future economic growth. 

Financial market analysts and economists believe that this asset
shuffling between the government and other investors could increase
stock prices and interest rates, undercutting the potential gain from
stock investing and increasing the government's cost of borrowing. 
Price effects could begin even before the government announces its
stock investment policy, reducing the potential gain to the
government from stock investing.  Higher interest rates would
increase the cost of government borrowing, even as higher stock
prices would reduce the government's expected return on its stock
investments.  The magnitude of price changes is uncertain and could
be small.  Long-term changes in asset prices are unpredictable and
would depend, in part, on the response of other borrowers and
investors to the short-term price effects of government stock
investing.  Although Social Security's stock portfolio would likely
be larger than that of any other single investor, its size in
comparison to the whole stock market would likely be relatively
small. 


--------------------
\3 This statement applies in a situation where the federal government
has a unified budget deficit after investing in the stock market. 
If, instead, the government ended up with a unified surplus, the cash
used to purchase stocks would not be available to reduce the level of
debt held by the public. 


      BENEFITS AND LIMITATIONS OF
      STOCK INDEXING
-------------------------------------------------------- Chapter 0:4.3

Proposals for government stock investing typically recommend
investing in a broad-based stock index to diversify the government's
stock portfolio, reducing the likelihood of concentrating investments
in individual companies, and to reduce administrative costs.  Still,
index investing could create price effects of its own as the
government would have to purchase new stocks added to an index or
sell stocks deleted from the selected index.  Analysts have found
significant price changes in stock prices of companies added to or
deleted from the Standard & Poor's 500 index, which represents about
two-thirds of the U.S.  stock market. 

Index investing does not eliminate the possibility that nonfinancial
objectives would influence stock selection.  For example, the
government could start with a broad-based index and modify it to
target investments that offer competitive returns and also achieve
social goals.  Alternatively, the government could choose to
disinvest in specific companies or sectors of the economy that are in
conflict with government policies.  Critics have expressed concerns
that pressures to include or exclude stocks for nonfinancial reasons
might reduce the rate of return on the government's portfolio and
hinder the overall economic efficiency of capital markets.  Some
analysts believe that it might be possible to establish in law that
investments be made solely for the financial benefit of Social
Security participants and not for other social objectives.  They cite
as an example the federal Thrift Savings Plan, which is managed
solely in the interest of participating federal employees and their
beneficiaries. 

Regardless of the type of indexing strategy the government adopts,
policymakers would need to decide how to handle the stock voting
rights for the government's portfolio.  Because index investors as a
general rule do not alter the portfolio composition to increase
financial performance, they have a stronger incentive to exercise
stock voting rights actively.  Instead of selling stocks of an
underperforming company, an index investor can choose to participate
in corporate decisions that might enhance the company's performance. 
However, critics have expressed concern that the government's right
to vote its sizable number of shares would allow it to influence
corporate decisions.  To blunt such concerns, the government's stock
voting rights could be restricted by statute, but any restriction
would need to be designed carefully.  For example, simply prohibiting
the government from exercising its voting rights would favor other
stockholders or investment managers by effectively increasing their
voting rights. 


      EFFECTS OF GOVERNMENT STOCK
      INVESTMENTS ON THE FEDERAL
      BUDGET AND FISCAL POLICY
-------------------------------------------------------- Chapter 0:4.4

In the short term, stock investing would increase the reported
unified deficit or decrease any unified surplus because, under
current budget scoring rules, stock purchases would be treated as
budget outlays.  The magnitude of the change in the reported
deficit/surplus could exceed $100 billion annually, depending on how
much the trust fund invested in stocks.  If after accounting for this
effect the government were in deficit, the Treasury would have to
borrow more from the public, unless action were taken to reduce other
spending or raise revenues. 

Any increased borrowing from the public would be offset by reduced
borrowing from the Social Security trust fund, leaving the federal
government's gross debt largely unchanged.  More important, the
reported decline in the government's budget balance caused by stock
investing would not significantly affect national saving.  While any
federal borrowing from the public would absorb money from capital
markets, the trust fund's stock investments would offset this effect
by adding money to the markets.  As a result, the government's
fundamental fiscal position would be largely unchanged. 

While stock investing would have a negative effect on the budget
deficit/surplus today, over time its impact on the unified budget
could largely be neutral.  As with any budget outlay, the purchase of
a stock would mean money flowing out of the government and, thus, the
reported budget deficit/surplus would deteriorate.  However, the sale
of a stock would mean money flowing back into the government.  So,
when Social Security begins running cash deficits in the future, it
could sell stocks to finance benefits, rather than drawing on the
Treasury.  This approach would result in smaller future budget
deficits or larger future budget surpluses than under current policy. 
This longer term improvement could offset the near-term deterioration
in the deficit/surplus.  Indeed, on balance, the government could
actually come out ahead in the long term if its stock earnings were
to exceed any increase in federal borrowing costs that might result
from a stock investing policy.  However, without any accompanying
action to raise national saving, the government would be capturing a
portion of stock returns that would otherwise have accrued to private
investors. 

Whether the short-term increase in reported budget deficits or the
decline in reported budget surpluses would lead to any changes in
fiscal policy is unclear.  Since stock investing would not
substantially change the impact of federal finances on the economy,
policymakers might decide to maintain fiscal policy as is.  Since the
government acquires a financial asset when it buys stock, the case
could be made that the purchase should not be treated as a budget
outlay.  Policymakers could choose to change budget scoring rules to
explicitly recognize the distinct nature of stock purchases. 
However, such a change would conflict with the way other asset
purchases are treated in the budget.  Generally, asset purchases are
scored as outlays.\4 In addition, creating different budget scoring
rules for stocks would also raise some complicated technical issues,
such as how to recognize changes in their market values.  If, despite
these considerations, stock purchases were not treated as outlays,
stock investing would have no major impact on the reported budget
deficit or surplus.\5

If current budget scoring rules were maintained, stock investing
would make more visible the underlying condition of the government's
finances excluding the Social Security surplus.  Currently, in the
unified budget presentation, the Social Security surplus masks the
financial status of the rest of the government.  By helping to
finance current spending, Social Security's cash surplus may result
in the government spending more or taxing less than it would if these
funds were not available to finance other programs.\6

By making at least part of Social Security's surplus unavailable to
the Treasury, stock investing would reduce or eliminate the masking
effect.  In effect, stock investing would make the unified budget
measure look more like the "on-budget" measure that excludes Social
Security's finances.  If Social Security's cash surplus and its
interest were invested in stocks, the "new" unified budget measure
would virtually match the on-budget measure.\7

Even though stock investing does not change the government's fiscal
position, the higher reported deficits or lower reported surpluses
that result could indirectly lead policymakers to change fiscal
policy by focusing more attention on the budget imbalance that exists
when Social Security's surplus is excluded.  Policymakers could react
to a higher unified deficit by cutting spending and/or raising taxes. 
Or, if stock investing were expected to reduce or eliminate a unified
budget surplus, instead of creating or adding to a unified budget
deficit, policymakers might be reluctant to enact tax cuts or
additional spending.  In this case, fiscal restraint might not
promote higher saving, but it would avoid policy actions that could
cause saving to decline. 

Though stock investing could help highlight the budget shortfall that
exists when Social Security's surplus is excluded, it represents a
circuitous way of essentially duplicating an existing measure--the
on-budget deficit.  If policymakers wanted to take actions to boost
national saving, they certainly could do so directly by running
annual surpluses in the unified budget and devoting the surplus funds
to reducing the level of outstanding debt held by the public.  If the
government ran a unified budget surplus equal to Social Security's
cash surplus, the Treasury would no longer need to rely on Social
Security revenues to finance federal spending on other activities. 
While attaining and sustaining surpluses could prove extremely
challenging, such a policy would strengthen the fiscal position of
the government and, by promoting higher saving, better position the
economy to handle the baby boomers' retirement costs. 


--------------------
\4 Major exceptions include debt transactions, such as those that
occur when Social Security invests funds in special Treasury
securities, and cases in which the government purchases an asset that
is considered equivalent to cash. 

\5 Even if stock purchases were not treated as outlays, stock
investing could result in minor changes in the budget deficit/surplus
(e.g., interest costs on any additional borrowing from the public). 

\6 This masking effect could have important implications for any
Social Security reforms.  Reforms that increase the size of the
Social Security surplus would not necessarily improve the long-term
picture for the budget as a whole.  A larger Social Security surplus
might serve to intensify the masking of the financial condition of
the rest of the government.  If policymakers responded to a larger
trust fund surplus by exercising less restraint in the rest of the
budget, the improvement in Social Security's finances would not
contribute to increased national saving.  Instead, it would only
allow the trust fund to build up more claims on the Treasury without
enhancing the nation's ability to meet these future claims. 

\7 In addition to Social Security, the on-budget deficit excludes the
operations of the U.S.  Postal Service.  However, this amount is very
small in comparison to Social Security. 


   AGENCY COMMENTS
---------------------------------------------------------- Chapter 0:5

GAO requested comments on a draft of this report from the
Commissioner of the Social Security Administration, the Secretary of
the Treasury, the Chairman of the Securities and Exchange Commission,
and the Director of the Office of Management and Budget or their
designees.  These agencies provided technical comments from their
staffs, which were incorporated where appropriate. 


INTRODUCTION
============================================================ Chapter 1

The Social Security program faces a long-term financing challenge
primarily due to changing demographics.  In response, numerous
options have surfaced to reduce benefits or increase revenues.  One
proposed option is to invest a portion of the Social Security trust
fund\1 in the stock market with the intention of earning higher
returns.  To better understand the potential implications of the
federal government investing in the stock market, the Chairman and
Ranking Member of the Senate Special Committee on Aging asked us to
analyze how such a proposal would affect the Social Security trust
fund, the U.S.  economy, and the federal budget. 


--------------------
\1 Social Security consists of two separate trust fund accounts: 
Old-Age and Survivors Insurance (OASI), which funds retirement and
survivor benefits, and Disability Insurance (DI), which provides
benefits to disabled workers and their families.  These two accounts
are commonly combined in discussing the Social Security program.  For
the purposes of this report, any reference to the Social Security
trust fund refers to the combined OASDI trust funds. 


   SOCIAL SECURITY'S FINANCING
   CHALLENGE PRIMARILY CAUSED BY
   DEMOGRAPHIC TRENDS
---------------------------------------------------------- Chapter 1:1

Social Security's long-term financing problem is primarily caused by
the aging of the U.S.  population.  As the baby boom generation
retires, labor force growth is expected to slow dramatically. 
According to Social Security's intermediate actuarial projections,
the number of workers supporting each Social Security beneficiary is
projected to drop from 3.3 to 2.0 between 1997 and 2030--a decline of
nearly 40 percent.\2 Beyond 2030, the population is expected to
continue aging due to relatively low birth rates and increasing
longevity.  These demographic trends will require substantial changes
in Social Security benefits and/or revenues.  Without such changes,
Social Security tax revenues are expected to be insufficient to cover
benefit payments in about 2012, less than 15 years from now.  To
cover this annual cash shortfall, the trust fund will begin drawing
on the Treasury, first relying on its interest income and,
eventually, on its assets.\3 As shown in figure 1.1, the trust fund's
annual cash deficits will grow rapidly, reaching a projected 1.6
percent of the nation's gross domestic product (GDP) in 2028.  The
trust fund is projected to be exhausted by 2029 and, after this
point, Social Security's annual tax revenues will cover only about 75
percent of expected benefit payments.  In short, the program as
currently structured is unsustainable.\4

   Figure 1.1:  Social Security's
   Projected Cash Surpluses and
   Deficits as a Percentage of GDP
   (1997-2029)

   (See figure in printed
   edition.)

Note:  Calendar year data. 

Source:  GAO analysis of 1997 Trustees' Report, intermediate
assumptions. 

These trends in Social Security's finances will place a significant
burden on future workers and the economy.  Without major policy
changes, the relatively smaller workforce of tomorrow will bear the
brunt of tax increases, spending cuts, or borrowing\5

needed to finance Social Security's cash deficit.  In addition, the
future workforce also would likely be affected by any reduction in
Social Security benefits or increased payroll taxes needed to resolve
the program's long-term financing shortfall. 


--------------------
\2 Throughout this report, we relied on data from The 1997 Annual
Report of the Board of Trustees of the Federal Old-Age and Survivors
Insurance and Disability Insurance Trust Funds, hereafter "the 1997
Trustees' Report." Under the Social Security Act, the Board of
Trustees is required to report annually to the Congress on Social
Security's financial and actuarial status.  We used the intermediate
assumptions, which reflect the Board of Trustees' best estimate.  Due
to the inherent uncertainty surrounding long-term projections, the
Trustees' report also includes two other sets of assumptions, a high
cost and a low cost alternative. 

\3 Regardless of whether the trust fund is drawing on its interest or
principal to pay benefits, the Treasury will need to raise the
required cash. 

\4 Retirement Income:  Implications of Demographic Trends for Social
Security and Pension Reform (GAO/HEHS-97-81, July 11, 1997). 

\5 If the unified budget were in surplus, then financing Social
Security's cash deficit would result in less debt redemption rather
than requiring increased borrowing. 


   STOCK INVESTING SUGGESTED TO
   HELP SOLVE SOCIAL SECURITY'S
   FINANCING PROBLEM
---------------------------------------------------------- Chapter 1:2

In its report to the Social Security Commissioner, the 1994-1996
Advisory Council on Social Security ("the Advisory Council") offered
three alternative reform proposals.  While the Advisory Council could
not agree on a single plan for dealing with Social Security's
difficulties, the members did agree that investing in the stock
market would yield higher returns for financing retirement benefits. 
One approach was to maintain Social Security's current benefit and
taxation structure and allow the government to invest directly in the
stock market.  The other two approaches would significantly
restructure Social Security and allow stock investing through
accounts owned and controlled by individuals. 

The "Maintain Benefits" plan--supported by 6 of the Advisory
Council's 13 members--recommended a traditional package of tax and
benefit changes.  These changes were projected to solve about
two-thirds of Social Security's long-term financing problem.  To
close the remaining gap, the plan called for studying the possibility
of allowing the Social Security trust fund to invest in the stock
market.  According to the plan's estimates, gradually investing up to
40 percent of the Social Security trust fund in stocks could obviate
the need for further tax increases or benefit cuts.  The plan also
outlined that the government's portfolio would be passively managed
and selected using a broad market index ("indexing"). 

The seven other Advisory Council members opposed government
investment in the stock market.  They believed that there would be
tremendous political pressures to steer the Social Security trust
fund's investments to achieve other economic, social, or political
purposes rather than basing decisions solely on the trust fund's
expected risk and return.  Even the Maintain Benefits proponents were
concerned about how to handle the government's stock voting rights so
as to neutralize the potential effect on individual companies or
industries. 


      PENSION FUNDS INVEST IN A
      MIX OF ASSETS
-------------------------------------------------------- Chapter 1:2.1

The Advisory Council's Maintain Benefits plan pointed out that
although investing in the stock market would be a new concept for
Social Security, it is a standard practice for public and private
sector pension funds.  Pension funds invest in a wide mix of assets,
often placing a majority of their funds in the stock market.  Pension
funds accumulate substantial assets from contributions by employers
and employees to finance future retirement benefits.  Investment
earnings on these funds contribute considerable revenue and reduce
the amount of money that would otherwise have to be contributed to
pay pension benefits.  In determining how to allocate their
portfolios among different types of assets, pension funds must
balance the trade-off between earning a good long-term return and
minimizing the risk of loss.  As of the third quarter of 1997,
private sector pension funds had $3.5 trillion in assets, and state
and local pension funds had $2 trillion.\6

Overall, pension funds held about $3.4 trillion in stocks, or about
60 percent of their total assets.  Pension funds own about a quarter
of total U.S.  stock holdings, valued at $12.8 trillion at the end of
the third quarter of 1997.\7 U.S.  government securities accounted
for about 15 percent of pension funds' holdings.  Pension funds also
invest in mortgages, real estate, and other assets, including venture
capital.  Table 1.1 illustrates the mix of assets held by private
sector as well as state and local government pension funds as of the
third quarter of 1997. 



                               Table 1.1
                
                 Asset Mix of Private Sector and State
                   and Local Government Pension Funds

                             (Percentages)

                                                       State and local
Asset type                          Private sector         governments
------------------------------  ------------------  ------------------
U.S. and foreign stocks                         62                  61
U.S. government securities\a                    12                  15
Corporate and foreign bonds\b                    9                   9
Cash and short-term                              5                   4
 instruments
Other investments                             12\c                  11
----------------------------------------------------------------------
\a Includes U.S.  Treasury and agency securities. 

\b Municipal securities represent less than 0.5 percent of pension
assets.  Pension funds as a whole do not invest in municipal
securities issued by state and local governments.  Most municipal
securities are exempt from federal income taxation, and their yield
is lower than that of Treasury securities with the same maturity. 

\c Includes 7 percent in contracts with private insurance companies. 

Source:  Flow of Funds Accounts of the United States, Federal Reserve
statistical release for the third quarter of 1997, tables L.119 and
L.120, p.76. 

Although Social Security is not a pension fund, the experience of
public pension funds may yield some insight into the implications of
the federal government allowing the Social Security trust fund to
invest in the stock market.\8 A public pension fund operates under a
legal framework established by its sponsoring government.\9 A
government may choose to restrict its public pension fund from
investing heavily in risky assets.  Also, a government may use public
pension investments to further other policy objectives.  Moreover,
some state and local pension plans actively participate in the
management decisions of the companies in which they invest by
exercising their shareholder voting rights.\10

The Advisory Council's Maintain Benefits plan pointed out that, in
fact, a few federal pension plans invest in assets other than
Treasury securities.  Most notably, the federal employees' Thrift
Savings Plan (TSP) has passively-managed indexed stock investments
similar to those that have been proposed for Social Security. 
However, in one major aspect, the Thrift Savings Plan is not
comparable to the Social Security trust fund--federal employees, not
the government, own and control the investment of their individual
account balances.  For this reason, the transactions of TSP are not
included in the unified federal budget. 

There are other limits to comparing the Social Security trust fund to
pension funds.  Although Social Security resembles a traditional
pension plan in that retirement benefits are based on earnings and
work time, the Congress has the ability to change the legislation
governing benefits at any time.  Social Security is not just a
retirement program--it is also a social insurance program with
additional goals for income redistribution and protection for
survivors, dependents, and disabled workers.  Moreover, Social
Security is financed largely on a pay-as-you-go basis, whereas
pension plans are generally funded in advance.\11 Finally, the sheer
size and scope of Social Security dwarfs U.S.  pension funds. 

Given that Social Security is national in scope, the experiences of
central governments in other nations may yield insights.  As
described in figure 1.2, Canada is implementing a new investment plan
allowing its national program that provides retirement and disability
benefits to invest in stocks. 

   Figure 1.2:  Canadian Pension
   Plan to Invest in Stocks

   (See figure in printed
   edition.)


--------------------
\6 Flow of Funds Accounts of the United States, Federal Reserve
statistical release for the third quarter of 1997, tables L.119 and
L.120, p.  76. 

\7 Flow of Funds Accounts of the United States, Federal Reserve
statistical release for the third quarter of 1997, table L.213, p. 
90. 

\8 For further information, see (1) Paul Zorn, 1995 Survey of State
and Local Government Employee Retirement Systems (The Public Pension
Coordinating Council, 1996), (2) Catherine Baker Knoll, "The Knoll
Survey of State Public Pension Funds Investment Policies and
Practices," presented at the National Association of State Treasurers
Public Pension Fund Conference (1995), (3) Olivia S.  Mitchell and
Ping Lung Hsin, "Public Pension Governance and Performance" (National
Bureau of Economic Research Working Paper No.  4632, January 1994),
and (4) Girard Miller, Pension Fund Investing (Chicago, Illinois: 
Government Finance Officers Association, 1987). 

\9 See Cynthia L.  Moore, Protecting Retirees' Money:  Fiduciary
Duties and Other Laws Applicable to State Retirement Systems, 3rd ed. 
(National Council on Teacher Retirement, 1995). 

\10 See Roberta Romano, "Public Pension Fund Activism in Corporate
Governance Reconsidered," Columbia Law Review, Vol.  93, No.  4
(1993), pp.  795-853. 

\11 The Employee Retirement Income Security Act of 1974 (ERISA)
requires private sector pension plans to be funded on a sound
actuarial basis.  ERISA does not apply to state and local pension
plans; their funding requirements are set by their sponsoring
governments.  See Public Pensions:  State and Local Government
Contributions to Underfunded Plans (GAO/HEHS-96-56, March 14, 1996)
for a discussion of funding of these plans. 


   OBJECTIVES, SCOPE, AND
   METHODOLOGY
---------------------------------------------------------- Chapter 1:3

To better understand the implications of the Social Security trust
fund investing in the stock market and of the federal government
owning and managing a portfolio of stocks, we considered three
different perspectives:  the trust fund, the U.S.  economy, and the
federal budget.  At the request of the Senate Special Committee on
Aging, we specifically addressed the impact of government stock
investing on (1) the investment earnings, investment risks, and
financial solvency of the Social Security trust fund, (2) national
saving and the financial markets, including implementation issues
presented by the government selecting and managing its stock
portfolio, and (3) the federal budget and federal debt. 

We reviewed the final report of the 1994-1996 Advisory Council on
Social Security as well as the technical reports and presentations to
the Advisory Council.\12 Our report, however, neither evaluates the
Advisory Council's specific government investment proposal nor
addresses its other proposals to establish individually-owned
accounts.  To fully understand the effects of government investing in
stocks, we also studied such a change in trust fund investment policy
in isolation from any other program changes in Social Security. 
Therefore, our report is limited to an analysis of altering Social
Security's investment policy and only indirectly addresses the
broader issue of the program's long-term financing needs. 

For the trust fund analysis, we concentrated on the potential returns
and risks of stock market investing for Social Security in isolation
from other program changes.  We researched economic and finance
literature and interviewed market experts in both academia and the
private sector to identify historical stock return data and major
trends that could affect future stock performance.  Also, we
identified the general risks inherent in stock market investing and
attempted to distinguish risks unique to the Social Security trust
fund investing in the stock market in isolation from other program
changes.  As a point of comparison, we also examined pension
literature and research documenting the general investment practices
of public pension plans, but a detailed discussion of those practices
was beyond the scope of this report. 

We simulated two stock investment scenarios to illustrate how
changing the investment policy can affect the future outcome for the
Social Security trust fund.  We developed an aggressive scenario
where the trust fund would maintain a reserve of Treasury securities
equal to at least 100 percent of annual expected expenditures and
invest its future annual cash surplus and interest earned on its
Treasury securities in the stock market.  We also tested a more
conservative scenario investing only Social Security's annual cash
surplus in the stock market.  The simulations use the historical
average real return on stocks that the Advisory Council assumed in
estimating future stock performance as well as its assumption about
administrative costs on the trust fund's investments.  At our
request, the Social Security Administration's (SSA) Office of the
Chief Actuary simulated the potential effect on the trust fund of
these two investment scenarios using the Social Security Trustees'
1997 intermediate assumptions about future program revenues and
expenditures as well as their demographic and economic assumptions. 
We did not audit or validate SSA's actuarial projections.  For
details of the simulation assumptions, see appendix I. 

Simulations are useful for comparing alternative investment policies
within a common framework; however, they should not be interpreted as
forecasts of future stock performance.  In light of the uncertainty
about future stock returns, we also tested a stock return that is 1
percentage point lower than the historical average.  This alternative
return is intended only to demonstrate that stock investment
simulation results are sensitive to the rate of return assumed.  The
results of our simulations are not intended to represent the full
range of expected future returns on stocks or possible outcomes for
the Social Security trust fund. 

For the market analysis, we considered (1) the potential impact of
government stock investing on national saving as well as the
financial markets, (2) the benefits and limitations of an indexing
investment strategy typically proposed for government stock
investing, and (3) implementation issues the government would face in
selecting and managing its stock portfolio.  We also reviewed
economic and finance literature pertaining to saving, asset prices,
and market behavior in general.  To identify specific market effects
that could result from government stock investing and other
implementation issues, we interviewed officials at the Department of
the Treasury, the Securities and Exchange Commission (SEC), and the
Federal Retirement Thrift Investment Board as well as economists and
finance experts in academia and the private sector.  We reviewed
records of congressional hearings, studies, and articles pertaining
to the concept of the government investing in private financial
markets. 

For the budget perspective, we considered the short- and long-term
fiscal effects of stock market investing for the federal budget and
federal debt.  To address these questions, we surveyed the literature
on Social Security and the federal budget and interviewed Social
Security officials and federal budget experts from the Office of
Management and Budget (OMB), the Congressional Budget Office, and the
Congressional Research Service, as well as private organizations. 

While this report focuses on stocks, there are other investment
options beyond Treasury securities.  A number of policymakers and
Social Security reformers have offered proposals to allow Social
Security to invest in other assets, such as bank certificates of
deposit, municipal bonds, and corporate bonds.  The choice of the
most appropriate investment policy is a decision to be made by the
Congress and the President after weighing the potential returns and
risks of various asset types. 

We performed our work from September 1996 to January 1998 in
accordance with generally accepted government auditing standards.  We
requested comments on a draft of this report from the Commissioner of
SSA, the Secretary of the Treasury, the Chairman of SEC, and the
Director of OMB.  Staff from these agencies provided technical
comments, which we incorporated in the report as appropriate.  We
also received comments from other experts in budget policy and Social
Security financing and have incorporated them as appropriate. 


--------------------
\12 Report of the 1994-1996 Advisory Council on Social Security, 2
vols.  (Washington, D.C.:  Government Printing Office, 1996). 


THE FEDERAL BUDGET, THE ECONOMY,
AND THE SOCIAL SECURITY TRUST FUND
============================================================ Chapter 2

Any analysis of changes in Social Security's investment policy must
recognize the role of Social Security in the federal budget and the
economy.  Within the federal budget, Social Security is a trust fund
account.  It is also the single largest spending program in the
budget.  The Social Security trust fund currently receives more tax
revenue each year than is needed to pay current benefits.  The
resulting cash surplus, by law, is invested in Treasury securities,
which reduces the Treasury's need to borrow from all other sources to
finance spending on other governmental activities.  While the trust
fund's Treasury securities represent assets for Social Security, they
are future claims against the Treasury.  Social Security's size means
not only that it dominates the budget but also that any program
reforms could have significant implications for the national economy. 
A critical factor to consider in looking at those implications is how
changes in Social Security might affect national saving, which is a
key determinant of long-term economic growth. 


   SOCIAL SECURITY TRUST FUND IS A
   BUDGET ACCOUNT
---------------------------------------------------------- Chapter 2:1

Social Security is a trust fund account within the federal budget.\1
Trust funds are budget accounts that are used to record receipts and
expenditures earmarked for specific purposes and designated as "trust
funds" by law.  Most federal trust funds, including Social Security,
do not have the fiduciary relationships that characterize private
trust funds.  The Office of Management and Budget (OMB) summarizes
the differences between federal and private trust funds as follows: 

     "The beneficiary of a private trust owns the trust's income and
     often its assets.  A custodian manages the assets on behalf of
     the beneficiary according to the stipulations of the trust,
     which he cannot change unilaterally.  In contrast, the Federal
     Government owns the assets and earnings of most Federal trust
     funds, and it can unilaterally raise or lower future trust fund
     collections and payments, or change the purpose for which the
     collections are used, by changing existing law."\2

As OMB's description makes clear, the Congress has the ability to
alter the legislation establishing trust fund revenues and spending. 
For example, the Congress has made a number of changes to Social
Security since the program's inception in the mid-1930s, either to
expand the program or strengthen its financial condition. 

Federal trust fund income and expenditures are typically recorded by
posting special interest-bearing nonmarketable securities to the
Treasury's accounts.  These special securities represent obligations
that the government has issued to itself.  Therefore, from the
standpoint of the Treasury, they are not assets, but, instead, future
claims.  For the Social Security trust fund, special Treasury
securities do represent assets--they signify a reserve of budget
authority for meeting its benefit obligations.  As long as the Social
Security trust fund has a balance, the Treasury is authorized to make
payments on the trust fund's behalf.  When a trust fund runs an
annual surplus as is currently the case with Social Security, its
balance of Treasury securities increases.  When a trust fund runs an
annual deficit as has recently been the case with Medicare's hospital
insurance fund,\3

it redeems some of its Treasury securities in order to pay benefits,
and its balance declines. 


--------------------
\1 Of the federal budget's approximately 1,300 accounts, about 180
are trust fund accounts.  For more details on trust funds, see David
Koitz and Dawn Nuschler, Federal Trust Funds:  How Many, How Big, and
What Are They For?  (Congressional Research Service, 96-686 EPW,
August 30, 1996); Analytical Perspectives, Chapter 17, "Trust Funds
and Federal Funds" (Washington, D.C.:  Government Printing Office,
February 1998); and Social Security:  The Trust Fund Reserve
Accumulation, the Economy, and the Federal Budget (GAO/HRD-89-44,
January 19, 1989).  For more details on budget accounts in general,
see Compendium of Budget Accounts:  Fiscal Year 1998 (GAO/AIMD-97-65,
April 1997) and Budget Account Structure:  A Descriptive Overview
(GAO/AIMD-95-179, September 18, 1995). 

\2 Analytical Perspectives, p.  321. 

\3 Annual outlays for Medicare's Hospital Insurance trust fund
exceeded annual income, including credited interest, in fiscal years
1995, 1996, and 1997.  These deficits required the fund to draw down
its balance of Treasury securities. 


      SOCIAL SECURITY AND FEDERAL
      RETIREMENT TRUST FUNDS HAVE
      LONG-TERM PERSPECTIVE
-------------------------------------------------------- Chapter 2:1.1

Social Security and federal retirement trust funds differ from other
federal trust funds in that they necessarily have a long time
horizon.  Workers covered by Social Security and federal retirement
programs build up their eligibility for benefits in the present and
receive benefits in the future when they retire.  This long-term view
is reflected both in the annual projections for Social Security and
federal retirement funds, which look several decades into the future
and, to some extent, in the longer maturities of the special Treasury
securities that the funds hold.  (See table 2.1.) For example, 78
percent of special Treasury securities held by the Social Security
trust fund and 68 percent in the civil service retirement fund were
due to mature after 2005.  In contrast, all special Treasury
securities held by the transportation trust funds are due to mature
before 2001, reflecting a shorter term outlook.  Since financial
experts agree that a successful stock investing strategy should be
grounded in a long-term outlook, the idea of government stock
investing could be considered more appropriate for
retirement-oriented accounts. 



                               Table 2.1
                
                 Maturities of Treasury Securities Held
                        by Selected Trust Funds

                                       Percent of Treasury securities
                                         due to mature in the period
                                       -------------------------------
                                Total
                             holdings
                             (dollars
                                   in      1997-      2001-      After
Trust fund budget account   billions)       2000       2005       2005
--------------------------  ---------  ---------  ---------  ---------
Social Security's Old Age        $631          6         16         78
 Survivors' and Disability
 Insurance trust fund
Civil Service Retirement         $398         13         19         68
 and Disability Fund
Military Retirement Fund         $126         11         38         51
Bank Insurance Fund               $26         46         46          8
Airport & Airways and             $29        100          0          0
 Highway trust funds
----------------------------------------------------------------------
Source:  Bureau of the Public Debt, Government Account Series,
Monthly Principal Outstanding Report, September 30, 1997. 

While federal employee retirement funds share with Social Security
the long-term outlook that is a prerequisite for stock investing,
Social Security's greater size and prominence mean that more
attention has been focused on changing its investment policy.  Social
Security is the largest federal trust fund--in terms of both annual
spending and account balance.  With fiscal year 1997 outlays of $365
billion, the fund accounts for over one-fifth of all federal spending
and is the largest single item in the federal budget.  At the end of
fiscal year 1997, the Social Security trust fund held $631 billion of
special Treasury securities (39 percent of all outstanding Treasury
securities in the Government Account Series). 


   SOCIAL SECURITY'S CURRENT
   INVESTMENT POLICY
---------------------------------------------------------- Chapter 2:2

Traditionally, Social Security has been financed on a pay-as-you-go
basis under which each current working generation pays for the
benefits of the retired generation.  In this type of system, annual
revenues and benefit payments are roughly equal, while a contingency
reserve is maintained to weather short-term events such as economic
downturns.  However, during the late 1970s and early 1980s, Social
Security's expenditures regularly exceeded revenues, causing a rapid
decline in the fund's balance and raising concerns about the
program's solvency.  In response, the Congress passed reforms in 1977
and 1983 that together were intended to assure Social Security's
solvency for a 75-year period.  These reforms included both tax
increases and reductions in benefits, such as raising the normal
retirement age for future workers.  In enacting the 1983 reforms,
policymakers erred on the side of caution by using fairly pessimistic
assumptions for the short term.  As a result, annual surpluses turned
out to be larger than anticipated.\4 One of the goals of the reforms
was to restore Social Security's contingency reserve.  According to
analysts of Social Security, a prudent reserve ranges from 1 to 1-1/2
year's worth of anticipated spending. 

A combination of annual cash surpluses in recent years and the
interest credited on its Treasury securities has brought the Social
Security trust fund to the top of this range.  As shown in figure
2.1, the trust fund's balance as a percentage of expected annual
benefits is projected to continue growing until it peaks in 2011 at
over 2-1/2 years' worth of expected benefits.  At this level, the
fund balance would substantially exceed the reserve recommended for
short-term contingencies.  In theory, this "excess" amount could be
used to help cover some of the costs of the baby boomers' retirement. 
However, this amount is relatively small when compared to Social
Security's expected future costs. 

   Figure 2.1:  Projected Buildup
   in Social Security Reserve
   (1997-2012)

   (See figure in printed
   edition.)

Source:  1997 Trustees' Report, intermediate assumptions. 

The Social Security trust fund, by law, must invest in
"interest-bearing obligations of the United States or in obligations
guaranteed as to both principal and interest by the United States."
This investment policy, described more fully in chapter 3, dates back
to the program's origin in 1935 and is intended to ensure the safety
of the trust fund's assets.  Treasury securities are widely
considered to be a risk-free investment in the sense that the federal
government has never defaulted on its debt obligations and is not
expected to do so in the future.  Throughout its history, the Social
Security trust fund has invested mostly in a special type of Treasury
security that cannot be sold on the open market.  The trust fund's
Treasury securities represent a reserve for the program, allowing
Social Security to pay benefits as long as it has a balance in its
account.  When the trust fund needs to draw down its balance, the
Treasury must obtain the necessary money to repay the trust fund by
borrowing from the public, unless the Congress and the President take
offsetting actions to raise taxes or cut spending. 

The trust fund's special Treasury securities earn interest credits at
a statutory rate linked to market yields.  These interest credits
take the form of additional Treasury securities and add to the trust
fund's balance.  In recent years, interest credits have accounted for
a growing share of annual trust fund income.  Between calendar years
1985 and 1996, the trust fund's interest credits grew from 1 percent
of its total income to 9 percent.  By 2012, interest credits are
projected to rise to 14 percent of total income.  While the interest
is added to tax revenues in assessing the finances of the Social
Security trust fund, the cash surplus is more meaningful in the
broader budgetary and economic context. 


--------------------
\4 For more details on the intent of the 1983 reforms, see David
Stuart Koitz, Social Security Financing Reform:  Lessons From the
1983 Amendments (Congressional Research Service, 97-741 EPW, July 24,
1997). 


   SOCIAL SECURITY'S PLACE WITHIN
   THE FEDERAL BUDGET
---------------------------------------------------------- Chapter 2:3

Social Security is not included in the measure of the federal budget
that is used as the basis for the budget controls under the Budget
Enforcement Act.  This measure is known as the "on-budget" deficit.\5
However, Social Security's "off-budget" status does not change the
impact that its finances have on the government's overall fiscal
position, as explained in the following passage from a Congressional
Budget Office (CBO) report:\6

     "Social Security's benefits alone account for more than
     one-fifth of federal spending, and its payroll taxes account for
     about one-fourth of government revenues.  Therefore, most
     economists, credit market participants, and policymakers, when
     they seek to gauge the government's role in the economy and its
     drain on the credit markets, look at the total budget
     figures--including Social Security."

As indicated by CBO, Social Security is included in the most commonly
used measure of the government's financial balance, known as the
unified, or "total," budget deficit/surplus.\7 The unified budget
measure includes all federal spending and revenue and generally
approximates the amount of annual federal borrowing from the public. 
Including the Social Security cash surplus in the unified budget
means that it partially offsets a deficit in all other government
accounts.  For example, in fiscal year 1997, a $62 billion cash
deficit in other government accounts was offset by a $40 billion cash
surplus in Social Security, resulting in a unified budget deficit of
$22 billion.  The interest credited to the trust fund does not
currently affect the unified deficit because it is an internal
transaction of the government.  One part of the government (the
Treasury) credits the interest to another part (the trust fund), so
the two transactions offset one another and there is no net budgetary
effect.\8

Since Social Security's annual cash surplus is available to the
Treasury to help finance current spending, it reduces the Treasury's
need to borrow from other sources.\9 Treasury borrowing from Social
Security and other federal trust fund accounts is referred to as
"debt held by government accounts." Treasury borrowing from all other
sources (e.g., individuals, pension funds, and financial
institutions) is called "debt held by the public." Together, these
two types of debt comprise the gross federal debt.\10 Debt held by
the public is a more commonly used measure because it best represents
the current impact of past federal borrowing on the economy.  In
contrast, debt held by trust funds is an internal government
transaction that has no current impact--its economic effects are not
felt until a trust fund needs to draw on its Treasury securities to
make program expenditures.\11


--------------------
\5 Social Security was legally removed from all calculations of
budget totals by the Budget Enforcement Act of 1990 (Public Law
101-508).  For further details on Social Security's budgetary
treatment, see David Koitz, Social Security:  Its Removal From the
Budget, and Procedures for Considering Changes to the Program
(Congressional Research Service, 93-23 EPW, Revised January 4, 1993). 

\6 The Economic and Budget Outlook:  Fiscal Years 1999-2008,
Congressional Budget Office, January 1998, p.  34. 

\7 Unless otherwise noted, any reference to the deficit in this
report means the unified budget deficit. 

\8 Besides interest credited to trust funds, other intragovernmental
transactions (such as the payments from the Treasury's general fund
to Medicare's Supplementary Medical Insurance trust fund) also do not
affect the government's current borrowing requirement. 

\9 If, as currently projected, the federal government experiences a
period of unified budget surpluses, the Treasury would essentially no
longer need to borrow additional funds from other sources except to
roll-over existing debt.  In this case, at least a portion of Social
Security's cash surplus could be used to reduce the level of debt
held by the public.  In any case, by law, the Social Security trust
fund would continue to invest in Treasury securities as long as
Social Security were in a surplus position. 

\10 With a few minor exceptions, the gross debt is the measure that
is subject to the debt limit. 

\11 For further information on federal debt, see Federal Debt: 
Answers to Frequently Asked Questions (GAO/AIMD-97-12, November 27,
1996). 


   FEDERAL FISCAL POLICY AND
   NATIONAL SAVING
---------------------------------------------------------- Chapter 2:4

As described above, Social Security's finances have a significant
influence on the government's overall fiscal position.  The
government's fiscal position, in turn, affects national saving, which
is a key determinant of long-term economic growth.  Saving is that
portion of income not used for current consumption.  The nation's
saving is composed of the private saving of individuals and
businesses and the saving or dissaving of all levels of government. 
In general, government budget deficits subtract from national saving
by absorbing funds that otherwise could be used for private
investment.  Conversely, government budget surpluses add to saving. 
In the case of the federal government, surpluses allow the government
to pay off some of its maturing debt, thereby reducing the
outstanding level of debt held by the public and freeing up
additional funds for private investment. 

Since the 1970s, private saving has declined, while federal budget
deficits have consumed a large share of these increasingly scarce
savings.  In recent years, federal deficits have declined sharply
from the levels of the 1980s and early 1990s, freeing up some
additional funds for investment.  Nevertheless, total national saving
and investment remain significantly below the levels of the 1960s and
1970s.  Raising saving and investment levels would improve the
long-term productivity of the economy, thereby boosting economic
growth. 

The most direct way for the federal government to contribute to
increased saving and long-term economic growth is to achieve and
maintain a balanced budget or a budget surplus.  While such a fiscal
policy would raise the government's contribution to national saving,
it would probably not raise total national saving dollar for dollar. 
The overall impact on national saving depends on how consumers
respond to the government's fiscal actions.  For example, cuts in
federal spending or increased taxes could cause a reduction in
private saving that would diminish the total impact on national
saving. 

While additional fiscal restraint would probably reduce consumption
in the short term, it would promote higher living standards over the
long term.  In a report issued last fall and in subsequent testimony,
we found that alternatives to current fiscal policy that emphasize
additional restraint could, over the long term, significantly boost
per capita GDP.\12 Taking actions to increase the size of the future
economy is particularly important because of the aging population.  A
bigger economic pie would make it easier for future workers to meet
the dual challenges of paying for the baby boomers' retirement and
achieving a rising standard of living for themselves. 


--------------------
\12 Budget Issues:  Analysis of Long-Term Fiscal Outlook
(GAO/AIMD/OCE-98-19, October 22, 1997) and Budget Issues:  Long-Term
Fiscal Outlook (GAO/T-AIMD/OCE-98-83, February 25, 1998). 


BALANCING POTENTIAL RETURNS AND
RISKS FOR THE SOCIAL SECURITY
TRUST FUND
============================================================ Chapter 3

Designing an investment policy for the Social Security trust fund
involves balancing potential returns and risks.  Currently, the
Social Security trust fund, by law, invests in securities backed by
the federal government and receives a relatively low return with
minimal risk.  Investing in the stock market offers the prospect of
higher returns but greater risk. 

The higher returns possible with stock investing would allow the
trust fund to pay benefits longer, even without other program
changes.  Assuming the historical average rate of return, investing
Social Security's future annual cash surplus and interest on its
Treasury securities in the stock market could delay the trust fund's
exhaustion by about a decade.\1 If only Social Security's cash
surplus is invested, then the possible delay in the trust fund's
exhaustion would be reduced to 3 years.  It is also important to
recognize that any possible gain for the Social Security trust fund
from a stock investment scenario would be significantly less if
future stock returns fall below the historical average.  The only way
for the trust fund to earn the higher returns possible with stock
investing is to take on greater risk.  Diversifying the stock
portfolio would reduce the risks associated with individual stocks,
but the trust fund would still be vulnerable to loss in the event of
a general stock market downturn. 

Stock investing, by itself, is unlikely to solve Social Security's
long-term financing problem.  If stock investing is implemented in
isolation from other program changes, the trust fund would face the
certain need to liquidate its assets to pay benefits with no
certainty of what future stock prices would be.  The more the trust
fund is counting on stock sales to pay promised benefits, the greater
its vulnerability in the event of a general stock market downturn. 
In the future, the trust fund would be less vulnerable to the
necessity of selling stocks at a loss if a stock investment policy
were implemented in combination with other program changes that
improve Social Security's financing.  As part of a package, investing
in stocks could increase the trust fund's expected investment income
and reduce the size of benefit cuts and/or payroll tax increases. 


--------------------
\1 We assumed that the trust fund would continue to hold enough
Treasury securities as a contingency reserve to equal at least 100
percent of the next year's expected expenditures. 


   SPECIAL TREASURY INVESTMENT
   POLICY:  LOW RETURNS AND
   MINIMAL RISK
---------------------------------------------------------- Chapter 3:1

Under the current law, the Secretary of the Treasury--managing
trustee for the trust fund-- is required to invest in special
Treasury securities unless the Secretary determines that purchasing
marketable Treasury and agency securities is "in the public
interest." In the past, the Treasury Department has, at times,
determined that such purchases would be in the public interest,
although such purchases have been rare.\2 With the practice of
investing in special Treasury securities, the Social Security trust
fund receives a relatively low return with minimal risk. 

By law, the interest rate on special Treasury securities is equal, at
the time of issue, to the average market yield on outstanding
marketable government securities not due or redeemable for at least 4
years.  According to the Congressional Research Service's analysis of
the law and practice governing Social Security's investment policy,
this statutory rate was intended to confer neither an advantage nor a
disadvantage on the trust fund.\3 From the trust fund's perspective,
the statutory rate represents a longer-term interest rate, and
long-term interest rates have historically been higher than
short-term rates.  From the government's perspective, the statutory
rate was intended to approximate the cost of long-term borrowing from
the public.\4

Like Treasury securities sold to the public, special Treasury
securities are backed by the full faith and credit of the U.S. 
government and are viewed as having no risk of default.  Although it
cannot sell its holdings in the open market, the Social Security
trust fund faces no liquidity risk because, by law, it can redeem
special Treasury securities before maturity without penalty.  This
liquidity feature is particularly important for the trust fund if it
needs to dip into its assets to cover a payroll tax shortfall during
a general economic downturn.  Moreover, redeeming special Treasury
securities before maturity presents no risk of loss due to interest
rate fluctuations because the trust fund can recover the par value
plus accrued interest.  In contrast, the trust fund would have to
sell marketable Treasury securities at the market price--which
fluctuates inversely with market interest rates.  The market price of
a Treasury security falls when the current interest rate on Treasury
securities of equal maturity rises.  Selling marketable Treasury
securities before maturity when market interest rates are rising
could result in a sizable loss.  In practice, the Treasury has
allowed the Social Security trust fund to redeem its special Treasury
securities at any time to pay benefits but not to do so solely for
the purpose of maximizing the trust fund's return. 

Like any investor, the Social Security trust fund faces the risk that
its investment returns will be eroded by inflation.  This is a
particular concern given that Social Security benefits are indexed
for inflation.  Social Security beneficiaries receive an annual cost
of living adjustment that is normally based on the Consumer Price
Index.  Under the intermediate scenario for the next 75 years, which
the Trustees regard as their "best estimate," the ultimate nominal
interest rate assumed over the long term is 6.2 percent, while annual
inflation is assumed to be 3.5 percent.  Thus, the trust fund is
expected to receive an ultimate real (after inflation) interest rate
of 2.7 percent on its Treasury holdings. 

Although the current debate focuses on allowing the Social Security
trust fund to invest in the stock market, there also are investment
options within the federal government.  Although the trust fund is
not specifically authorized to do so by the Social Security Act, it
may purchase securities issued by the Government National Mortgage
Association, the Federal National Mortgage Association, and other
Federal farm and home credit entities.\5

The Advisory Council's Maintain Benefits approach suggested
considering such investments to increase the trust fund's return. 
Agency securities typically pay more because they are not uniformly
guaranteed as to principal and interest and there is some risk of
default.  And, like marketable Treasury securities, agency securities
would expose the trust fund to potential losses due to fluctuating
market prices. 

Another option could be to change the statutory rate of interest for
special Treasury securities.  For example, the Congress could raise
the rate by a fixed percentage or link the rate to a stock market
index, such as the Standard and Poor's index of 500 large stock
companies (S&P 500).  The interest premium in excess of the average
rate Treasury pays on debt held by the public would represent a
general revenue transfer to the Social Security trust fund. 
Increasing the rate credited to the Social Security trust fund
account would appear to boost the program's finances inasmuch as the
trust fund's balance would increase.  However, crediting more
interest to the trust fund would not generate revenue for the
government, so the government's capacity to finance retirement
benefits would be unchanged. 


--------------------
\2 As of 1996, marketable Treasury securities represented only 0.009
percent of the trust fund's holdings. 

\3 Geoffrey Kollmann, Social Security:  Investing the Surplus
(Congressional Research Service, 91-129 EPW, January 27, 1991). 

\4 The average nominal interest rate on new special Treasury
securities issued in 1996 was 6.6 percent.  The average nominal rates
for marketable medium- and long-term Treasury securities outstanding
in 1996 was 6.5 percent for Treasury notes issued with a term of at
least 1 year but not more than 10 years, and 9 percent for Treasury
bonds with a term of more than 10 years. 

\5 The Social Security Act requires that the Secretary of the
Treasury invest trust funds in "interest-bearing obligations of the
United States or in obligations guaranteed as to both principal and
interest by the United States." Although certain federally sponsored
agency obligations do not meet these criteria, the Secretary may
invest in such obligations based on a 1966 opinion of the Attorney
General.  The opinion held that notwithstanding the absence of
statutory language pledging the "faith" or "credit" of the United
States, agency guaranties or other contractual liabilities issued in
pursuance of an agency's statutory functions constitute "general
obligations of the United States backed by its full faith and
credit." Op.  Atty.  Gen.  327 (1966). 


   STOCK RETURNS HAVE BEEN HIGHER
   BUT ARE MORE UNCERTAIN
---------------------------------------------------------- Chapter 3:2

Historically, the rates of return on stocks have exceeded interest
rates on Treasury securities, although stock returns are more
variable.  According to an analysis prepared for the Advisory
Council, real yields on stocks--i.e., adjusted for inflation--have
averaged about 7 percent.\6 In its deliberations, the Advisory
Council agreed to use this rate in estimating average future yields
on stocks.  Of course, an average return over a long period of time
obscures the reality that stock returns fluctuate substantially from
year to year, and there have been years in the past with negative
returns.  Figure 3.1 shows the annual returns--not adjusted for
inflation--for large company stocks from 1950 to 1996.  Actual
nominal returns varied widely from the annualized average return over
the period and ranged from a low of -26.5 percent in 1974 to a high
of 52.6 percent in 1954. 

   Figure 3.1:  Annual Returns on
   Large Company Stocks in
   Comparison to the Annualized
   Average Return for 1950 through
   1996

   (See figure in printed
   edition.)

Note:  Large company stock returns are based upon the S&P 500 index;
before 1957, the index consisted of 90 of the largest stocks.  The
total annual return reflects capital appreciation and cash income
during the year, assuming any income is reinvested, and does not
reflect any transaction costs.  The annualized average return for the
period was 12.8 percent.  This compound annual rate reflects the
return over the period figured on a constant year basis, which is not
the same as the arithmetic average of rates for each year. 

Source:  Stocks, Bonds, Bills, and Inflation 1997 Yearbook.  Used
with permission.  Copyright 1997 Ibbotson Associates,
.  All rights reserved.  (Certain portions of this
work were derived from copyrighted works of Roger G.  Ibbotson and
Rex Sinquefield.)

Ibbotson Associates, Inc.  has granted permission solely for display
in this medium and any copying, printing, modifying, distributing,
disclosing transferring, displaying in another medium or format, or
incorporating in another, of any portion of this work is expressly
prohibited. 

According to an analysis by the Congressional Research Service,\7 the
annualized average return for the S&P 500 since 1950 was almost twice
the average rate credited on the Social Security trust fund's
Treasury securities.\8 Although the 30-year moving average of the S&P
500 since 1970 consistently outperformed the Treasury returns
credited to the Social Security trust fund, the 10-year moving
average of the S&P 500 underperformed the trust fund's Treasury
returns at times.  Again, a long-term average return does not reflect
fluctuations in year-to-year stock returns.  In fact, nominal stock
returns were less than the Social Security trust fund's annual yield
in 17 years from 1950 to 1996--more than 35 percent of the time. 

Short-run fluctuations generally are less of a concern for a
long-term investor who buys and holds investments.  Table 3.1
illustrates the actual best and worst nominal returns on stocks as
well as long-term government bonds for investment periods ranging in
duration from 1 year to 20 years.  An investor would face uncertain
returns in the short term given that annual returns range widely and
were negative in nearly 1 out of 4 years.  Likewise an investor can
lose money selling marketable government bonds before maturity
because of bond price fluctuations.  As table 3.1 shows, the range
between the best and the worst returns narrows as the investment time
horizon lengthens.\9 Given that from 1926 to 1996, there was no
20-year period with a negative stock return, an investor might
reasonably expect to earn a positive return over 20 years. 



                               Table 3.1
                
                Best and Worst Returns on Large Company
                 Stocks and Long-term Government Bonds
                for Varying Investment Periods From 1926
                              Through 1996

                             (Percentages)

                         Large company stock     Long-term government
                               returns               bond returns
                        ----------------------  ----------------------
Investment period            Worst        Best       Worst        Best
----------------------  ----------  ----------  ----------  ----------
1 year                      -43.34       53.99       -9.18       40.36
10 years                     -0.89       20.06       -0.07       15.56
20 years                      3.11       16.86        0.69       10.45
----------------------------------------------------------------------
Notes:  Annual compound rates of return were calculated for
overlapping holding periods from 1926 through 1996.  For the 71
1-year holding periods, annual returns on large company stocks and
long-term government bonds were positive in 72 percent of the years. 
For the 62 10-year holding periods, returns were positive in 97
percent of the periods for large company stocks and in 98 percent for
long-term government bonds.  For the 52 20-year holding periods,
returns on large company stocks and long-term government bonds were
positive in every period. 

Source:  Stocks, Bonds, Bills, and Inflation 1997 Yearbook.  Used
with permission.  Copyright 1997 Ibbotson Associates,
.  All rights reserved.  (Certain portions of this
work were derived from copyrighted works of Roger G.  Ibbotson and
Rex Sinquefield.)

Ibbotson Associates, Inc.  has granted permission solely for display
in this medium and any copying, printing, modifying, distributing,
disclosing transferring, displaying in another medium or format, or
incorporating in another, of any portion of this work is expressly
prohibited. 


--------------------
\6 Joel Dickson, "Analysis of Financial Conditions Surrounding
Individual Accounts," Report of the 1994-1996 Advisory Council on
Social Security, Volume II, pp.  484-488.  The stock market realized
an annualized real yield of approximately 7 percent from 1900 to
1995. 

\7 Statement of David Koitz of the Congressional Research Service
before the Subcommittee on Social Security of the House Committee on
Ways and Means on April 10, 1997, at a hearing on the future of
Social Security. 

\8 According to the Congressional Research Service's analysis, since
1950, the annualized average nominal return for the S&P 500 was 11.36
percent assuming annual administrative costs of 1 percent, compared
to 5.96 percent for the trust fund's nominal yield. 

\9 The variation of returns around the expected average can be
quantified in statistical terms, such as the standard deviation.  For
more data about stock and other asset returns, see Stocks, Bonds,
Bills, and Inflation Yearbook (Chicago, Illinois:  Ibbotson
Associates). 


      FUTURE STOCK RETURNS ARE
      UNCERTAIN
-------------------------------------------------------- Chapter 3:2.1

There is no guarantee that investing in the stock market, even over 2
or 3 decades, will yield the long-run average return.  According to
economic and financial literature, there are reasons to believe that
future stock returns could be less than the historical average.\10
Also, as discussed in chapter 4, government investing by itself could
affect stock prices and returns at least in the short run.  While the
average historical stock return is commonly used in assessing future
stock performance, assuming a moderately lower return could also be
consistent with the expected economic and demographic outlook. 

An investor entering the market at today's high stock prices may earn
less than the long-term historical average.  Two fundamental measures
for evaluating stock prices are the dividend yield--the ratio of
annual dividends to stock prices--and the price-earnings ratio. 
According to an analysis of historical stock performance, returns on
stocks over 10-year periods have been well below average when the
dividend yield was low and the price-earnings ratio was high.\11
Indeed, the stock market has been at record high levels in recent
years, and the dividend yield is below long-run historical values.\12
Likewise, the price-earnings ratio is well above its long-run
average.  Some analysts have estimated that recent price-earnings
ratios would be consistent with a 1 percentage point decline in the
long-run return on stocks; in other words, future stock returns could
decline from the historical real average of 7 percent to 6 percent
over the long run.\13

Another factor that may affect future stock returns is that the U.S. 
economy is expected to slow as the population ages.  The rate of
national saving and the growth in real wages and productivity,
factors that relate to economic growth, have slowed notably in the
past two decades.\14 Social Security's Trustees assume future growth
in the GDP will slow as the baby boom generation retires and
relatively fewer young people begin work.  Whereas the economy grew
at an average real rate of 2.2 percent from 1989 to 1997, real
economic growth under the Trustees' intermediate scenario is assumed
to be 2.0 percent annually over the next decade and then to slow to
1.3 percent by 2020.  Some economists have estimated that the
macroeconomic and demographic outlook would be consistent with
long-run stock returns lower than the historical average.\15

Another uncertainty is whether the baby boomers' retirement might
affect the stock market.  As they retire, baby boomers are expected
to sell stocks to finance consumption, and private pension plans
likewise will sell stocks to finance retirement benefits.  The
Advisory Council's technical panel reported that selling pressure
resulting from the sheer number of the baby boom retirees could
possibly depress stock prices but that estimating any baby boom price
effect would be highly speculative.  Given that the financial markets
and investors anticipate the aging of the population, asset prices
may adjust downward gradually beforehand rather than dropping
abruptly when the baby boomers begin retiring.  Also, some analysts,
including those we interviewed, suggest that global demand for stocks
could offset any baby boom price effect.  For example, investors from
countries with relatively younger populations might invest in the
U.S.  stock market to save for their own retirement even as the baby
boomers are selling their stocks. 

While future returns on stocks might reasonably be expected to be
less than the historical average, analysts, including those we
interviewed, expect stock returns to be higher than those on Treasury
securities over the long term.  How much higher is uncertain.  The
spread between the rates of return on stocks and Treasury securities
has been shrinking.  Historically, stocks have earned higher rates of
return than those of Treasury securities to compensate for the
additional risk associated with stocks.  This "equity risk premium"
has declined since the 1950s from about 7 percent to around 3 to 4
percent today.\16 It is unclear whether this change will be
long-lasting or whether the equity premium will decline even further. 
Some economists have suggested that the shrinking risk premium
reflects a structural change in that the economy appears less
susceptible to recessions.\17

To the extent that corporate profits fluctuate with general economic
conditions, fewer downturns translate into less volatility in
corporate earnings.  If investors perceive that the outlook for
corporate earnings is more certain and that stocks may be less risky
than they have been historically, stock investing might carry a lower
premium and, therefore, relatively lower returns.  The uncertainty
about the risk and size of the risk premium have implications for
analyzing the benefits of a stock investment proposal.\18

Although investors believe that over the long run stock returns will
always be higher than bond returns, one study has questioned this
conventional wisdom.\19 According to standard analytical models,
stocks are virtually certain to outperform bonds over a long enough
investment period.  However, extrapolating from these standard
models, the study estimated that a stock portfolio has a 32 percent
chance of underperforming a bond portfolio over a 10-year horizon. 
Even over a 30-year investment time horizon, there is still about a
20 percent chance that a bond portfolio would provide a higher
return.  As this study indicates, investing in the stock market does
not ensure a higher return than might be possible investing in
government and corporate bonds. 


--------------------
\10 For example, see John E.  Golob and David G.  Bishop, "What
Long-Run Returns Can Investors Expect from the Stock Market?" Federal
Reserve Bank of Kansas City Economic Review, Vol.  82, No.  3 (Third
Quarter 1997), pp.  5-20 and John H.  Cochrane, "Where is the Market
Going?  Uncertain Facts and Novel Theories," Economic Perspectives,
Vol.  XXI, Issue 6 (November/December 1997), pp.  3-37. 

\11 Burton G.  Malkiel, A Random Walk Down Wall Street (New York
City, New York:  W.W.  Norton & Company, 1996), pp.  384-389. 

\12 Golob and Bishop, pp.  7-8, estimated that whereas the dividend
yield for the S&P 500 had averaged about 4 percent since the 1950s,
this ratio dropped below 2 percent for the first time in 1996 and has
remained below 2 percent most of the time since then. 

\13 Golob and Bishop, pp.  13, 14, and 16. 

\14 See Retirement Income:  Implications of Demographic Trends for
Social Security and Pension Reform (GAO/HEHS-97-81, July 11, 1997). 

\15 Golob and Bishop estimated that macroeconomic trends could reduce
stock returns over the next decade by about one-half a percentage
point, for a long-run real return of 6.5 percent.  A 1997 report for
the Twentieth Century Fund/Economic Policy Institute, Saving Social
Security With Stocks:  The Promises Don't Add Up, estimated that
given the Trustees' other assumptions, future stock returns could be
as low as 4.0 percent. 

\16 For an historical examination, see Olivier J.  Blanchard,
"Movements in the Equity Premium," Brookings Papers On Economic
Activity, 2:1993, pp.  75-118. 

\17 Goldman Sachs, "The Equity Risk Premium and the Brave New
Business Cycle," U.S.  Economics Analyst, No.  97/8, February 21,
1997. 

\18 For further reading, see Jeremy J.  Siegel and Richard H. 
Thaler, "Anomalies:  The Equity Premium Puzzle," Journal of Economic
Perspectives, Vol.  11, No.  1 (Winter 1997), pp.  191-200. 

\19 Martin Leibowitz and William Krasker, "The Persistence of Risk: 
Stocks Versus Bonds Over the Long Term," Financial Analysts Journal,
November/December 1988, pp.  40-47. 


   HIGHER RETURNS COULD DELAY
   EXHAUSTION OF THE SOCIAL
   SECURITY TRUST FUND
---------------------------------------------------------- Chapter 3:3

If the stock market continues to outperform Treasury securities, the
Social Security trust fund could increase its investment revenue with
a stock investment policy.  The higher returns possible on stocks
would allow the Social Security trust fund, even without other
program changes, to pay benefits longer before depleting its
assets.\20 The potential gain from stock investing would depend on
both how much the Social Security trust fund invests in the stock
market and how much future stock returns are. 

According to the Trustees' 1997 intermediate estimate, the trust fund
expects to collect roughly $30 billion more in cash than is needed to
pay benefits each year from 1998 until 2008 and continue to receive
some excess cash until 2012.  In addition, the interest credited on
the trust fund's special Treasury securities was roughly $40 billion
in 1997.  Given that the Social Security trust fund's balance,
beginning in 1997, was expected to exceed 150 percent of its annual
expenses,\21 the trust fund theoretically could start investing in
stocks in 1998.  Under the Trustees' intermediate projections, the
trust fund does not anticipate that it would need to tap its
investment income and assets to pay current benefits for nearly 15
years. 

We developed two scenarios to illustrate the trust fund's potential
gain from stock investing.  Under an aggressive scenario, the trust
fund would invest both its future annual cash surplus and interest in
the stock market, while maintaining a contingency reserve of special
Treasury securities equal to at least 100 percent of the next year's
expected expenditures.  We also tested an alternative scenario under
which only Social Security's cash surplus would be invested in
stocks, and Social Security's cash deficit, beginning in 2012, would
be financed from stock earnings and sales.  At our request, the
Social Security Administration's Office of the Chief Actuary
simulated the potential effect on the trust fund of these two
investment scenarios using the Trustees' 1997 intermediate
assumptions.  The simulations use the 7-percent real yield on stocks
assumed by the Advisory Council in estimating future stock
performance.  This historical average stock return is 4.3 percentage
points higher than the trust fund expects on its special Treasury
securities.  In the simulations, stock earnings are assumed to be
reinvested in the market unless the trust fund needs cash to pay
benefits or to invest in Treasury securities to maintain a
100-percent contingency reserve.  The simulations also reflect the
Advisory Council's assumption that annual administrative costs for
the trust fund's stock holdings would be 0.5 basis points.\22 See
appendix I for further discussion of our assumptions. 

Figure 3.2 shows the estimated trust fund assets under the aggressive
investment scenario compared to the cash surplus scenario as well as
to the current statutory policy of investing solely in special
Treasury securities.  These simulation results illustrate some
outcomes associated with our two alternative stock investment
policies.  They should not be interpreted as forecasts and do not
represent the full range of possible outcomes for the Social Security
trust fund. 

   Figure 3.2:  Estimated Trust
   Fund Assets Under Current Law
   and Two Alternative Stock
   Investment Scenarios

   (See figure in printed
   edition.)

Note:  Assets at the beginning of year. 

Source:  Social Security Administration. 

Assuming the historical 7 percent real yield on stocks, investing
both Social Security's future annual cash surplus and the interest on
its Treasury securities in the stock market could delay the projected
exhaustion of the trust fund for about a decade, from 2029 to 2040. 
This potential delay would extend the trust fund's life well into the
baby boomers' retirement years.\23 However, this delay results from a
scenario representing an outer bound of how much the trust fund might
invest in the stock market.  Under the aggressive scenario, the trust
fund would hold only enough Treasury securities to cover its
contingency needs and would amass a sizable stock portfolio.  Within
5 years, the trust fund would have about $500 billion invested in the
stock market.  Stocks as a share of the trust fund's portfolio would
peak in 2017 at more than 70 percent.  While a 70 percent stock
allocation is not necessarily unsound from an investment perspective,
it would be a dramatic shift from investing solely in Treasury
securities. 

In nominal dollars, the trust fund's stock holdings under the
aggressive scenario would peak in 2025 at about $4 trillion.\24
According to Social Security's Trustees, the ratio of trust fund
assets at the beginning of a year to that year's expected expenses is
a useful measure of the trust fund's asset level.  At their peak, the
trust fund's stock holdings would represent about twice Social
Security's expected expenses in 2025.\25 By 2034, the trust fund's
assets would drop to about 150 percent of expected annual expenses. 
At that time, the trust fund would still have more than $1.5
trillion--about one-third of its portfolio--in stocks.  In 2036, the
trust fund would liquidate all of its stocks, and its special
Treasury securities would drop below a 100 percent contingency
reserve level. 

If only Social Security's cash surplus were invested, still assuming
a 7 percent real rate of return, then the trust fund's projected
exhaustion could be delayed by only 3 years, from 2029 to 2032.  This
scenario is somewhat more conservative; the trust fund would invest
less than half as much as under the aggressive scenario.  Within 5
years, the trust fund's stock investments would be about $200
billion.  Stocks as a share of its portfolio would peak at about 35
percent, which is conservative in comparison to the 60 percent held
by state and local government pension plans as a whole.  The trust
fund's stock holdings, in nominal dollars, would peak in 2017 at
approximately $1.2 trillion, an amount roughly equivalent to that
year's expected expenses.  In 2024, the trust fund would liquidate
all of its stocks, but its special Treasury securities would still
represent more than 150 percent of the next year's expected expenses. 
In 2028, however, the trust fund's balance would drop below a 100
percent contingency reserve level. 

Again, these estimates of the potential delay of the trust fund's
exhaustion were based on a 7 percent average real return on stocks. 
The possible gain for the Social Security trust fund would be
significantly less if future stock returns are lower than this
historical average.  As an illustration, if the future real return on
stocks is 1 percentage point lower, the aggressive scenario would
extend the trust fund's life by not 11 years, but only 6 years to
2035.  Again assuming the real return on stocks is 1 percentage point
lower, the possible delay under the cash surplus scenario would not
be 3 years, but only 2 years to 2031.  These results demonstrate the
sensitivity of the rate of return assumption and are not intended to
represent the worst or the most likely outcomes for the trust fund. 


--------------------
\20 This statement would also apply to other assets, such as
corporate bonds, which could yield potentially higher returns than
the current statutory policy of investing solely in Treasury
securities. 

\21 As discussed in chapter 2, a balance of 100 to 150 percent of
anticipated annual spending is considered a prudent contingency
reserve. 

\22 A basis point is 1/100 of 1 percentage point, so one-half of a
basis point is 0.00005. 

\23 In 2040, the oldest baby boomers would be 94 years old and the
youngest would be 76 years old. 

\24 Nominal asset levels are not comparable over time due to
inflation, economic growth, and growth in the Social Security
program. 

\25 Total assets, including special Treasury securities, would be
approximately $6.1 trillion--about three times expected expenses in
2025. 


   STOCK INVESTING ENTAILS GREATER
   RISK
---------------------------------------------------------- Chapter 3:4

Investing in the stock market involves a clear trade-off.  In
exchange for the prospect of higher returns, the trust fund must
accept greater risk.  The trust fund would face greater uncertainty
about its future returns and even the chance of losing money.  Under
the current policy, the trust fund receives a relatively low rate
investing in Treasury securities but can readily liquidate its
special Treasury holdings to pay benefits.  In contrast, the trust
fund would face uncertainty as to the amount or timing of future
stock earnings and dividends.  Moreover, just as the trust fund
expects to liquidate its Treasury securities to pay benefits, it
would have to sell its stocks to get cash to pay benefits.  There is
no certainty about what stock prices would be when the trust fund has
to sell or whether it could recover amounts invested. 

The primary risk that the trust fund would face is "market risk," or
the possibility of financial loss caused by adverse market movements. 
When the stock market drops, prices of stocks--regardless of their
individual quality--fall and can stay depressed for a prolonged
period of time.  Fluctuations in overall market rates of interest can
affect the stock market, and rising interest rates tend to depress
stock prices.  Market risk does not disappear over time.  Although a
long investment time horizon provides more time to recover from
short-term fluctuations, an investor also would have more time to
encounter a prolonged stock market downturn. 

Depending on the composition of its stock portfolio, the trust fund
could also be exposed to "concentration risk," or the potential loss
resulting from a heavy investment in a group of related companies or
an industry susceptible to the same economic dynamics.  Like any
investor, the trust fund would face "default risk," or the exposure
to loss due to an individual company failing. 


      DIVERSIFICATION REDUCES
      DEFAULT AND CONCENTRATION
      RISK
-------------------------------------------------------- Chapter 3:4.1

According to portfolio theory, diversification reduces risk. 
Diversifying a stock portfolio across companies and industries
reduces both default and concentration risk.  Diversification also
reduces the risk that the portfolio's return will vary widely from
the expected market return.  Indexing, discussed in more detail in
chapter 4, is one way to broadly diversify a stock portfolio and to
match the approximate market return.  Under the Advisory Council's
Maintain Benefits approach, the trust fund would invest in stocks
indexed to the broad stock market. 

A diversified stock portfolio, however, does not protect against the
risk of a general stock market downturn.  An investor can shield
against stock market risk by diversifying into other types of assets,
such as corporate bonds.  Also, one way to mitigate U.S.  stock
market risk is to diversify into international markets.  To minimize
exposure to short-term stock market fluctuations, an investor can
hold less risky, albeit lower-yielding, assets to cover liquidity
needs in the short run. 


      SOCIAL SECURITY TRUST FUND
      WOULD BE VULNERABLE TO STOCK
      MARKET RISK
-------------------------------------------------------- Chapter 3:4.2

Higher stock returns could delay the trust fund's exhaustion, but,
without other program changes, the trust fund inevitably will have to
liquidate its stocks to pay benefits.  Social Security's tax revenues
are projected to be inadequate to cover annual benefits beginning in
2012.  To pay benefits after that point, the trust fund will have to
draw upon its investment earnings and eventually its assets to cover
the shortfall.  Riding out a general stock market downturn could be
difficult for the Social Security trust fund as it faces a cash
deficit and growing numbers of retirees.  The trust fund might have
to sell its stock holdings at a loss to raise cash to pay benefits. 
The more the trust fund is counting on stock sales to finance current
benefits, the greater its vulnerability in the event of a general
stock market downturn. 

Conceivably, the trust fund could draw on its contingency reserve to
avoid selling its stocks at a loss during a general market downturn. 
Once the trust fund depletes its special Treasury holdings though, it
would be wholly subject to the vagaries of the stock market to get
cash needed to pay benefits.  Under such circumstances, a contingency
reserve of 100 or even 150 percent of expected annual expenses may be
inadequate for the trust fund to ride out a prolonged market
downturn. 

Again, if stock investing is implemented in isolation from other
program changes, the trust fund would have to liquidate a sizable
stock portfolio.  The size of the trust fund's stock holdings as a
share of the stock market and possible price effects are discussed
further in chapter 4.  In the simulations, the liquidation of the
trust fund's stock portfolio would coincide with the baby boomers'
retirement.  A sustained stock market downturn during this period not
only would decrease the value of the trust fund's stock holdings but
would affect retirees' personal savings as well. 


      A GENERAL STOCK MARKET
      DOWNTURN COULD COINCIDE WITH
      A SOCIAL SECURITY TAX
      SHORTFALL
-------------------------------------------------------- Chapter 3:4.3

The degree of risk facing the Social Security trust fund under a
stock investment policy would depend, in part, on the relationship
between stock returns and the trust fund's predominant revenue
source--payroll taxes.  Like stock returns, payroll tax revenues
fluctuate with changes in overall economic conditions.  If stock
returns tend to be high when payroll tax revenues drop, the trust
fund theoretically could reduce its overall risk by diversifying into
stocks.  If, however, stock returns move in tandem with payroll tax
revenues and tend to fall during recessionary periods, the trust fund
would face greater risk investing in stocks.  A general stock market
downturn coinciding with a payroll tax shortfall would exacerbate
Social Security's need for cash to pay benefits.  One economic study,
done from the perspective of the government as a whole, concluded
that stock returns and tax revenues are positively correlated.\26 The
Advisory Council's technical panel reported in September 1995 that
further research on the relationship between stock returns and
payroll tax revenues is critical in evaluating whether stock
investing is appropriate for the Social Security trust fund. 


--------------------
\26 Henning Bohn, "Tax Smoothing with Financial Instruments," The
American Economic Review, Vol.  80, No.  5 (December 1990), pp. 
1217-1230. 


      WHO BEARS THE INVESTMENT
      RISK OF THE SOCIAL SECURITY
      TRUST FUND? 
-------------------------------------------------------- Chapter 3:4.4

The Social Security trust fund could expect to earn a higher return
by diversifying into stocks, but it is reasonable to anticipate that
its return could be lower than the long-term average market
return.\27 As our simulations illustrate, as long as its return on
stocks is greater than the expected return on special Treasury
securities, the trust fund would be able to pay benefits longer than
is possible under the current investment policy.  If, however, the
real return on stocks over the next 20 or 30 years averages less than
the expected return on Treasury securities or is negative, the trust
fund would be exhausted sooner than in 2029, exacerbating Social
Security's long-term financial imbalance. 

The increased risks associated with the Social Security trust fund's
stock investments would be borne collectively through the government
and ultimately by taxpayers.  According to recent research, the
increased risk of any shortfall if stock investing does not work as
expected would be borne largely by future taxpayers.\28 Even if the
trust fund earns more investing in stocks, the full gain may not
accrue to future generations.  The prospect of higher returns from
stock investing, as shown in our simulations, could be used to delay
benefit reductions or tax increases for current generations; or,
there even could be pressure to cut Social Security taxes or raise
benefits now. 


--------------------
\27 This discussion focuses on the real possibility that the trust
fund would earn less than the 7 percent real return assumed by the
Advisory Council and used in our simulations.  Alternatively, it is
theoretically possible that the trust fund could earn more. 

\28 For research on intergenerational risk-sharing, see (1) Henning
Bohn, "Social Security Reform and Financial Markets," Social Security
Reform:  Links to Saving, Investment, and Growth, Federal Reserve
Bank of Boston Conference Series No.  41, June 1997, (2) Kent
Smetters, "Investing the Social Security Trust Fund in Equities:  An
Option Pricing Approach," Congressional Budget Office Technical Paper
1997-1, August 1997, and (3) Peter A.  Diamond, "Macroeconomic
Aspects of Social Security Reform," Brookings Papers on Economic
Activity, 2:1997. 


   STOCK INVESTING COULD
   COMPLEMENT OTHER SOCIAL
   SECURITY REFORMS
---------------------------------------------------------- Chapter 3:5

Investing in the stock market is one option to increase the trust
fund's revenues, but by itself, is not the solution to Social
Security's financing problem.  Stock investing would have a
relatively modest impact on long-term solvency as long as Social
Security remains largely a pay-as-you-go program.  Restoring Social
Security's long-term solvency will require some combination of
benefit reductions and revenue increases.\29 Recent reform proposals,
such as those the Advisory Council suggested, would increase the
funds set aside to pay for future Social Security benefits.  As part
of a reform package that moves towards more advance funding, stock
investing could have a more significant effect on Social Security's
long-term financing. 

Stock investing could complement traditional reforms in that higher
stock returns could serve to reduce the size of benefit cuts and/or
tax increases needed to restore Social Security's financial solvency. 
Increased funding generated by other program reforms could be
invested in the stock market to earn a higher rate of return, further
boosting Social Security revenues.  Typically, reform proposals
incorporating stock investments envision that the Social Security
trust fund would hold its stock portfolio and mainly draw on its
stock earnings.  In this context, the trust fund would be less
vulnerable to the risk inherent in liquidating stocks to pay promised
benefits. 

However, caution is warranted when counting on future stock returns
in designing a Social Security reform package.  If stocks do not
deliver the expected returns, then Social Security could again face a
financial shortfall, and the trust fund might have to quickly sell
its stocks to pay benefits.  In light of the variability of stock
returns and the range of possible outcomes for the trust fund,
investing in the stock market may not necessarily bolster public
confidence in Social Security's finances.  Even if the trust fund
earns a solid return over the long haul, short-term fluctuations
could heighten public concern about the stability of the program's
financing. 


--------------------
\29 Social Security:  Restoring Long-Term Solvency Will Require
Difficult Choices (GAO/T-HEHS-98-95, February 10, 1998). 


ECONOMIC AND MARKET EFFECTS OF
GOVERNMENT STOCK OWNERSHIP
============================================================ Chapter 4

The economic effects of government stock investing would likely be
minimal because stock investing by itself does not increase national
saving.\1 Without additional saving, government stock investing might
improve the finances of the Social Security trust fund, but it would
not increase the size of the economy because it would represent an
asset shuffle between the government and private investors.  This
asset shuffle would likely be accompanied by changes in bond and
stock prices that might, to some extent, undercut the government's
expected gains on stock investments and increase the government's
cost of borrowing.  While government stock investments might be small
in relation to the total stock market, the federal government could
become the largest single investor within a few years. 

Investing in a broad-based index to diversify the government's
portfolio would reduce the risk of losses and administrative costs,
but it has price effects of its own and its potential to prevent the
use of nonfinancial objectives in stock selection is limited.  Even
if the government chooses an indexing strategy to select its stock
investments, the issue of how to handle stock voting rights remains. 
The perception of political influence over the government's stock
investments could undermine public confidence in the government's
handling of Social Security's finances. 


--------------------
\1 The economic effects of allowing the trust fund to invest in other
non-Treasury assets, such as bank certificates of deposit or
municipal bonds, would be similar to the effects that are described
in this chapter for stock investing. 


   ECONOMIC EFFECTS WOULD LIKELY
   BE MINIMAL
---------------------------------------------------------- Chapter 4:1

The economic effects of government stock investing would likely be
minimal because stock investing by itself does not increase national
saving.\2 Without additional national saving and investment,
government stock investing might improve the financial position of
the Social Security trust fund but would not increase the productive
capacity of the economy.  There would be no additional saving
because, without other changes in federal spending or revenue, the
Treasury would have to borrow more from the public to offset Social
Security's stock purchases.\3 On balance, private investors would end
up with fewer stocks and more government bonds, while the government
would then hold some stocks and fewer of its own bonds.  Any higher
returns earned by the government would otherwise have accrued to
other investors.  So, while the public might gain by a higher return
to the trust fund, people would have to accept a lower return on
their privately held assets. 

There may also be offsetting increases and decreases in income to
investors, the Social Security trust fund, and the U.S.  Treasury due
to changes in interest rates and stock yields.  According to
analysts, the asset shuffling between the government and other
investors could lead to higher stock prices and interest rates.  The
magnitude of price changes is uncertain and could be small.  Any
price effects might occur even before the government starts buying
shares of stocks.\4 In the stock market, the new government demand
for stocks could tend to increase stock prices, inducing some private
investors to sell stocks to the government.  In the bond market, the
Treasury likely would have to issue additional bonds to the public to
substitute for Social Security's stock investments.  This increase in
Treasury borrowing from the public could tend to reduce bond prices
or, equivalently, raise their rate of interest, which would raise
federal interest costs.\5 These changes would tend to redistribute
income to a degree, but would not provide the public in general with
any additional income. 

The initial effects of price changes could be reduced as firms
increase their equity financing and foreign buyers increase their
purchases of U.S.  securities.  Looking to the future when members of
the baby-boom generation reach retirement, analysts have expressed
concern that at about the same time that the Social Security trust
fund expects to liquidate its assets, pension funds, which have been
a major source of national saving, are expected to become enormous
net sellers of assets.\6 Unexpected, sizable stock sales by pension
funds and the government, such as those described in the investment
scenarios discussed in chapter 3, could tend to depress stock prices. 
It is difficult to predict precisely how stock prices and interest
rates might be affected by the simultaneous sales of pension fund and
Social Security's stock holdings more than a decade in the future. 


--------------------
\2 See (1) Alan Greenspan, "Remarks at the Abraham Lincoln Award
Ceremony of the Union League of Philadelphia," December 6, 1996, (2)
Congressional Budget Office, Implications of Revising Social
Security's Investment Policies, September 1994, (3) "Economic
Challenges of an Aging Population," Chapter 3, The Annual Report of
the Council of Economic Advisers, February 1997, and (4) Eric M. 
Engen and William G.  Gale, "Effects of Social Security Reform on
Private and National Saving," Social Security Reform:  Links to
Saving, Investment, and Growth, (Federal Reserve Bank of Boston,
Conference Series No.  41, June 1997). 

\3 This statement applies in a situation where the federal government
has a unified budget deficit after investing in the stock market. 
See chapter 5 for a complete discussion of the budgetary effects of
government stock investments and the implications for federal fiscal
policy and government saving. 

\4 According to financial market analysts, stock markets are
reasonably efficient in that stock prices would incorporate any
relevant information about government stock investing as soon as it
is publicly available.  See Hendrik S.  Houthakker and Peter J. 
Williamson, The Economics of Financial Markets (Oxford, New York: 
Oxford University Press, 1996). 

\5 Higher interest rates would be paid on debt roll-overs and new
borrowing.  According to the Treasury, $3 trillion of marketable debt
is held by private investors, about one-third of which matures within
1 year and over half of which matures within 2 years. 

\6 See (1) Sylvester J.  Schieber and John B.  Shoven, "The
Consequences of Population Aging on Private Pension Fund Saving and
Asset Markets," National Bureau of Economic Research Working Paper
No.  4665, March 1994, and (2) Report of the 1994-1996 Advisory
Council on Social Security, Volume II,
pp.  56-58. 


   THE GOVERNMENT COULD BECOME THE
   LARGEST STOCK INVESTOR
---------------------------------------------------------- Chapter 4:2

Proponents of government stock investing and some analysts believe
that investing some Social Security funds in equity securities would
not have a major disruptive effect on the stock market.\7 According
to an analysis prepared for the Advisory Council, Social Security's
stock portfolio under the Maintain Benefits plan would probably
account for a relatively small share of the stock market.  The
Advisory Council's analysis estimated that the share would be less
than 3.5 percent of the value of all stocks, assuming that the stock
market would grow at the real stock yield of 7 percent.\8 Starting
with the stock market value of $12.8 trillion at the end of the third
quarter of 1997 and using the Advisory Council's assumption about
market growth, our aggressive scenario would peak at about 3.5
percent of the stock market. 

While the government's investments might be small in relation to the
stock market, their size would probably be larger than the
investments of other investors.  Under our aggressive investment
scenario, the federal government could become the largest single
stock investor within several years.  If the Social Security trust
fund invests both its future cash surplus and interest in the stock
market, assuming a 7 percent real rate of return, the federal
government's stock portfolio would be estimated at more than $80
billion by the end of 1998, more than $170 billion by year-end 1999,
and nearly $275 billion by year-end 2000.  As a point of comparison,
table 4.1 shows the five largest pension funds and money managers. 
The largest money manager held $522 billion at the end of 1996, and
the largest pension fund held almost $128 billion as of September 30,
1997. 



                               Table 4.1
                
                    Top Five Pension Funds and Money
                    Managers Ranked by Total Assets

                         (Dollars in billions)

Pension fund/money manager                                Total assets
--------------------------------------------------  ------------------
Top five pension funds/sponsors\a
----------------------------------------------------------------------
California Public Employees                                       $128
New York State Common                                              $96
General Motors                                                     $91
California State Teachers                                          $79
Florida State Board                                                $72

Top five money managers\b
----------------------------------------------------------------------
Fidelity Investments                                              $522
Barclays Global Investors                                         $386
Prudential Insurance                                              $333
State Street Global                                               $301
Bankers Trust                                                     $227
----------------------------------------------------------------------
\a Total assets at September 30, 1997. 

\b Total assets at December 31, 1996. 

Source:  Pensions & Investments, May 12, 1997, and January 26, 1998. 


--------------------
\7 See (1) Robert M.  Ball, Edith U.  Fierst, Gloria T.  Johnson,
Thomas W.  Jones, George Kourpias, and Gerald M.  Shea, "Social
Security for the 21st Century," in Report of the 1994-1996 Advisory
Council on Social Security, Volume I, pp.  83-86, (2) P.  Brett
Hammond and Mark J.  Warshawsky, "Investing Social Security Funds in
Stocks," Benefits Quarterly (1997), and (3) Francis X.  Cavanaugh,
"The National Debt and Social Security," chapter 8, The Truth About
the National Debt (Boston, Massachusetts:  Harvard Business School
Press, 1996). 

\8 The Advisory Council's analysis started with a stock market value
of $8 trillion.  P.  Brett Hammond and Mark J.  Warshawsky, in
"Investing Social Security Funds in Stocks" (1997), compared
government stock investments to the indexed equity portion of the
stock market and concluded that there may be a need to strengthen the
institutional capacity of this portion of the stock market before
proceeding with a large Social Security investment program. 


   BENEFITS AND LIMITATIONS OF
   STOCK INDEXING
---------------------------------------------------------- Chapter 4:3

Proposals for government stock investing typically recommend
investing in a broad-based stock index to diversify the government's
stock portfolio, reducing the likelihood of concentrating investments
in individual companies, and to reduce administrative costs. 
Purchasing stocks listed in an index is a form of passive investing
that seeks to match the performance of the group of securities listed
in a market benchmark, or index.  Figure 4.1 defines several widely
recognized stock indexes, such as the Dow Jones Industrial Average,
S&P 500, and Wilshire 5000. 

   Figure 4.1:  Examples of Widely
   Recognized Stock Indexes

   (See figure in printed
   edition.)


      INDEXING WOULD REDUCE SOME
      RISKS
-------------------------------------------------------- Chapter 4:3.1

Indexing reduces risk or exposure to loss associated with an
individual company failing and industry-specific downturns.  The
securities held in a broadly-based indexed portfolio would represent
many different sectors of the economy and many individual companies. 
This diversification reduces the risk that any loss related to the
performance of an individual security or group of securities would
greatly affect the overall performance of the portfolio.  In spite of
the diversification benefits of indexing, stock investing would still
be riskier than the government's current investment in special
Treasury securities.  Inclusion in a stock index does not represent
any guarantee about a company's future performance, and indexing
across the stock market will not reduce the government's risk of loss
in the event of a general stock market downturn. 


      INDEXING WOULD REDUCE
      ADMINISTRATIVE COSTS
-------------------------------------------------------- Chapter 4:3.2

Index investing has grown more popular in recent years as empirical
studies have shown that active stock managers generally underperform
the stock market due to their high transaction costs.  Active
managers incur higher expenses in the process of doing research and
trading the stocks of companies or industries that could be
undervalued or offer good growth potential.  In contrast, index
managers generally do not research individual companies, and the
securities in an index are not changed frequently, both of which
result in lower trading costs.  As a result, the costs of managing an
index of large company stocks, like the S&P 500, typically average
about 10 basis points, while the active management equivalent may
cost 40 to 50 basis points or more when all costs are included.\9

Most of the cost of managing an S&P 500 index fund is attributable to
the cost of maintaining thousands of shareholder accounts.  With
government stock investing there would be only one account to
maintain; thus, it would incur negligible costs as a percentage of
assets.  For this reason, the Advisory Council assumed that annual
administrative costs on the government's stock investments would be
only one-half a basis point of total assets.\10

Active managers may have an advantage over index managers in
selecting small and medium corporations because data on these
companies are less widely available, giving active managers with
research capabilities more opportunity to find companies with growth
potential.  Also, index managers that try to match an index of small
companies will incur larger transaction costs than a large company
index because the stocks tend to be lower priced and less liquid. 


--------------------
\9 For example, the federal Thrift Savings Plan's (see figure 4.2)
1997 total expenses divided by the average monthly fund balance were
7 basis points for the government securities fund, 7 basis points for
the stock index fund, and 8 basis points for the bond index fund. 

\10 According to Joel Dickson's analysis prepared for the Advisory
Council, fees charged for managing a broad stock index within the
Maintain Benefits plan could be less than 1 basis point; moreover, if
the investment managers are allowed to earn profits through
securities lending activities, management expenses could be bid down
to zero.  Report of the 1994-1996 Advisory Council on Social
Security, Volume II, p.  487. 


      INDEXING CREATES PRICE
      EFFECTS OF ITS OWN
-------------------------------------------------------- Chapter 4:3.3

Large company indexes, like the S&P 500, have become vulnerable to
price disruptions as they have become more popular.  Studies have
shown that newly included companies in the S&P 500 tend to appreciate
in price by about 5 percent simply because they become part of the
index.  This occurs because managers who run index funds linked to
the S&P 500 have to purchase the stocks of the new companies in the
index so that their performance will continue to replicate the S&P
500.  Additions and deletions from certain indexes, like the Dow
Jones Industrial Average and S&P 500, represent staff's judgments
about an individual company's relative standing and do not
necessarily reflect a change in its financial outlook. 

Choosing a broader market index, like the Wilshire 5000, could reduce
price distortions somewhat by diversifying across thousands of medium
and small capitalization corporations.  Investing in a broad-based
market index, however, could lead to active selection of stocks by
the government and/or its stock managers.  Managers that track the
performance of a broad market index typically engage in sampling
among thousands of small capitalization stocks to reduce transaction
costs.  For example, Vanguard's Wilshire 5000 index replicates the
top 1,400 companies and takes a sample of the rest of the companies. 


      NONFINANCIAL OBJECTIVES
      COULD STILL INFLUENCE STOCK
      SELECTION UNDER INDEXING
-------------------------------------------------------- Chapter 4:3.4

Analysts and critics of government stock investing have expressed
concern that political objectives would influence Social Security's
investments.\11 While indexing greatly reduces this possibility, it
does not eliminate it because the government, as the owner of the
investments, would be able to drop or add shares of certain companies
from a stock index to achieve other public goals.  The Congress could
start with a broad indexing strategy to reduce price distortions and
the possibility of owning a significant percentage of the shares of
any individual company.  The Congress could then modify the index's
composition to target investments that could offer competitive
returns while providing other social benefits.\12 Additionally, a
tailored index could automatically exclude stock investments in
companies engaged in activities that are in conflict with government
policies.  For example, some state governments have been disinvesting
from tobacco companies because they are pursuing a lawsuit against
them.  According to critics of government stock investing, pressures
to exclude or include stocks for nonfinancial reasons might reduce
the rate of return on the government's portfolio and hinder the
overall economic efficiency of capital markets. 

The issue of how to select stock investments also emerged when the
federal employees' Thrift Savings Plan (TSP) was created (see figure
4.2).  To eliminate political influence in TSP's stock investment
decisions, the Congress restricted TSP investments to widely
recognized broad-based stock indexes, prohibited TSP board members
and employees from exercising stock voting rights, and subjected TSP
board members and employees to strict fiduciary rules.  These
fiduciary rules require board members and employees to invest the
money and manage the funds solely for the benefit of the owners of
the funds--the participating federal employees and beneficiaries. 
Board members and employees cannot consider nonfinancial objectives,
such as job creation and environmental protection, in selecting
stocks, and the portfolio composition is automatically determined by
the stock index chosen.  Breaching their fiduciary duty would expose
board members and employees to civil and criminal liabilities. 

   Figure 4.2:  Overview of the
   Federal Thrift Savings Plan

   (See figure in printed
   edition.)

Some analysts believe that TSP's fiduciary rules may not be
applicable to government stock investing.  The government owns the
assets in the Social Security trust fund, and the Congress would be
able to define the federal government's "best interests."\13 Other
analysts believe that it might be possible to establish in law that
investments be made solely for the financial benefit of Social
Security participants and not for other economic, social, or
political objectives.  And, analysts believe that the Congress could
establish an investment policy board, similar to TSP's board, subject
to strict fiduciary rules when investing and managing the
government's money. 


--------------------
\11 A restriction to invest only in publicly listed stocks would
automatically favor the group of large, publicly held corporations
over thousands of small and medium businesses that do not have
publicly held stock.  See Lawrence J.  White, "Investing the Assets
of the Social Security Trust Funds in Equity Securities:  An
Analysis," Investment Company Institute Perspective, May 1996.  Other
critics include (1) Joan T.  Bok, Ann L.  Combs, Sylvester J. 
Schieber, Fidel A.  Vargas, and Carolyn L.  Weaver, "Restoring
Security to Our Social Security Retirement Program," Report of the
1994-1996 Advisory Council on Social Security, Volume I, 1997, (2)
Michael Leidy, "Investing U.S.  Social Security Trust Fund Assets in
Private Securities," IMF Working Paper 97-112, and (3) Gene Steuerle,
"Investing Social Security Surpluses in the Stock Market," Tax Notes,
April 3, 1995. 

\12 See Theodore J.  Angelis, "Investing Public Money in Private
Markets:  What Are the Right Questions," Framing the Social Security
Debate:  Values, Politics, and Economics, (National Academy of Social
Insurance Conference, January 29-30, 1998). 

\13 Securities and Exchange Commission economists suggested that
appropriate safeguards to preserve the integrity of stock markets
against politically motivated investments should be in place before
the government invests in the stock market. 


      COMPLEXITIES OF STOCK VOTING
      WOULD REMAIN UNDER INDEXING
-------------------------------------------------------- Chapter 4:3.5

Even if the government chooses an indexing strategy to select its
stock investments, the issue of how to handle stock voting rights
remains.  In general, because index investors cannot alter their
portfolio composition to increase financial performance, they have a
stronger incentive to actively exercise stock voting rights.  Instead
of selling stocks of an underperforming company, an index investor
can choose to participate in corporate decisions that could enhance
the company's performance.  However, critics have expressed concern
that the government's right to vote its sizable number of shares
would allow it to influence corporate decisions.\14

To blunt critics' concerns about potential federal meddling in
corporate affairs, the government could choose not to exercise its
stock voting rights.  The Congress could, by statute, expressly
prohibit the government from directly exercising voting rights
associated with its ownership of securities or delegating these
rights to stock managers.  However, prohibiting the exercise of
voting rights by the government or its stock managers would, in
effect, create a situation favoring certain stockholders and
corporate managers.  If, for example, the government does not
exercise its voting rights, other stock owners would have their own
unencumbered voting rights increased and could conceivably act in
ways that take advantage of the government's passivity.  Also, no
matter what stock voting policy is chosen when the government begins
investing, policymakers could change the rules for stock voting
rights in the future. 

Despite these complexities, analysts and Maintain Benefits plan
supporters believe that some features of TSP might be used to achieve
neutrality of Social Security investments in matters of corporate
policy.  For example, voting on Social Security stock might be done,
as for TSP, by investment managers according to their own guidelines. 
Since delegating the government's stock voting rights would shift the
government's sizable voting rights to its selected managers, Maintain
Benefits plan supporters have suggested spreading the government
assets among many stock index managers.\15

Investing in stocks or other assets outside the government would be a
new concept for the Social Security trust fund.  How the government
chooses to handle the implementation issues associated with stock
investing would affect the public's perception of such a new
investment policy.  If political factors appear to influence stock
selection, stock voting, or the selection of stock managers, the
public confidence in government's handling of Social Security's
finances could be undermined. 


--------------------
\14 Even a 2 or 3 percent block of shares could allow substantial
influence over the policies of publicly traded companies.  An
activist shareholder can disproportionately magnify the power of
small holdings in a company.  See Angelis, pp.  26-27. 

\15 Spreading the government's stock portfolio among many managers
might also help to address a concern raised by Securities and
Exchange Commission economists that government stock investing could
have an impact on competition and efficiency in the stock market
because the federal government's stock managers would have enormous
clout through their ability to place billions of dollars in orders to
purchase and sell shares of stocks. 


EFFECTS OF GOVERNMENT STOCK
INVESTMENTS ON THE FEDERAL BUDGET
AND FISCAL POLICY
============================================================ Chapter 5

Under current budget scoring rules, in the short term stock investing
would increase the reported unified deficit or decrease any reported
surplus, because stock purchases would be treated as budget outlays. 
Any money used to purchase stocks would no longer be invested in
Treasury securities, reducing the Treasury's available cash and
making the resulting budget measure look more like the on-budget
deficit that excludes Social Security's finances.  Each dollar
invested in stocks is $1 less available to the Treasury to finance
other government spending or reduce debt held by the public.  If
after accounting for this effect the federal government were in
deficit, the Treasury would need to borrow from the public to replace
cash used to buy stocks, unless offsetting spending or revenue
actions were taken.  Despite the change in the reported
deficit/surplus, stock investing would not significantly change the
government's fiscal position.  Any additional borrowing from the
public would be accompanied by less borrowing from Social Security,
leaving the federal government's gross debt largely unchanged.  And,
more important, the government's impact on the economy would not
change significantly because the effects on national saving of
additional borrowing from the public would be offset by the purchase
of stocks. 

Whether the higher reported deficits or lower reported surpluses
would lead to changes in fiscal policy is unclear.  Since the change
in the deficit/surplus would not alter the government's fiscal
position, policymakers might choose to maintain the status quo in
fiscal policy.  In fact, recognizing that purchasing stocks means
acquiring financial assets, they might consider changing budget
scoring rules so that stock purchases would not count as budget
outlays.  If scoring rules were changed in this way, the reported
deficit/surplus would not be significantly affected by stock
investing.  Alternatively, the higher reported deficits or lower
reported surpluses could make more visible the underlying condition
of the government's finances excluding the Social Security surplus
and prompt policymakers to initiate compensating spending or revenue
actions.  If so, such fiscal actions could raise national saving. 
Or, if the net effect of stock investing were to reduce or eliminate
an anticipated unified surplus, policymakers might be reluctant to
devote surplus funds to additional spending or tax cuts.  In this
case, fiscal restraint might not promote higher saving, but it would
avoid policy actions that could cause saving to decline. 

Of course, if policymakers wanted to take actions to promote higher
national saving, they could certainly do so directly by running
annual surpluses in the unified budget and devoting the surplus funds
to reducing the level of outstanding debt held by the public.  While
sustaining surpluses would likely prove challenging, such a policy
would strengthen the government's fiscal condition and enable it to
better handle the baby boomers' retirement costs. 


   IMPACT OF STOCK INVESTING ON
   REPORTED DEFICITS/SURPLUSES
---------------------------------------------------------- Chapter 5:1

Under current budget scoring rules, stock purchases would be treated
as budget outlays.  Therefore, in the short term, investing surplus
Social Security funds in stocks would make unified budget deficits
larger or unified budget surpluses smaller.\1 Each dollar invested in
stocks is $1 less available to the Treasury to finance other
government spending or reduce debt held by the public.  If, after
accounting for the effects of stock investing, the government were in
deficit, the Treasury would have to borrow more from the public,
unless action were taken to reduce other spending or raise revenues. 
If, instead, after adjusting for the effects of stock investing, the
government were running a budget surplus, the Treasury would have
less cash available to reduce debt held by the public. 

Depending on how much the Social Security trust fund invests in
stocks, the impact on the budget deficit or surplus could be
substantial.  If Social Security's cash surplus were invested, the
government's budgetary balance would be expected to decline by about
$30 billion to $35 billion\2 annually for the next several years.\3
For example, the $8 billion surplus that the Congressional Budget
Office (CBO) projects for fiscal year 1998\4 would be eliminated,
leaving a deficit of about $25 billion in its place.  If the trust
fund were to also invest its annual interest, the change in the
annual deficit or surplus could exceed $100 billion a year.  Under
current CBO projections, investments of this magnitude would prevent
the realization of any unified budget surpluses over the next
decade.\5 However, as discussed below, this deterioration in the
reported deficit/surplus would not significantly affect the
government's fiscal position, and, over the long term, the effect of
stock investing on the reported deficit/surplus could largely be
neutral. 


--------------------
\1 Since, in the short term, stock investing would worsen the
reported budgetary balance of the government, it could also change a
unified surplus into a deficit depending on the size of the initial
surplus and the amount that Social Security invests in stocks. 

\2 GAO analysis of 1997 Trustees' Report, intermediate assumptions. 

\3 This amount excludes the effects of the additional interest
payments to the public from any increased borrowing. 

\4 Congressional Budget Office, An Analysis of the President's
Budgetary Proposals for Fiscal Year 1999, March 1998. 

\5 Ibid. 


      HIGHER SHORT-TERM DEFICITS
      UNDER STOCK INVESTING WOULD
      NOT SIGNIFICANTLY AFFECT
      NATIONAL SAVING
-------------------------------------------------------- Chapter 5:1.1

Higher short-term deficits would mean more borrowing from the public
and less borrowing from the Social Security trust fund.\6 This shift
from one type of debt to another would not significantly change the
federal government's gross debt.\7 However, the annual change in the
gross debt is generally less noticeable than the annual change in
borrowing from the public,\8 which is reflected in the unified
deficit.  It is the unified deficit that is most commonly used by
policymakers, analysts, and the media because it best reflects the
current impact of federal borrowing on the financial markets and the
economy.  Therefore, changing Social Security's investment policy
would increase the type of federal borrowing that is most closely
monitored. 

Although stock investing would increase federal debt held by the
public, it would not by itself significantly affect national saving. 
Ordinarily, government deficits reduce national saving because
government borrowing absorbs money that would otherwise have been
available for private investment.  Therefore, the current unified
deficit measure closely approximates the economic effects of federal
borrowing.  However, under stock investing, although the reported
deficit would increase, the government's draw on capital markets
would not.  While the additional federal borrowing from the public
would absorb money from capital markets, the trust fund's stock
investments would add funds to the markets.  This new flow of money
from the trust fund would offset the additional borrowing from the
public, resulting in no significant net change in the annual funds
available for private investment.\9 (See figure 5.1.)

   Figure 5.1:  No Net Change in
   National Saving with Government
   Stock Investing

   (See figure in printed
   edition.)

Note:  Figure excludes the effects of additional interest paid to the
public. 


--------------------
\6 This section discusses stock investing's impact in an environment
of budget deficits.  Although the effects of stock investing in a
budget surplus environment would differ technically, they would be
the same conceptually.  For example, if after accounting for stock
investing, the government were still in a surplus, it would
potentially have cash available to reduce the level of debt held by
the public.  However, this result does not change the fact that the
immediate effect of stock investing is to remove cash from the
Treasury, which means that, absent stock investing, the government
would have had more money available for such debt reduction. 

\7 If the government borrowed more from the public in response to
government stock investing, the federal government would pay more
interest to the public while crediting less to the trust fund.  This
change in interest flows would probably have no significant effect on
the gross debt. 

\8 The debt limit provides an exception to this statement, because it
applies not just to debt held by the public, but to nearly all of the
gross debt.  In this case, changes in the gross debt are more
noticeable than changes in debt held by the public. 

\9 As long as the government were in a deficit position, the Treasury
would have to borrow more to pay the interest costs on additional
borrowing from the public.  This borrowing to pay interest would
slightly reduce national saving. 


      LONG-TERM EFFECTS OF STOCK
      INVESTING ON REPORTED
      DEFICITS/SURPLUSES COULD
      LARGELY BE NEUTRAL
-------------------------------------------------------- Chapter 5:1.2

While stock investing would have a negative effect on the budget
deficit/surplus today, over time its impact on the unified budget
could largely be neutral.  As with any budget outlay, the purchase of
a stock would mean money flowing out of the government and, thus, the
reported budget deficit/surplus would deteriorate.  However, the sale
of a stock would mean money flowing back into the government.  So,
when Social Security begins running cash deficits in the future, it
could sell stocks to finance benefits, rather than drawing on the
Treasury.  This approach would result in smaller future budget
deficits or larger future budget surpluses than under current policy. 
This longer-term improvement could offset the near-term deterioration
in the deficit/surplus. 

Because it is impossible to predict the precise effects of stock
investing, it is important to qualify any discussion of its fiscal
impact.  For example, over the long term, stock investing could have
at least a slight positive effect on the federal government's
finances if the earnings on stock investments exceed any potential
increase in federal borrowing costs.\10 If the federal government
borrows from the public to replace cash used to purchase stocks, it
could put upward pressure on interest rates.  Even a small rise in
interest rates could significantly increase federal interest costs
because of the size and structure of the outstanding debt held by the
public.  About $1 trillion of the federal government's outstanding
debt securities are rolled over every year.  Higher interest rates
would affect these roll-overs, as well as the additional debt issued
to cover the stock purchases.\11

If the federal government's stock earnings did exceed any increased
borrowing costs, the government would benefit from an improved
budgetary position.  However, any such gain to the government would
not increase national income.  If the federal government were
investing in stocks without taking other actions designed to raise
national saving, it would be capturing a portion of stock returns
that would otherwise have accrued to private investors, who would now
own fewer stocks and more Treasury securities.  In short, simply
altering the ownership of stocks and bonds between the public and
private sectors would not boost long-term economic growth. 


--------------------
\10 Stock investing would have several additional effects on federal
spending and revenue.  These effects would be relatively small and
could have either a positive or negative net impact on the budget
deficit/surplus.  For example, the investment earnings on the
government's stock portfolio would be exempt from taxation.  However,
the additional interest income earned by federal bondholders would be
subject to taxation. 

\11 Higher interest rates would also affect any new issues of the
nonmarketable Treasury securities held by the Social Security trust
fund or any old issues that are rolled over. 


   INDIRECT IMPACT OF GOVERNMENT
   STOCK INVESTING ON FISCAL
   POLICY IS UNCLEAR
---------------------------------------------------------- Chapter 5:2

Whether the short-term increase in reported budget deficits or
decline in reported budget surpluses would lead to changes in fiscal
policy is unclear.  Since stock investing would not substantially
change the impact of federal finances on the economy, policymakers
might choose to maintain a status quo fiscal policy.  Changing budget
scoring rules to exclude stock purchases from budget outlays would be
consistent with this approach because it would result in no
significant change in the reported deficit/surplus.  If budget
scoring were not altered, however, stock investing would more clearly
reveal the budget imbalance that exists when Social Security's
temporary surplus is excluded.  The greater visibility of this
imbalance could lead to further actions to restrain federal spending
or increase revenues, which could boost national saving.  Of course,
stock investing would be a very indirect way of prompting additional
actions to raise national saving.  If policymakers choose to enact
additional fiscal restraint, they can do so directly without any
changes in the way the unified budget deficit/surplus is presented. 


      STOCK INVESTING COULD PROMPT
      CHANGE IN BUDGET SCORING
      RULES
-------------------------------------------------------- Chapter 5:2.1

While stock investing could lead to additional federal borrowing from
the public, the additional borrowing would not substantially reduce
the pool of funds available for private investment.  Any increased
borrowing would be largely offset by the government's stock
purchases.  Since stock investing does not significantly alter the
impact of the government's finances on the economy, policymakers
might decide to maintain current fiscal policy as is.  In fact, they
could choose to change budget scoring rules to explicitly recognize
the distinct nature of stock purchases.  Since the government
acquires a financial asset when it buys stock, one could argue that
the purchase should not be treated as a budget outlay.  However, such
a determination would conflict with the way other asset purchases are
treated in the budget--they are generally scored as outlays.\12 And,
in addition, creating different budget scoring rules for stocks would
also raise some complicated technical issues, such as how to
recognize changes in their market values.  To some extent, new
budgetary scoring procedures would have to be developed for handling
stocks.  If, despite these considerations, stock purchases were not
treated as outlays, stock investing would have no major impact on the
reported budget deficit or surplus.\13 (For additional details on
stock investing and budgetary accounting, see appendix II). 


--------------------
\12 Major exceptions are debt transactions (for example, when Social
Security invests in special Treasury securities) and cases in which
the government purchases assets that are considered equivalent to
cash. 

\13 If stock purchases were not treated as outlays, stock investing
could still result in minor changes in the budget deficit/surplus. 
For example, if stock investing resulted in additional borrowing from
the public, the associated interest costs would increase budget
outlays and, hence, the deficit. 


      FINANCIAL STATUS OF
      GOVERNMENT EXCLUDING SOCIAL
      SECURITY SURPLUS WOULD BE
      MORE VISIBLE UNDER STOCK
      INVESTING
-------------------------------------------------------- Chapter 5:2.2

If current budget scoring rules were maintained, stock investing
would make more visible the underlying condition of the government's
finances excluding Social Security's temporary surplus.  Currently,
in the unified budget presentation, the Social Security surplus masks
the financial status of the rest of the government.  By helping to
finance current spending, Social Security's cash surpluses may result
in the government spending more or taxing less than it would if these
surpluses were not available to finance other programs.\14 Stock
investing, by removing surplus Social Security funds from the
Treasury would reduce or eliminate the masking effect, making the
unified budget measure look more like the on-budget measure that
appears in budget documents but is not widely used.  If Social
Security's cash surplus plus interest were invested in stocks, the
"new" unified budget measure would virtually match the on-budget
measure.\15

The masking that occurs under current policy is an important
budgetary issue because of its implications for the future.  Social
Security's surpluses are temporary; when they disappear, there will
be relatively fewer funds available to support other government
activities.  At that time, the federal government would need to find
an alternative source of funds if it wanted to maintain the same
amount of spending.  As the trust fund slips into deficit and begins
drawing on the Treasury in increasing annual amounts, the challenge
of financing the rest of the budget would intensify.  (See figure
1.1.) The fiscal choices needed to redeem the trust fund's Treasury
securities could be especially difficult because the need to make
such trade-offs may not be well understood by the public.  Current
fiscal policy may be creating the misleading impression that today's
spending levels and commitments can be maintained indefinitely
without additional revenues or borrowing from the public.  More
clearly acknowledging the government's budgetary status excluding
Social Security's surplus, rather than deferring that recognition
until the trust fund begins drawing on the Treasury, could dispel
this impression. 

Our concern about the impact of Social Security's temporary surplus
on the unified budget does not imply that the program's revenues and
expenditures should be excluded from the budget.  The unified budget,
which includes Social Security, is the way to measure the current
impact of the federal government's finances on the nation's economy. 
However, in considering the long-range implications of government
policies, the temporary nature of the Social Security surplus
suggests that an alternative presentation would more clearly reveal
the underlying nature of federal commitments.  Such a presentation
would help today's decisionmakers better understand the long-term
budgetary condition and underlying commitments of the federal
government. 


--------------------
\14 This masking effect could have important implications for any
Social Security reforms.  Reforms that increase the size of the
Social Security surplus might not necessarily improve the long-term
picture for the budget as a whole.  A larger Social Security surplus
might serve to intensify the masking of the financial condition of
the rest of the government.  If policymakers responded to a larger
trust fund surplus by exercising less restraint in the rest of the
budget, the improvement in Social Security's finances would not
contribute to increased national saving.  Instead, it would only
allow the trust fund to build up more claims on the Treasury without
enhancing the nation's ability to meet these future claims. 

\15 In addition to Social Security, the on-budget deficit excludes
the operations of the U.S.  Postal Service.  However, this amount is
very small in comparison to Social Security. 


      HIGHER REPORTED DEFICITS OR
      LOWER REPORTED SURPLUSES
      COULD INFLUENCE FISCAL
      POLICY
-------------------------------------------------------- Chapter 5:2.3

Even though stock investing does not alter the government's fiscal
position, it could indirectly lead to changes in fiscal policy by
focusing more attention on the budget imbalance that exists when
Social Security's surplus is excluded.  Policymakers could react to a
higher unified deficit by cutting spending and/or raising taxes. 
Such fiscal restraint could contribute to a higher level of national
saving.  Or, if the net effect of stock investing were to reduce or
eliminate an anticipated budget surplus, policymakers might be
reluctant to devote surplus funds to additional spending or tax cuts. 
In this case, fiscal restraint might not promote higher saving, but
it would avoid policy actions that could cause saving to decline. 

Though stock investing could help highlight the budget shortfall that
exists when Social Security's surplus is excluded, it represents a
circuitous way of essentially duplicating a measure that already
exists:  the on-budget deficit.  If policymakers wanted to take
actions to boost national saving, they could certainly do so directly
by running annual surpluses in the unified budget and devoting the
surplus funds to reducing the level of outstanding debt held by the
public.  If the government ran a unified budget surplus equal to
Social Security's cash surplus, it would mean that the Treasury would
no longer need to rely on Social Security revenues to finance federal
spending on other activities.  While attaining and sustaining such
surpluses would likely prove challenging, such a policy would
strengthen the fiscal condition of the government and, by promoting
higher saving, better position the economy to handle the baby
boomers' retirement costs. 


ASSUMPTIONS USED IN ESTIMATING HOW
HIGHER RETURNS AFFECT THE SOCIAL
SECURITY TRUST FUND
=========================================================== Appendix I

At our request, the Social Security Administration's (SSA) Office of
the Chief Actuary simulated the potential effect of higher returns of
stock investing on the Social Security trust fund.  Simulations are
useful for comparing alternative investment policies within a common
framework but should not be interpreted as forecasts given the range
of uncertainty about the amount and timing of any Social Security
stock investments as well as about future stock returns and potential
economic changes in response to government stock investing.  While
this report discusses potential investment alternatives, it does not
suggest any particular course of action, since the choice of the most
appropriate investment policy is a decision to be made by the
Congress and the President. 


   STOCK INVESTING IN ISOLATION
   FROM OTHER PROGRAM CHANGES
--------------------------------------------------------- Appendix I:1

We examined the potential effect of stock investing in isolation from
other changes in the Social Security program.  At our request, the
SSA actuaries used the Social Security Trustees' 1997 intermediate
assumptions about future tax revenues, benefit expenditures,
demographic trends, and economic growth.  Under the Trustees' 1997
intermediate estimate, the trust fund's balance of special Treasury
securities at the beginning of 1998 was expected to be equivalent to
about 165 percent of estimated expenditures in that year.  Given that
a level of 100 to 150 percent is suggested as a prudent contingency
reserve, we assumed that the Social Security trust fund could begin
investing in the stock market in 1998.  The trust fund, assuming no
program changes, expects to collect more cash from Social Security
tax revenues than is needed to pay benefits each year until 2012.  At
that point, Social Security's tax revenue will be insufficient to pay
benefits each year, and the trust fund will finance the program's
cash deficit by drawing on its investment income and eventually
depleting its assets. 

We assumed that the trust fund would continue to hold a contingency
reserve equal to at least 100 percent of the next year's expected
expenditures.  Moreover, we assumed that the contingency reserve
would continue to be in the form of special Treasury securities,
given that stock prices are highly variable in the short term.  Under
the Trustees' 1997 intermediate estimate, the trust fund's balance of
special Treasury securities at the beginning of 1998 was expected to
be more than $647 billion.  This balance would be adequate as a
contingency reserve of at least 100 percent of annual expenditures
for about 10 years.  However, as the baby boomers begin collecting
benefits, Social Security's annual expenditures will increase, and
the trust fund will need an increasingly larger amount of Treasury
securities for its contingency reserve.  At our request, the SSA
actuaries assumed that the trust fund's minimum reserve of special
Treasury securities would equal 100 percent of the combined annual
costs of Social Security's two trust fund accounts:  Old-Age and
Survivors' Insurance (OASI) and Disability Insurance (DI).  Combining
these accounts results in a shorter stock investment period than if
the two accounts were treated separately in determining the 100
percent reserve.  On a separate basis, the OASI trust fund account
could hold more stocks for a longer period, but the DI trust fund
account is expected to be exhausted in 2015 under the Trustees' 1997
intermediate assumptions.  In the past, the DI trust fund account has
received additional financing through a reallocation of a portion of
Social Security taxes from the OASI trust fund account. 


   STOCK RETURN ASSUMPTIONS
--------------------------------------------------------- Appendix I:2

The potential gain from stock investing would depend on what future
stock returns are.  To illustrate the potential effect of higher
returns on the Social Security trust fund, we used the 7 percent real
stock yield assumed by the Advisory Council in estimating future
stock performance.  To illustrate how changing the stock return
assumption affects the estimated delay in the trust fund's
exhaustion, we also tested a real stock yield, which is 1 percentage
point less than the historical average.  This alternative return is
intended only to demonstrate that stock investment simulations are
sensitive to the rate of return assumed and does not represent the
lower bound of a confidence interval of expected returns. 

The 7 percent long-term historical average return on stocks is 4.3
percentage points more than the ultimate 2.7 percent yield on special
Treasury securities under the Trustees' 1997 intermediate
assumptions.\1 The real stock yield of 6 percent is 3.3 percentage
points greater than the expected yield on special Treasury
securities.  We also used the Advisory Council's assumption that the
trust fund's annual administrative costs would be 0.5 basis points,\2
which reduces the spread between the real yields on stocks and
Treasury securities by 0.005.\3 Based on the Trustees' 1997
intermediate assumption for inflation, the ultimate nominal yield on
special Treasury securities would be 6.29 percent.  Thus, the
ultimate nominal yields on stocks would be 10.74 percent (assuming a
7 percent real yield) and 9.70 percent (assuming a 6 percent real
yield).  Under the Trustees' intermediate assumptions, the yields on
special Treasury securities are expected to be higher than the
ultimate rate over the next decade or so.  Likewise, the nominal
stock yields over these years are higher than the ultimate stock
yield assumptions, as shown in table I.1. 



                               Table I.1
                
                   Nominal Yields on Special Treasury
                  Securities and on Stocks Assuming 7
                 Percent and 6 Percent Real Yields From
                           1998 Through 2011

                             (Percentages)

                           Special               7 percent   6 percent
                          Treasury              real stock  real stock
Year                         yield   Inflation       yield       yield
----------------------  ----------  ----------  ----------  ----------
1998                          7.41         3.2       11.84       10.81
1999                          7.30         3.2       11.74       10.70
2000                          7.23         3.4       11.67       10.64
2001                          7.15         3.5       11.60       10.56
2002                          7.09         3.5       11.54       10.50
2003                          7.04         3.5       11.49       10.45
2004                          6.98         3.5       11.42       10.39
2005                          6.91         3.5       11.36       10.33
2006                          6.85         3.5       11.29       10.26
2007                          6.73         3.5       11.18       10.14
2008                          6.62         3.5       11.07       10.03
2009                          6.51         3.5       10.96        9.92
2010                          6.40         3.5       10.85        9.81
2011                          6.29         3.5       10.74        9.70
----------------------------------------------------------------------
Source:  SSA based on 1997 Trustees' intermediate assumptions about
yields on special Treasury securities and inflation. 


--------------------
\1 The 7 percent real stock yield was 4.7 percentage points greater
than the 2.3 percent yield expected under the Trustees' 1995
assumptions, which were used in the Advisory Council's analysis. 

\2 In its analysis, the Advisory Council assumed that the annual
administrative costs would apply to the trust fund's entire asset
balance and not just to its stock holdings. 

\3 The spread over the real yields on Treasury securities would be
4.295 percent under the 7 percent real stock return assumption and
3.295 percent under the 6 percent assumption.  The spread between the
nominal yields is slightly higher. 


   STOCK INVESTMENT SCENARIOS
--------------------------------------------------------- Appendix I:3

The potential gain from stock investing also would depend on how much
the Social Security trust fund invests in the stock market.  We
developed an aggressive scenario investing all future amounts beyond
a 100 percent contingency reserve level and a more conservative
scenario investing only Social Security's cash surplus.  Under the
aggressive scenario, the trust fund would hold its balance of special
Treasury securities constant as of the beginning of 1998.  From 1998
until 2008, all of Social Security's cash surplus and the interest on
its special Treasury securities would be invested in the stock
market.  Beginning in 2008, the trust fund would need to begin
investing more in Treasury securities to maintain a 100 percent
reserve level.  Under the cash surplus scenario, the trust fund would
invest in the stock market until 2012 and then it would begin drawing
on its stock earnings and sales to finance Social Security's cash
deficit.  In both scenarios, stock earnings are reinvested in the
market unless the trust fund needs cash to pay benefits or to invest
in Treasury securities to maintain its contingency reserve. 

Table I.2 shows, under current law and the two stock investment
scenarios, the years when (1) the trust fund would be exhausted, (2)
its asset level would fall below 100 percent of expected annual
expenditures, and (3) its asset level would fall below 150 percent. 
These simulation results illustrate some outcomes associated with two
alternative investment policies.  These results should not be
interpreted as forecasts and do not represent the full range of
possible outcomes for the Social Security trust fund. 



                               Table I.2
                
                  Key Dates Under Current Law and Two
                       Stock Investment Scenarios

                                             Assets less   Assets less
                                                than 100      than 150
                                Trust fund    percent of    percent of
                                 exhausted  annual outgo  annual outgo
----------------------------  ------------  ------------  ------------
Current law                           2029          2025          2022

Aggressive scenario
----------------------------------------------------------------------
7 percent real yield                  2040          2036          2034
6 percent real yield                  2035          2032          2029

Cash surplus scenario
----------------------------------------------------------------------
7 percent real yield                  2032          2028          2026
6 percent real yield                  2031          2027          2025
----------------------------------------------------------------------
Source:  SSA, Office of the Chief Actuary. 

In our opinion, the aggressive scenario represents the outer limit of
how much the Social Security trust fund might invest in the stock
market.  Hypothetically, the trust fund could redeem its existing
balance of Treasury securities to invest more in the stock market
now.  Or the trust fund could hold an amount of Treasury securities
less than 100 percent of expected annual expenditures.  We do not
believe either of these scenarios would be sound from an investment
perspective.  Under the aggressive scenario, the trust fund's stock
holdings would peak at more than 70 percent of its portfolio.  A
scenario investing even more in stocks would increase the trust
fund's exposure to the risk of loss in the event of a general stock
market downturn.  Moreover, the trust fund would not hold enough
Treasury securities to maintain an adequate reserve to cover
unforeseen liquidity needs. 

There are also other reasons why we question whether it would be
feasible to invest more than under the aggressive scenario.  For one,
under the Trustees' 1997 intermediate estimates, the trust fund's
special Treasury securities at the beginning of 1998 were expected to
be approximately $647 billion, which is equivalent to about 5 percent
of U.S.  stock holdings, valued at $12.8 trillion as of the third
quarter of 1997.  Attempting to rapidly shift the trust fund's
existing balance into the stock market would magnify any upward
pressure on stock prices resulting in the short term from government
stock investing.  Finally, the aggressive scenario presents budgetary
challenges:  Social Security's cash surplus would not be available to
finance other government activities, and the Treasury would have to
raise additional cash to finance interest payments to the trust fund. 
The trust fund reinvesting its existing balance in the stock market
would create another budgetary challenge--the Treasury would have to
repay money previously borrowed from the trust fund. 

The cash surplus scenario is somewhat conservative.  The trust fund's
stock holdings would peak at about 35 percent of its portfolio, which
is conservative in comparison to the 60 percent held by state and
local pension funds as a whole.  Also, under the cash surplus
scenario we specified, the trust fund would sell its stocks first to
finance Social Security's cash deficit even before tapping the
interest on its special Treasury securities.  It is possible to
construct a more conservative investment scenario, for example, by
limiting stock holdings to a lower percentage of the trust fund's
portfolio.  Alternatively, it is possible to construct a less
conservative cash surplus scenario.  The trust fund could draw on its
special Treasury securities in excess of the contingency reserve
level instead of selling its higher-yielding stocks first.  Under
this less conservative scenario, investing Social Security's cash
surplus in the stock market, still assuming a 7 percent real yield,
would delay the trust fund's exhaustion until 2034.  This result is 5
years longer than expected under current law and 2 years longer than
is possible under the more conservative scenario. 


GOVERNMENT STOCK INVESTING AND
FEDERAL BUDGETARY ACCOUNTING
========================================================== Appendix II

While the costs of government stock investing would show up
immediately under the current budgetary accounting system, the
potential benefits would not appear for several years or more.  This
discontinuity between costs and benefits would make it difficult for
policymakers and the public to evaluate a stock investment policy. 
Changing the budgetary accounting rules for stock purchases to
exclude them from government outlays could alleviate the confusing
budget signals that would arise under current scoring methods. 
However, changing the scoring treatment would also eliminate the
clearer view of the government's finances excluding the Social
Security surplus that would otherwise result under stock investing. 
Alternatively, supplemental information on the stock portfolio could
be used to track changes in its value. 


   STOCK PURCHASES WOULD INCREASE
   BUDGET OUTLAYS
-------------------------------------------------------- Appendix II:1

The federal budget is largely cash-based.  Receipts are recorded when
received and outlays are recorded when paid, without regard to the
period in which taxes and fees were assessed or the costs incurred. 
According to budget experts that we interviewed, the purchase of
stocks, like any other purchase, would be treated as an expenditure
or cash outlay.  In this way, stock purchases increase federal
spending and, therefore, raise the reported unified budget deficit or
lower the reported surplus. 

While budget analysts consider a stock purchase to be an expenditure,
the issue is less clear when viewed from the perspective of Social
Security.  The trust fund is not spending more money; it is simply
investing in a different asset.  Why would choosing a different asset
increase federal spending?  The answer lies in the distinct nature of
each type of asset transaction.  Purchases are generally treated as
budget outlays.  However, the trust fund's purchase of a Treasury
security is a debt transaction.  Such transactions do not constitute
outlays or receipts; if they did, the budget would approximately be
balanced by definition.  In addition to debt transactions, another
exception to the general treatment of purchases as outlays applies to
assets that are equivalent to cash.\1 Stock purchases do not fall
under either of these exceptions:  they are not debt transactions and
budget analysts do not consider them to be equivalent to cash because
they are not as liquid and their value fluctuates. 

Any income from dividends on stocks or sales of stock would be
counted as offsetting receipts, i.e., it would reduce total spending. 
However, in the case of dividends, any reinvestment would be treated
as additional spending.  So, on balance, a policy of reinvesting
dividends would have no immediate impact on the budget deficit or
surplus as the effects on spending would offset one another. 


--------------------
\1 For example, gold is considered a cash-equivalent asset. 


   POTENTIAL GAIN FROM STOCK
   INVESTING WOULD NOT BE VISIBLE
   IN THE SHORT TERM
-------------------------------------------------------- Appendix II:2

In addition to its cash-based nature, federal budgeting has a
relatively short-term focus--generally 1 to 5 years.  Together, these
factors would tend to obscure the potential benefit of stock
investing.  Most of the stock purchases would occur over the next 10
to 15 years, while any revenue gains would likely be concentrated in
later years.  Thus, any long-term benefits would not be immediately
visible.  The budget would not reflect any change in value of the
stock portfolio until the trust fund sells its stocks.  In the
short-term, then, stock investing might not look particularly
attractive--cash would go out to purchase stocks, but little or no
new cash would come in to the Treasury. 

The long-term signals sent by stock investing could also be confusing
to policymakers and the public.  When the trust fund sells stocks in
the future to help pay benefits, the cash proceeds would appear in
the budget as revenue, but it would not be clear how much the stocks
had gained or lost in value. 

Budgetary accounting rules could be changed to avoid these confusing
signals, but such a change could carry costs as well as benefits. 
For example, if government stock investing were scored under
different rules than other types of federal spending, stock purchases
could be excluded from government outlays.  As noted previously, the
purchase of stock does not have the same effect on the economy as
other types of federal spending.  Different treatment in the budget
would underscore this distinction.  However, creating different rules
for stock purchases would also undermine the possible incentive for
addressing the deficit that exists when Social Security's surplus is
excluded from the government's finances.  If stock purchases were not
treated as outlays, Social Security's surpluses would continue to
obscure the size of this deficit.  Counting stock purchases as
outlays, as current scoring requires, would make this deficit more
visible. 

Instead of changing budgetary accounting procedures to deal with
stock purchases, it might be preferable to provide additional
information on a government stock portfolio as supplementary budget
data.  The Social Security Trustees already issue annual reports that
include information on trust fund assets.  These reports could be
expanded to include an annual statement of the government's stock
portfolio that would contain information such as the amounts
invested, the nature of the investments, and the change in market
value of the holdings.  In addition, similar information could be
included in the budget documents themselves. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix III


   ACCOUNTING AND INFORMATION
   MANAGEMENT DIVISION,
   WASHINGTON, D.C. 
------------------------------------------------------- Appendix III:1

Jose Oyola, Assistant Director
MaryLynn Sergent, Senior Evaluator


   BOSTON FIELD OFFICE
------------------------------------------------------- Appendix III:2

Andrew D.  Eschtruth, Evaluator-in-Charge


   HEALTH, EDUCATION, AND HUMAN
   SERVICES DIVISION, WASHINGTON,
   D.C. 
------------------------------------------------------- Appendix III:3

Barbara D.  Bovbjerg, Associate Director
Francis P.  Mulvey, Assistant Director
Michael Packard, Economist
Ken Bombara, Economist
Ken Stockbridge, Senior Evaluator

RELATED GAO PRODUCTS

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

Social Security:  Restoring Long-Term Solvency Will Require Difficult
Choices (GAO/T-HEHS-98-95, February 10, 1998). 

Social Security Reform:  Demographic Trends Underlie Long-Term
Financing Shortage (GAO/T-HEHS-98-43, November 20, 1997). 

Budget Issues:  Analysis of Long-Term Fiscal Outlook
(GAO/AIMD/OCE-98-19, October 22, 1997). 

Retirement Income:  Implications of Demographic Trends for Social
Security and Pension Reform (GAO/HEHS-97-81, July 11, 1997). 

Compendium of Budget Accounts:  Fiscal Year 1998 (GAO/AIMD-97-65,
April 1997). 

Federal Debt:  Answers to Frequently Asked Questions (GAO/AIMD-97-12,
November 27, 1996). 

Public Pensions:  State and Local Government Contributions to
Underfunded Plans (GAO/HEHS-96-56, March 14, 1996). 

Budget Account Structure:  A Descriptive Overview (GAO/AIMD-95-179,
September 18, 1995). 

*** End of document. ***