View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

For release on delivery
7:00 p.m., EST
March 7, 1994

The Best of Times - - The Worst of Times
Remarks by
Lawrence B. Lindsey
The Baltimore Chapter of
the Commercial Finance Association
Baltimore, MD
March 7, 1994

The Best of Times -- The Worst of Times

Thank you.
economic trends.

It's a pleasure to be here tonight to discuss
As economic and investment professionals,

you've all come to appreciate how easy it is to maKe a forecast
these days.

Whatever the question, the answer is three.

will real growth be?
About three.

About three.


What will inflation be?

What about short rates?

Three something.

Tonight, however, I am going to resist the temptation of
just uttering the number three and sitting down.

After all, if I

did so, you might think that I didn't earn my dinner.

And if

word got out that economics was as easy as one, two, three, we'd
all be in for a pay cut.
While the magic number for the economy is three, the grade
we should give our current economic performance is a ten.
believe that the reason for this is clear.


For the past twelve

years, my predecessors and colleagues at the Federal Reserve have
labored successfully to establish credibility as opponents of

The result has been a steady and continuous decline

in the long bond rate of more than sixty basis points per year -a remarkable cumulative reduction of 800 basis points in the
benchmark ten year Treasury bond, for example.
This long term sustained reduction in inflation expectations
and interest rates has provided our economy with one of the most
favorable environments for maximizing America's economic
potential in more than two decades.

It has, for example, created

the highest level of housing affordability since 1973 and one of

the lowest costs of productive capital since the early 1960s.


believe that the potential benefits of this environment are
widely recognized and are the principal cause of our currently
improving economy.
The financial health of the private sector has also improved

Chairman Greenspan's metaphor of diminished

financial headwinds is a particularly apt description of this

The deleveraging process which began in the late

1980s has left corporate balance sheets in a much improved
many years.

Our nation's banks are as healthy as they have been in
Furthermore, the decline in mortgage interest rates

and other key lending rates have worked to lower the overall debt
payment burden of the household sector.
• But tonight I would like to focus on the household sector
and take a detailed look at consumer spending patterns and
balance sheets.

I believe that the details suggest more cause

for concern than the aggregate statistics indicate.

For what

seems to be one of the best of times financially for our country
as a whole stands, in contrast, to what is arguably one of the
riskiest times that large parts of the household sector have
faced in many years.
While the conventional forecast of sustained three percent
growth is certainly the most likely economic scenario, I believe
that the financial condition of much of' the household sector
poses one- of the most significant downside risks to the current
economic environment.

Note that I say "much of" the household

sector, for many households are doing quite well, historically

Furthermore, my concerns do not reflect on the short

term outlook, but on the long-term sustainability of the current
level of household spending.
My case tonight is that the aggregate statistics mask the
potential risk to key components of the household sector for two

First, financial innovations and tax law developments

have induced many households to assume a less liquid 1 lancial
position than has previously been the case.

Second, demographic

and labor market developments have weakened the financial
position of the core households of our country -- the middle-aged
and middle-class -- far more than the aggregate statistics
To better understand how financial innovations and changes
in the tax laws have altered household behavior, let us begin
with a comparison of household spending behavior in 1993 versus

Personal income in 1993 was $244 billion higher than the

year before.

Uncle Sam took $37 billion more in personal taxes,

meaning that disposable personal income was $207 billion higher.
But, personal outlays rose $254 billion.

In Keynesian jargon,

the marginal propensity to consume last year was about 1.2 -$1.20 extra was spent for every dollar of extra income.
of course, unsustainable in the long term.

This is,

The short-term result

was a decline in the personal saving rate to 4.0 percent -- one
of the lowest on record.
In and of itself, this historically low saving rate suggests

that households were stretching their budgets last year to
maintain their consumption.

But, a closer look at the components

of savings suggests that their "stretching" was even more extreme
than the data indicate when it comes to the liquidity of their
balance sheets.
One of the most striking elements of the data was the
increased use of tax sheltered forms of saving, notably a buildup of pension fund and life insurance reserves.

In 1992, the

buildup of reserves in these two tax sheltered vehicles amounted
to half of the net acquisition of financial assets by the
household sector.

In 1993 however, these tax sheltered

investments rose to nearly 70 percent of net acquisition of
financial assets.

So while total personal saving fell $47

billion in 1993 compared with 1992, tax sheltered personal saving
rose $67 billion.

That means that saving outside of these two

vehicles had to fall by $114 billion between the two years.
The ramification of this $114 billion decline in other
savings becomes clear when one examines where this decline in
unsheltered saving came from.

Data on household balance sheets

suggest that it primarily came from bank accounts and sharply
reduced net purchases of liquid financial market securities.


are all aware that banks have received a smaller portion of
household wealth as a result of today's interest rate

Although bank accounts did manage to grow by some

$17 billion in 1992, the total amount of bank deposits actually
fell $18 billion in 1993.

On net, therefore, bank deposits


represented $35 billion of the decline in unsheltered saving in
1993 compared to 1992.
The household ownership of securities has also seen a rapid
transformation with the rapid growth of mutual funds.

In 1992,

households acquired, on net, individual securities • (including
security credits) of $58 billion and increased their mutual fund
balances by $171 billion.

Broadly speaking therefore, net

investment in all securities rose $229 billion.

In 1993, by

contrast, net acquisition of individual securities fell $115
billion while purchases of mutual funds rose $269 billion.


net investment in all securities in 1993 therefore was only $154
billion, down $75 billion from 1992.
Much has been made of the mutual fund boom in 1993.


the growth of funds that year did set a record, total purchase of
securities by households both directly and indirectly through
mutual funds was much lower in 1993 than 1992.

These numbers

show that the accelerated liquidation of individual stocks and
bonds by the household sector swamped the increased purchases of
mutual funds in 1993 relative to 1992.
The reason for such behavior is certainly understandable.
The increased use of tax sheltered vehicles is a normal reaction
to the higher rates which were retroactively imposed in last
year's tax bill.

Individuals are clearly increasing the share of

their assets which they hold in taxable form.

But tax sheltered

investments are not perfect substitutes for taxable investments.
In other words, the difference is more than just taxes.


are three key differences which impact the financial health of
the household sector.
The first difference is that the buildup of pension and life
insurance reserves is an inherently less liquid form of saving
than are either bank deposits or securities.

Withdrawal is

seldom easy, is often constrained by law or covenant, and often
involves punitive tax consequences.

This reduced liquidity

suggests that, in an emergency, households have less easy access
to their assets than they otherwise would.
The second difference involves the actual dollar value of
saving which is occurring.

The data on tax sheltered saving are

reported in before-tax magnitudes.

When withdrawn, most of this

saving will be taxed at a federal tax rate of 28 percent and a
state tax rate of around 5 percent.

As actual after-tax net

wealth goes, each dollar of tax sheltered saving is really only
worth 67 cents.

Furthermore, because of a 10 percent withdrawal

penalty, in liquid terms, each dollar is really only worth 57

On a tax-adjusted basis therefore, household savings is

even lower than the aggregate data indicate.
The third difference is a distributional one.


with access to defined contribution pension plans are generally
more affluent than, say, holders of bank accounts, or even mutual

While there are an estimated 14 million households with

401(k) or 403(b) plans, 18 million own mutual funds and 83
million of the country's 96 million households have bank

The shift into tax-sheltered saving by the household

sector as a whole may indicate that those households most likely
to be financially vulnerable are seeing the biggest reduction in
their financial cushion.

The reason this difference is of

concern is the qualitative version of the decrease in household

Not only are households less liquid in the aggregate,

but it may be that those who need liquidity the most are the most
financially stretched.
This data on household wealth is consistent with evidence I
will present later on household earning that middle income
households are increasingly stretched.

While the data are not

sufficiently disaggregated to prove the point conclusively, they
strongly indicate both lower levels and less liquid forms of
saving than aggregate evidence suggests.

Most important, the

data suggest that relying on historical comparisons are tricky
and liable to overstate the current financial health of
A look at the liability side of the household balance sheet
also reveals trouble spots.

In order to finance increased

spending, installment credit rose $38 billion faster in 1993 than
in 1992 while mortgages grew $20 billion faster.


residential real estate sales were growing along with purchases
of consumer durables, so we should expect some rise in this type
of debt.

But, a closer look at the numbers indicates that a good

part of the increase in mortgage debt was probably not being used
to finance home purchases, but was probably diverted to


Let's look at mortgages.

During 1993, total household

ownership of residential structures and the land on which they
sit rose $221 billion in value.

During the same period, mortgage

debt owed by households rose $191 billion.

In other words, the

rise in mortgage debt was 86 percent of the rise in the value of
residential property.

But the standard bank or finance company

mortgage has a standard loan to value ratio of only 80 percent.
This 80 percent figure actually exaggerates the maximum
plausible amount of mortgage financing of home purchases.


is because while new homeowners may be mortgaging their
properties to the hilt, existing homeowners are in the process of
retiring the mortgage debt on the property they bought in
previous years.

On average then, the increase in mortgages in

any one year should be well under 80 percent of the increased
value of residential property holdings.
Even in periods such as the 1980s, which are supposedly eras
of high debt growth, the rise in mortgages is much less than 80
percent of the rise in household ownership of residential

Between 1981 and 1989 for example, mortgages rose

$1253 billion while residential property holdings rose $2486
billion, implying a 50 percent ratio of additional mortgages to
additional property holdings.

Applying this 50 percent ratio to

1993 suggests that $80 billion of the $191 billion rise in
mortgage debt during that year was used for consumption or some
purpose other than strictly home purchase.
Tax advantages might explain why an individual household

would opt to borrow against the equity value of its home since
mortgage interest payments are tax deductible while personal
interest payments are not.

While explainable, the increased use

of mortgage finance contributes to the rise in long-term debt to
finance present consumption.

This might be considered a

potentially troubling development for household balance sheets.
On the debt front, much has been made of the declining debt
service share households face due to the decline in interest
rates and the retiring of some debt during the early 1990s.


example, debt service as a percent of disposable income declined
to 16 percent in 1993, back down to the 1985 level and not far
from the historic norm of 15 percent for the 1960-1985 period.
The case of the debt service ratio illustrates my second point
about aggregate statistics -- they can mask the effect that
demographic trends have on the household sector's financial
For example, consider that the elderly and the very well-todo are unlikely to engage in significant debt financing of
consumption for obvious reasons.

Yet, these two groups of the

population have enjoyed a very significant rise in the share of
income they receive.

In 19 60, for example, the elderly received

just 7 percent of household income.
16 percent.

By 1993 their share was over

Similarly, the share of income going to the top 1

percent of households has risen from 9 percent in 1960 to 13
percent today.

In aggregate therefore, the share of income going

to these two groups has risen from 16 percent to 29 percent.

Consequently the share of income received by everyone else - those likely to borrow -- has fallen from 84 percent of income to
71 percent.
This means that even a constant debt burden as a share of
total income actually involves a much higher burden on those
likely to be shouldering that debt.

If one adjusts the debt

service burden for these demographic factors, the average debt
service burden for those most likely to borrow during the 19 601985 period turns out to be roughly 18 percent.

The 1993 level

of this measure of debt service burden stands at 23 percent.
Instead of being within one percentage point of the historic
norm, debt service has risen by 5 percentage points to consume 2 8
percent more of disposable personal income for this group.


terms of total debt outstanding, the demographic adjustment
raises the ratio of total debt outstanding to disposable income
from 77 percent to 108 percent.
This squeeze on the relative income position and consequent
ability to service debt by the middle-aged middle-class is also
evident from an analysis of the functional distribution of income
-- how the pie is divided among wages, interest, and dividends.
By and large this middle group lives off the wages and salary
income it earns.

While 89 percent of all wages are received by

those under 65 in households making less than $200,000, only 30
percent of interest is received by such households and only 28
percent of dividends.

Of course, non-elderly middle class

households are also negligible recipients of transfer payments.

During 1993, however, only 38 percent of the gains in
personal income were paid in the form of wages, net of social
insurance taxes, 28 percent was paid as capital income and 22
percent paid in the form of transfers.
represented fringe benefits.

The remaining 12 percent

This is the lowest wage share of

personal income gains in recent times.

Even during the 1981-1989

period, an era which some political mythology describes as an era
in which the middle-class got squeezed, 52 percent of income
gains were paid in the form of wages, 28 percent was paid to
capital, 14 percent in the form of transfers and 6 percent in

Furthermore, forecasts of personal income growth

suggest that this wage share is unlikely to grow much and will
not attain the levels of the 1980s.
Compensation and productivity data tell much the same story.
Between 1981 and 1989, output per hour in the non-farm business
sector grew 8.7 percent.
3.9 percent.

Real compensation per hour rose only

This higher compensation absorbed about 45 percent

of productivity gains.

During 1993, output per hour in the non-

farm business sector grew 1.5 percent.
hour rose only 0.2 percent.

Real compensation per

Thus, higher compensation absorbed

about 13 percent of productivity gains.
Thus, data from a wide variety of perspectives show that the
income base of the middle class eroded in 1993 relative to the
experience of the 1980s.

This occurred at the same time that

debt burdens on this group maintained near-record high levels and
households were stretching their budgets to historically

unprecedented levels to maintain consumption.

There is not a lot

of evidence that these pressures will be alleviated at any point
in the foreseeable future.
Further, this data seems to be reflected in consumer

Last fall, Money magazine and ABC News did a

comprehensive interview of 2,154 American households.


respondents clearly recognized that a national recovery was
underway and were optimistic about the nation's economic future.
Furthermore, they rated this a very good time to buy both homes
and cars, no doubt a reflection of the current low interest rate
environment which holds down monthly payments.

But, households

were far less sanguine about their personal finances.

They did

not feel that the general national picture affected their daily

One other key finding in the Money/ABC poll summarizes

this concern.

In spite of a clearly better economy in late 1993

relative to a year earlier, and widespread media coverage that
this was so, the number of respondents seeing America in long
term decline was up 13 percentage points from 43 percent to 56

This may be evidence that, in their own way, households

recognize the potentially unsustainable condition of their
Americans seem to be currently stretching their finances
because they see the business cycle induced improvements in the
economy and are therefore willing to assume a riskier position
for their balance sheets.

This pattern was very similar to what

happened during the 1980s when the personal saving rate fell from

nearly 8.6 in 1982 at the trough of the recession to 4.3 percent
by 1987, with expansion fully underway. Following the 1973-1975
recession, the personal saving rate fell from 8.9 percent at the
recession's trough in 1974 to 6.3 percent by 1977, when that
expansion was fully underway.
The importance of these declines in saving as a spur to
inducing economic growth should not be underestimated.

In the

1980s, for example, had the personal saving rate stayed at its
19 82 level, personal consumption expenditures would have been
$131 billion lower in real terms in 1987, ceteris paribus.


$131 billion represents 17 percent of the growth in real GDP
during those 5 years.
In the current cycle, the personal saving rate dropped from
a high of 5.3 percent in 1992 to 4.0 percent in 1993.

Using the

same analysis as above, a real increase in personal consumption
of $56 billion was due to the decline in the saving rate.


$56 billion amounts to 38 percent of the $146 billion increase in
real GDP in 1993 over 1992.

Thus, the decline in personal saving

was more than twice as important last year as it was during the
1980s as a spur to growth in the overall economy.
An obvious problem for economic growth in the next two years
is that a sustained decline in the personal saving rate as a
means of financing higher consumption seems unlikely.
represented a record low in this measure.


Furthermore, the 1993

tax legislation is going to affect household cash flow for the
first time this year.

This tax increase is likely to lower the

saving rate still further, all else equal.
The Administration correctly argued that the drag on
spending from their proposed tax increase would be reduced
because the great majority of the increase fell on upper income

The Administration's 1995 Budget

(p.59) shows that of

a $41.3 billion increase in taxes, $32.9 billion or 80 percent,
falls on taxpayers earning over $200,000.

Just for argument

sake, let's say that the 80 percent share that the wealthy pay
comes out of personal saving while the remaining 20 percent comes
out of consumption.

That would mean that the tax increases would

cause a mere $8 billion drag on aggregate expenditure.

But, it

would involve a 14 percent decline in personal saving and a
personal saving rate of 3.5 percent, without any increase in
Where this fall in saving would come from on the household
balance sheet is certainly an area of pure conjecture, but the
magnitude of the adjustment involved is troubling.

For example,

let's say that bank deposits stay even in 1994 with their 1993
level while other parts of the household balance sheet continue
at their 1993 pace.

Then, net purchases of financial market

instruments including mutual funds would fall to just $83
billion, one third of their 1992 level.

Such a decline could

have significant repercussions for market performance.
In sum, I believe that the household sector poses one of the
most serious risks to the continuation of this recovery.


I believe that the consensus forecast of roughly 3 percent growth


is the most likely outcome for the rest of 1994.

But, we must

realize that the situation in the household sector is far less
positive than it was during the 1980s.

Not only is personal

saving at a record low, but households are less liquid than at
any time in memory.

Furthermore, the capacity for income growth

to improve those balance sheets is greatly inhibited as the wage
share of income growth is also at historic lows.


shocks to the system from higher taxes, higher energy prices, or
even significantly higher inflation, could cause difficulties for
what is already a challenging situation.