[Economic Report of the President (2006)]
[Administration of George W. Bush]
[Online through the Government Printing Office, www.gpo.gov]

 
CHAPTER 3


Over the past few decades, concerns have mounted that Americans have
been preparing inadequately for retirement. Recent newspaper
headlines suggest that Americans have stopped saving and are at risk
of sharp reductions in both their private and public pension
benefits. To be sure, these concerns have some basis: The aggregate
personal saving rate published in the National Income and Product
Accounts (NIPA) turned negative in 2005; high-profile bankruptcies
in airlines and other industries have led to substantial reductions
in retiree pension benefits; the collapse of technology stocks in
the early 2000s left many defined-benefit pension plans underfunded;
and promised Social Security benefits vastly exceed forecasted
revenues. Understanding how these events relate to retirement
security is important if public policy is to respond productively.
This chapter builds such an understanding. The main points are:
 Most working-age Americans are on track to have more
retirement wealth than most current retirees. However,
it is inherently difficult to assess whether these
preparations are adequate for most households, given
that incomes have also grown over time and people may
have markedly different plans for their retirement length
and standard of living.
 The decline in an often-cited aggregate personal saving
rate may not be cause for alarm. Much of this decline can
be attributed to spending triggered by wealth increases
from capital gains on housing and financial assets.
 There are, however, a number of risks to the retirement
preparations of Americans: People today are living longer
and could face higher health-care costs in retirement than
members of previous generations. In addition, Social
Security and many defined-benefit pension plans are at
risk.
 Both defined-benefit pensions and Social Security suffer
from fundamental financial problems, which expose not just
retirees but all U.S. taxpayers to risk of substantial
losses. The Administration is focused on addressing these
problems and protecting the Nation's retirement security.

What Does ``Retirement Preparedness'' Mean?

Retirement preparedness is defined here as the accumulation of
wealth necessary to maintain a desired standard of living in
retirement. Economists tend to agree that individuals want to smooth
consumption in retirement (i.e., limit the extent to which retirement
will decrease their consumption). However, individuals may have
disparate views about how much they want to smooth consumption, when
they plan to retire, and how much they intend to work in retirement.
Thus, two individuals, even with the same preretirement standard of
living, may have markedly different views about how much wealth
accumulation is adequate.
For the purposes of this discussion, we divide the wealth that
individuals can draw on in retirement into three categories: personal
net worth, including defined-contribution pension plans;
employer-sponsored defined-benefit pensions; and Social Security.
(Retirement wealth also includes other expected benefits, such as
retiree health care from employers and Federal programs, but such
benefits fall outside the scope of this chapter.) Personal net
worth is the sum of the value of financial assets (e.g., stocks
and bonds held in and out of retirement accounts such as 401(k)
plans, and savings accounts) and durable goods (e.g., houses and
cars) less the value of liabilities (e.g., credit card debt,
mortgages, and car loans). Net worth grows in part from personal
saving--the excess of after-tax income over consumption--and in
part from inheritances and capital gains on assets already owned.
Some portion of current workers' net worth, however, may be
drawn down before retirement. For instance, households may
liquidate financial assets or take out home-equity loans to make
tuition payments, pay health-care expenses, or offset negative
income shocks.
The other two sources of retirement wealth, employer-sponsored
defined-benefit pensions and Social Security, are sometimes referred
to as retirement income, since payments from both sources are
periodic. Employer-sponsored defined-benefit pensions generally
increase with years of employment and salary levels, while Social
Security payouts tend to increase with retirement age and average
lifetime earnings.
The next section of this chapter considers how prepared households
are for retirement. Because the definition of retirement adequacy
is somewhat subjective, we focus primarily on cross-generational
comparisons of retirement-wealth accumulation. Cross-generational
comparisons do not speak directly to the adequacy of retirement
preparations, but do shed light on the related question of whether
retirement preparations have deteriorated.

Estimates of Retirement Preparedness
This section begins with a brief description of the results from
studies that directly address the difficult question of whether
retirement preparations are adequate. The section then discusses
cross-generational comparisons, beginning with comparisons of net
worth and ratios of net worth to income, and then turning to
comparisons of retirement income from defined-benefit pensions and
Social Security. The section concludes with a discussion of the key
limitations of cross-generational approaches.
Studies that directly address the question of retirement adequacy
typically define adequate wealth accumulation as essentially that
which is expected to smooth consumption according to a particular
model of individual preferences. Given that these studies make different
key modeling assumptions, and in some cases include different
components of expected retirement wealth, they have generated a wide
range of results. Nevertheless, some recent studies find that most
baby-boom households have been preparing adequately. In any case,
conclusions about retirement adequacy based on these studies should be
regarded as suggestive only, given the inherent uncertainty surrounding
predictions of how much wealth is enough.
Comparing retirement wealth across generations, unlike evaluating the
adequacy of any one generation's preparations, can be done without reliance
on subjective assumptions. One such cross-generational study of
retirement wealth contrasts the net worth (defined as above) of
households in the baby-boom generation (individuals born between 1946
and 1964) and generation X (headed by individuals born between 1965
and 1976) with that of households in the pre-baby boom generation (headed
by individuals born between 1925 and 1945). The study considers the net
worth of the heads of these households when they were between 25 and 34
years old. Controlling for age is essential given that individuals tend
to save at different rates over their lifetimes.
The study finds that baby-boom and generation-X households tend to have
more net worth than pre-baby-boom households had when they were
roughly the same age. As shown in Table 3-1, the median net worth
of pre-baby-boom households at ages 25-34 was $6,072 in 1998
dollars. In contrast, the median net worth of baby-boom and
generation-X households was, respectively, $19,504 and $15,500 in
1998 dollars. The somewhat lower median net worth of generation-X
households mainly reflects their higher debt burdens. The table also
reveals that baby-boom and generation-X households with heads of all
types--low or high education, married or single--were better off
than pre-baby-boom households.
We might also want to compare household net worth to income for
each generation to see whether saving rates have kept pace with
increases in income.  Intuitively, households with greater
wealth-to-income ratios will be better able to maintain preretirement
living standards when they retire. As shown in Table 3-2, the same
study also finds that median net worth-to-income ratios are higher
for the baby-boom and generation-X households than for the pre-baby-boom
households, and these gains were experienced by a wide range of
demographic groups.
Finally, we can compare the median expected retirement income of
baby-boom households with that of generation-X households. The study
finds that median expected retirement income (including predicted
defined-benefit pension and Social Security payouts in inflation-adjusted
dollars but not personal net worth) for generation-X households is
greater than that for


baby-boom households. A second, less sanguine,
result is that if the Social Security system's expected funding
shortfalls are resolved by gradually reducing retirement benefits
(notably, not the Administration's proposed solution) and thus
lowering benefits for generation X more than for the baby boomers,
then the median expected retirement incomes of generation-X and
baby-boom households are about the same. This implies that, in terms
of retirement income relative to preretirement income, generation-X
households have not kept pace with the baby boomers.
The results shown above have a few important limitations. First,
cross-generational comparisons fail to adjust for the possibility
that current generations may live longer and could face higher
health-care costs in retirement than previous generations. As a result,
current workers may need more retirement wealth than previous
generations. On the other hand, longer life expectancies may encourage
current generations to work longer than previous generations, which,
all else equal, would lower retirement-wealth needs.
Another limitation of these cross-generational comparisons is that
they consider only a relatively early period in each generation's
lifecycle (although they allow the inclusion of more recent generations).
However, studies that compare somewhat older households from the
baby-boom generation to recent retirees find similar conclusions.
Nevertheless, retirement preparations of today's Americans may
veer off track as they age if they stop saving or if financial-asset
returns, house-price gains, or defined-benefit pension and Social
Security payouts turn out to be less than expected. The next section
of this chapter addresses some of the key risks to retirement preparations.

The Risks to Retirement Preparedness
Three risks to retirement wealth are discussed in this section: first,
the risk to household net worth created by the negative level of
the personal saving rate, as measured in the National Income and
Product Accounts (NIPA); second, the risk to defined-benefit pension
plans created by underfunding, in part due to investments in risky
assets; third, the risk to Social Security from the aging of the
population and other structural problems.

Are Low Saving Rates Putting Household Net Worth
at Risk?
The NIPA personal saving rate is the difference between the household
sector's after-tax personal income (disposable income) and personal
consumption, expressed as a percentage of disposable income. As a
technical matter, the household sector includes nonprofit
institutions. The NIPA personal saving rate was constructed as a
measure of the household sector's contribution to national saving--funds
set aside from the economy's current production to finance investment
(see Chapter 1, entitled The Year in Review and the Years Ahead,
and Chapter 6, entitled The U.S. Capital Account Surplus, for more
discussion of the national saving rate). However, the NIPA personal
saving rate is widely cited in newspapers as a gauge of retirement
preparedness. The discussion here details the NIPA saving rate's
limitations as a measure of the extent to which households are adding
to their retirement wealth. The goal of the discussion is to assess
whether the decline in the NIPA personal saving rate reflects a
widespread deterioration in household retirement preparations.
Chart 3-1 illustrates the decline in the NIPA personal saving rate. The
saving rate is volatile from quarter to quarter but has been trending
down at a relatively constant rate of about 0.5 percent per year
since the early 1980s. In the fourth quarter of 2005 (the most recent
quarter for which data are available), the NIPA personal saving
rate was -0.4 percent, not far above the post-World War II low
observed in the third quarter.


However, the relationship between the personal saving rate and
households' wealth accumulation is not always close. Household net
worth is what matters for retirement, but the NIPA personal saving
rate is not equal to the change in household net worth. First, the
NIPA personal saving rate excludes the acquisition of consumer durables,
a component of household net worth. Second, while business saving (such a
s businesses' retained profits) is ultimately owned by households, it
is also excluded from NIPA personal saving. Third, and arguably most
important, the NIPA personal saving rate excludes capital gains on
financial and other assets (e.g., the increase in the value of a house);
however, taxes on capital gains, which reduce the saving rate, are
included in the computation of personal saving. The exclusion of
capital gains is particularly problematic because capital gains may
encourage households to consume more, which in turn drives down the
measured saving rate. In other words, capital gains may be reflected
in the data as reductions in saving, even though these gains add to
household wealth on net--though some might argue that these gains can
be illusory.

Do Wealth Gains Explain the Decline in the NIPA Personal
Saving Rate?
The consumption-wealth effect (i.e., the tendency to consume more
as wealth increases) has been the subject of numerous empirical
investigations. Studies find that an additional dollar of wealth tends
to lead to a permanent rise in the level of household consumption
of about 2 to 5 cents. The link between aggregate wealth and spending
has proved to be one of the more enduring relationships in macroeconomics.
Estimates of the consumption-wealth effect suggest that it can explain
a sizable portion of the decline in personal saving since the
mid-1990s. As shown in Chart 3-2, the ratio of household net worth to
disposable income has risen from about 440 percent in the early
1980s to about 550 percent in the third quarter of 2005. This measure
of household net worth, obtained from the Federal Reserve's Flow of
Funds Accounts, is the difference between household assets--including
defined-benefit pension wealth--and household liabilities. The
ratio moved up and down with the rise and collapse of the stock
market in the late 1990s and early 2000s and then rebounded more
recently along with rising house prices and stock market gains.
An estimate of the impact of these wealth gains on the NIPA personal
saving rate is shown below in Chart 3-3. Under the assumption that
an additional dollar of wealth leads to a $0.035 permanent rise
in the level of consumption (the middle of the range cited above),
the chart shows that the personal saving rate would have declined
about half as much since 1980 if household wealth had grown at the
same pace as disposable income (keeping the ratio constant) over
that period.


Are Saving Rate Declines Widespread?
Yet another limitation of the NIPA personal saving rate as a
measure of households' wealth accumulation is its aggregate nature;
as such, it masks possible differences in behavior by households
at different income levels. Understanding the saving dynamics
in different parts of the income distribution requires
household-level data on saving.
However, household wealth at the individual level is difficult to
track over time. One study thus employed an innovative approach
to circumvent various data problems and found that the saving rate,
using NIPA definitions, for households in the upper two-fifths of
the income distribution declined over the 1990s, while the saving
rate for households in the middle fifth remained relatively steady,
and the saving rate for households in the bottom two-fifths
actually increased. Given that high-income households almost certainly
experienced the majority of capital gains in the 1990s, these results
suggest that the net worth component of retirement wealth may not be
at risk. Relatively high-income households may have accumulated net
worth from capital gains, while other households may have accumulated
net worth by saving.
Overall, the above discussion of household saving suggests that the net
worth component of retirement preparedness may not be in jeopardy. The
NIPA personal saving rate is a potentially misleading measure of
households' wealth accumulation. Moreover, much of the recent decline
in the NIPA personal saving rate may reflect consumption increases that
were triggered by capital gains on stocks and real estate. Finally,
some evidence suggests that the decline in household saving rates has
not been widespread but may have been concentrated among higher-income
households.

Policy Reforms
While the net worth component of retirement wealth does not appear
to be in jeopardy, policy reforms can still productively reduce
impediments to saving. Under current law, interest income is taxed,
creating a disincentive for households to set aside funds for retirement.
This disincentive is mitigated to some extent by policies that afford
favorable tax treatment to various types of retirement accounts (e.g.,
IRA and 401(k)). However, restrictions on these accounts limit their
value as retirement-saving vehicles. To make these accounts more effective,
Congress passed legislation that increases contribution limits and makes
retirement assets more portable. In addition, the Administration has
proposed simplifying the retirement account system in two important
ways: (1) creating a single Retirement Savings Account (RSA) to replace
the three types of Investment Retirement Accounts (IRAs) currently in
place; and (2) creating a Lifetime Savings Account (LSA) that could be
used for a variety of purposes, including retirement saving (see Chapter 5,
entitled The U.S. Tax System in International Perspective, for additional
discussion of tax recommendations in the President's Budget). Another
impediment to saving may be limited financial knowledge. The Department
of the Treasury is actively engaged in campaigns to improve financial
literacy. In addition, the President has instructed the Federal Deposit
Insurance Corporation (FDIC), the Small Business Administration (SBA),
and the Treasury Department to work with consumer groups to ensure
that financial literacy is widespread.

Defined-Benefit Pensions
Historically, defined-benefit pension plans have been an important
part of retirement preparedness. These employer-sponsored plans
compensate retirees through a specified monthly benefit, which tends to
vary with salary and years of service. In addition, most plans
sponsored by private employers are guaranteed in part by the Pension
Benefit Guaranty Corporation, and those sponsored by public employers
are ultimately backed by the ability of states to levy taxes. As such,
''DB'' plans may appear more stable than increasingly prevalent
''defined-contribution'' plans (such as 401(k) plans), which explicitly
depend on employee contributions, tie benefits more directly to market
performance, and may expose retirees to longevity risk (the risk of outliving
retirement resources).
Defined-benefit plans can, nevertheless, carry considerable risk. This
risk comes from employers (1) contributing less to plans than what is
promised to employees (funding risk), (2) investing contributions in
a hazardous manner (portfolio risk), and (3) encountering financial
distress (bankruptcy risk) in the case of private employers. When
these risks are realized, beneficiaries and taxpayers can be exposed
to substantial and oftentimes unanticipated losses.
An early example of these problems comes from the 1960s landmark case
of Studebaker Corporation. When this former carmaker defaulted on its
defined-benefit plan, it left about 11,000 participants without most
or any of their pensions. These losses eventually led Congress to set
minimum standards for private pension plans via the Employee Retirement
Income Security Act (ERISA) in 1974.
ERISA gave rise to the Pension Benefit Guaranty Corporation (PBGC), which
now partially insures the pensions of over 34 million workers and retirees.
The PBGC largely funds itself with premiums from private-sector sponsors
of defined-benefit plans (i.e., employers). When an employer becomes
financially distressed, the PBGC may take control of the plan's management
and use the plan's assets and its own funds to pay retirees a capped portion
of their promised benefits. Employees in contemporary cases like the
bankruptcy of United Airlines filed in 2002 are thus less exposed to
defined-benefit risks than were employees in cases like Studebaker.
Despite this insulation, employees with defined-benefit pension plans
sponsored by private employers remain exposed to considerable risks. As of
2005, for example, the limit on PBGC insurance increased with retirement age,
and topped out at about $46,000 per year. Employees whose plans default can
thus incur considerable losses when their promised benefits exceed these
limits. United's workers, for example, expect to receive about 80 percent
of their earned benefits, and thus stand to lose more than $3 billion of
total promised benefits. In addition, as the following sections show, the
combination of inadequate protections and a series of pension defaults has
left the PBGC with insufficient funds for paying even these limited claims.
Consequently, if losses overwhelm the pension insurance system, Congress
may step in and pass the bill to taxpayers.
For defined-benefit plans sponsored by public employers, the taxpayer exposure
is even more direct. Recall that the PBGC only insures plans sponsored by
private employers. In the event that a publicly sponsored plan's assets are
insufficient to pay benefits, absent renegotiation of benefits, such plans
could only be made whole with the support of state-level tax revenues.

Employee Exposure to Defined-Benefit Risks
Recently, market fluctuations and the rules that govern how employers
participate in the defined-benefit system appear to have turned risks into
reality. Decreasing interest rates and stock market valuations, coupled with
the exposure of pension plan assets to market fluctuations, coincided with a
marked increase in the underfunding of defined-benefit plans. Underfunding,
in turn, increased expected defaults on pension obligations, putting both
workers and the pension insurance program into jeopardy.
In the case of privately sponsored pensions, the value of assets set aside to
fund retirement obligations began to decrease in 2000 while the value of
promised benefits began to increase. The total underfunding of private pension
plans grew from less than $50 billion at the end of 2000 to over $400 billion
today. At the same time, as Chart 3-4 illustrates, PBGC's capacity to insulate
workers from employer defaults turned from a $10 billion surplus in 2000 into a deficit that now totals more than $20 billion.
This deterioration can plausibly be attributed to the exposure of pension plan
portfolios to coincident decreases in both interest rates and stock market
valuations. A decrease in interest rates can contribute to this problem by
increasing the measured present value of a pension plan's promised benefits. A
decrease in stock market valuations can further contribute by weakening the
ability of plan investments to pay benefits.
To see this relationship, suppose that an individual wants to buy a new appliance
next year for $500, and consider how much must be saved today to


fund this purchase. The answer depends on how much interest these savings will
earn: As this interest increases, the savings that are necessary to fund the
future purchase decrease. Extreme cases are illustrative: One would have to
save $500 today if the interest rate is 0 percent, but only $250 if it is
100 percent. This example reflects a more general relationship: When interest
rates decrease, the present value of future obligations increases.
For pensions, this relationship implies that employers must set aside more
funds to meet pension obligations when interest rates decrease. The decrease
in interest rates that started late in 2000 thus threatened the funding
status of defined-benefit pension plans.
A simultaneous decrease in stock market valuations from the peaks of the
late 1990s appears to have furthered this threat. At the same time that
interest-rate changes were increasing the value of employers' obligations,
a decrease in stock market valuations was diminishing the value of assets
that employers had set aside to fund those obligations. Together, these
changes coincided with the marked weakening in the funding status of both
defined-benefit plans and the PBGC.
While market fluctuations appear to have been an important contributor
to these woes, they could be made less so. To see why, recall from above
that the PBGC manages the pension plans it receives from financially
distressed employers. In doing so, it reduces exposure to interest-rate
fluctuations by matching investment payoffs with the timing of employee
benefits. The value of plan assets and liabilities will tend to move more
closely together under this strategy of duration matching than they would
under the strategies that employers appear to have used.

Taxpayer Exposure to PBGC's Deficit
The recent spike in underfunding has also exposed taxpayers to the prospect
of making up for the PBGC's deficit (recall that this exposure is more
immediate for publicly sponsored plans). While the PBGC's liabilities
are not explicitly backed by the Federal government, a future Congress
might decide that a taxpayer bailout is preferable to a PBGC default.
Indeed, taxpayers' exposure to the PBGC's deficit is especially concerning
since the manner in which it evolved mimics how the 1980s savings and
loan (S&L) crisis developed.
Like the insurance that PBGC offers, the insurance offered to depositors
at financial institutions can provide important benefits. But if they
are not prudently managed, these insurance programs can fall prey to
moral hazard (explained in Chapter 9, The U.S. Financial Sector) and
thus expose taxpayers to an undue liability. In the 1980s, for example,
loose regulatory oversight let savings and loans overly expose themselves
to market fluctuations (such as changes in real-estate values and
interest-rates) and ultimately left insufficient funds for paying off
depositors. Depositors did not fully bear the burden of this underfundng,
however. Instead, the Federal Savings and Loan Insurance Corporation
(FSLIC) insured depositors in much the same way that PBGC covers retirees.
In an analogous manner to the current pension situation, market fluctuations
and regulatory difficulties not only helped increase the rate at which
depositors drew on this insurance, they also compromised FSLIC's capacity
to pay insurance claims. Like the PBGC, FSLIC was structured to be
self-financing. Nevertheless, taxpayers ultimately paid about $150 billion
for the financial losses of failed institutions.
The PBGC faces a situation that is similar to what plagued FSLIC. Waiting
to implement productive reforms magnified taxpayers' burden in bailing out
the S&L industry. Postponing the issue of underfunded pension plans can
likewise make matters worse for pensioners and taxpayers. According to
testimony by the PBGC's executive director, the PBGC's present $23 billion
deficit could grow toward $80 billion over the next ten years. Without
prompt and effective action, taxpayers may thus find themselves bailing out
yet another ''self-financed'' public insurance program.


Policy Reforms
Prompt action, grounded in good economics and informed by lessons learned
from similar financial crises, can keep the current pension problem from
becoming even more burdensome. To help the private pension system move in
this direction, the administration has proposed to strengthen the requirements
for funding privately sponsored pension plans and improve the manner in which
plan sponsors disclose information. State-level policies that would address
the problems with plans sponsored by public employers are at an earlier
stage of development.
Current funding and disclosure rules can allow privately sponsored pension
plans to appear healthier than they actually are. Reforms such as restricting
the use of ''credit balances'' could help enhance funding adequacy and
transparency. Under present law, employers receive credit for contributions
that exceed minimum requirements and can later use those credits in lieu of
actual contributions. This treatment is problematic. For example, excess
contributions are characterized as earning interest even if the assets in
which those contributions were invested lose value. Moreover, credit balances
can delay plan sponsors from addressing funding problems and thus let even
grossly underfunded employers forgo actual contributions.
Limiting private employers' ability to use an average interest rate to value
plan liabilities could also strengthen funding and improve transparency.
Recall that, as interest rates decrease, the present value of an employer's
pension obligations increases. Current law lets employers use a moving average
of these rates spread out over several years, however, and thus mutes the
near-term effect of an interest-rate decrease on an employer's contribution
requirements.
To see this effect, suppose that employers can use a two-year average, and
that interest rates decrease from 6 percent to 5 percent. Using an average
rate, employers could discount their future obligations at 5.5 percent. But
if employers had to use the current rate of 5 percent, they would have to
increase contributions by more, and do so more quickly. Averaging the discount
rate can thus cloud the picture of a plan's status.
The Administration has similarly proposed limits on the ability of private
employers to smooth reported fluctuations in the value of their plan-assets.
Coupled with the related proposal for plans to accurately address the timing
of benefit payments, this reform could reduce the portfolio risks that are
characterized above as the proximate cause of the system's weakened funding
status.
Finally, the administration has proposed to increase funding targets, measure
the performance of plans in a uniform manner, and update assumptions like
those of mortality. These reforms, like the others discussed above, would
enhance the integrity of the defined-benefit system, and should be uniformly
applied across plan sponsors. Doing otherwise would give some economic
sectors, or firms within a sector, an artificial advantage. Economic
performance could deteriorate as scarce resources flow not to their most
productive uses, but to their most politically-favored uses. In addition,
exempting certain sectors or firms could exacerbate the underfunding problem
by breathing artificial life into risky plans and thus further exposing
workers, retirees, and taxpayers to economic risk.


Social Security
Along with personal savings and employer-provided pension plans, Social
Security has long stood as a pillar of retirement security. A response of
Franklin D. Roosevelt's administration to the Great Depression, the Social
Security Act was signed into law on August 14, 1935, and first issued
monthly retirement checks in January 1940. At that time, about 200,000
retirees received aggregate benefits valued at about $35 million. Since
then, both the number of beneficiaries and the level of benefits has
steadily grown. In 2004, more than 47 million beneficiaries received a
total of about $493 billion through the Old Age, Survivor, and Disability
Insurance programs (OASDI).
These benefits are funded by taxes on wage income. In an accounting sense,
employers and employees equally share this funding by contributing
6.2 percent of taxable payroll each. Since employers focus on the total
cost of labor, however, workers bear most of this combined 12.4 percent
tax. For each worker, this tax applies to payroll beneath a ceiling that
annually adjusts with the average wage index. That ceiling, which stood
at $90,000 in 2005, increased to $94,200 for 2006.


Taxpayer Exposure to an Increasingly Large Social Security Burden
The overall cost of Social Security is substantial. The Office of Management
and Budget (OMB) estimates that Social Security transfers amounted to
4.2 percent of GDP in 2005. During the coming decades, Social Security's
share of GDP is expected to increase, reaching 6 percent in 2035.
In the short term, this increase will largely come from the retirement
of baby boomers, which begins in 2008. It will persist in the long run,
however, due to a combination of relatively low fertility rates and
relatively high life expectancies. These factors will push the ratio
of workers to retirees down from its current level of 3.3 to 1 to around
2 to 1 by the time that most baby boomers retire.
Since the benefits of those currently retired mostly come from taxes on
those currently working, these developments will create considerable
pressure to increase payroll taxes. Indeed, the Social Security
Administration's actuaries estimate that, starting in 2017, the system's
annual cost will exceed its total tax income (which includes taxes on
payroll and Social Security benefits themselves).
From an accounting perspective, Social Security can still fully fund
benefits at this point because the system has run surpluses since 1984,
holding special Treasury bonds as IOUs. Although they are assets to the
Trust Fund, however, these IOUs are equally debt to the Federal
government, and thus an obligation that faces taxpayers.
The actuaries estimate that without legislative action, the Trust Fund's
IOUs will run out by 2041, leaving a system that can fulfill only
74 percent of currently scheduled benefits. Even more, promised
Social Security benefits from 2005 to 2080 are expected to exceed
the sum of revenues and Trust Fund IOUs by $4 trillion in present
value. Given these mounting costs, taxpayers and workers would be
better off dealing with this problem now rather than later.
Social Security reform has been on the national radar for decades
(see Box 3-1). Notably, former President Clinton convened an Advisory
Council which, in 1996, released several recommendations. Two of the
three plans supported by the Advisory Council involved some kind of
voluntary personal retirement accounts (through publicly held individual
accounts in one case and privately administered personal accounts in
another), and the other plan also envisioned moving to a system of
advance funding, albeit through government-directed investment in
equities. Importantly, the longer it takes to initiate reforms, the
greater any changes must be, because they will be shared by fewer
generations.

Policy Reform: Progressive Indexing
Projections suggest that, under current law, the Social Security
system will soon be unable to pay for itself. Many of the proposals
to address this problem fall short of a productive and durable
reform. Removing the cap on wages that are subject to the payroll
tax, for example, would not only increase contributions to the
system but also increase the system's promised benefits in the long
term. Progressively reducing future benefit growth, on the other hand,
may strike an attractive balance by closing roughly two-thirds of the
system's long-range annual cash shortfalls while maintaining the system's
capacity to act as a social safety net.
Initial benefits for new retirees are currently indexed to wage inflation
rather than price inflation.  Since wages typically increase at a
faster rate than prices (reflecting gains in productivity), wage
indexation results in increasingly large benefits in real dollar terms.
Progressive indexing would decrease the rate of benefit growth for
individuals whose lifetime earnings are the highest (less than the
highest 1 percent of all wage earners) by linking their benefit growth
to price increases. At the same time, it would maintain the current
law's more generous benefit-growth rate for individuals whose lifetime
earnings are relatively low. Benefits of retirees in the upper
70 percent of the
-------------------------------------------------------------------
Box 3-1: Earlier Attempts to Shore Up Social Security

Congress has responded to developing problems with Social Security
finances in the past. For example, both 1977 and 1983 saw the signing
of significant amendments to improve the system's deteriorating
financial condition.
Why were the system's finances deteriorating then, and why are they
continuing to do so today? There are several answers. First, the
1972 amendments to Social Security effectively indexed benefit growth
for those working at the time to both wage and price inflation,
essentially providing two cost-of-living adjustments. This
double-benefit indexation was amended in 1977 to establish the
current method of wage indexation. But while wage indexation addressed
the double-indexation issue, some experts warned that, coupled with
demographic changes, it would still require future taxpayers to
shoulder larger Social Security tax burdens than is required today.
Second, the economic projections following the amendments of 1972,
1977, and 1983 proved overly optimistic. From 1972 to 1976, for
example, real wages grew by nearly 11 percent less than expected,
resulting in lower than anticipated growth of the payroll income base
on which Social Security taxes were collected. Similarly, from 1977
to 1981, real wages decreased by about 6.9 percent rather than
increasing by 12.9 percent as projected. Assumptions made following
the 1983 reforms were not as far off as those of 1972 and 1977, but
are nonetheless responsible for some of the overstatement of Social
Security's financial strength. Consequently, although the year for
the exhaustion of the Trust Fund was forecast to be 2063 in 1983,
it has been pushed forward and now stands at 2041.
Third, and perhaps most importantly, the 1983 reforms did not attain
sustainable solvency. The 1983 reforms envisioned several decades
of Social Security surpluses, followed by several decades of large
and growing deficits. This meant that with the passage of time,
Social Security would again become financially imbalanced. Even as
early as the 1985 Social Security Trustees' report, it could be
seen that the system was again heading out of long-term balance.
This is one reason why a number of bipartisan commissions have since
recommended that future Social Security reforms place the program
on a sustainable, as opposed to merely a solvent, footing.
--------------------------------------------------------------------

distribution would depend on a combination of price
and wage increases. The system would be progressive because benefit
growth would slow the most for those with higher earnings. This method
of benefit growth would let future retirees enjoy benefits that are
higher than those paid today while eventually ensuring that no person
who works a full career would retire with a Social Security benefit below
the poverty level.
Progressive indexing would slow the benefit-growth rate for high-income
individuals in a manner that strongly pushes the system toward solvency.
In addition, by maintaining a relatively fast rate of benefit growth for
low-income individuals, progressive indexing would further protect
retirement incomes from falling below the poverty level.
Policy Reform: Personal Accounts
The traditional Social Security system largely funds retirement benefits
by transferring payroll taxes from current workers to beneficiaries.
In addition to being subject to the risk of insolvency (which, as explained
above, can be addressed in part through progressive indexing), this
type of pay-as-you-go system runs the risk of future workers voting
to cut back on their contributions. This risk may be considerable,
as additional changes needed to restore solvency would leave future
retirees with substantially smaller benefits than the current
system's promises.
This problem comes in large part from a system that relies on
future
generations to fulfill promises made today. By letting individuals
pre-fund their retirements, personal accounts allow current
generations to rely in part on their own savings, rather than
solely upon contributions that future generations may be unwilling
or unable to make.
Because this issue is separate from that of solvency, personal accounts
need not (and under the President's proposals, would not) adversely
affect the system's long-term finances. If traditional benefits are
offset by the amount that individuals could obtain by investing in
low-risk assets, such a reform
can be made approximately neutral with respect to the capacity to
fulfill remaining traditional benefits. Such offsets are said to be
roughly neutral on an actuarial basis because they leave (1) beneficiaries
who remain wholly invested in government bonds with the same expected
future benefit and (2) the Trust Fund with nearly the same expected
long-term balance.
While they leave the long-term balance mostly unchanged, allocations to
personal accounts do alter the timing of the system's future obligations.
Their basic effect is to take some of the long-term obligation and shift
it to an earlier time. Moving a portion of payroll taxes to personal
accounts will take money off of the government ledger today, some of which
is used to pay for current benefits and some of which has long been used
to finance other Federal spending. At the same time, because voluntary
personal retirement accounts will replace a portion of unfunded future
benefits, they also reduce future strains on the system.
Shifting the future imbalance forward in time could increase transparency
by making the system's impending shortfalls less of an abstraction. Financial
markets tend to applaud such solutions to fiscal challenges and might do
so again in this context by keeping interest rates at productive levels.
Pre-funding a portion of future benefits appears attractive in other
dimensions as well. Every dollar of benefits funded today through personal
accounts is a dollar of benefits that need not be paid by taxpayers in the
future. Because rising benefit obligations would under current law lead to
increased tax burdens over time, shifting forward the funding of some
benefits could create a more equitable treatment of different generations.
In addition, redirecting assets to personal accounts increases the likelihood
that real savings will be accumulated to meet tomorrow's retirement needs.
If these assets are owned and controlled by individuals, they will be
less available for the government to spend than if these assets are left
on the Federal ledger. Finally, personal accounts would provide an
opportunity for individuals to diversify their investment in Social
Security, which may add to their retirement security.

Conclusion

This chapter's first section shows that today's generations are on track to
have more retirement wealth than previous generations, though it is
unclear whether these wealth gains have kept pace with rising
preretirement incomes. Going forward, the relative security of
retirement wealth may be compromised by fundamental problems with
defined-benefit pensions and Social Security.
Both of these systems could be improved by more-effective funding
rules and safeguards that protect against the opportunistic handling
of retirement assets. Strengthening pension-contribution requirements,
and watching more carefully how those contributions are managed, would
go far to mitigate the growing risks to pensioners and taxpayers
alike. Progressively targeting the rate of future benefit growth and
expanding ownership over payroll contributions, likewise, would help
strengthen Social Security for the future. In both cases, waiting to
act allows the present problems to grow and increases the costs of
adopting effective reforms.