ALFRED
P.
WORKING PAPER
SLOAN SCHOOL OF MANAGEMENT
THE ROLE OF PRIVATE PENSIONS
Daniel M. Holland
WP #1031-78
November, 1978
MASSACHUSETTS
INSTITUTE OF TECHNOLOGY
50 MEMORIAL DRIVE
CAMBRIDGE, MASSACHUSETTS 02139
THE ROLE OF PRIVATE PENSIONS
Daniel M. Holland
WP #1031-78
November, 1978
Grovs/th
of Pension Plans
,
While private industrial pension plans have
least,
at
set
it
is
long history,
their accumulated funds have become
capacity for vigorous growth
is
A picture of rapid growth
a
in
retirement.
major financial
is
It
institution.
deserves
a
plans
Is
to the point,
This
closer look.
2
for state-local
Table
in
1
government planb.
While, on the various dimensions of growth, the Tables "tell
trial
major
Over this same period,
borne out by any of the indicia
industrial plans or Table
few highlights can be noted.
a
important evidence of the viability and inherent
strength of private pension plans;
a
back to 1875
only within the last 35 years that they have emerqed as
of arrangements for income support
for private
a
their own story,"
The number of vvorkers covered by private Indus-
now (1975) over seven times as great as thirty years ago.
the percentage of employees
in
non-agricultural
More
industry (excluding
government workers who are covered by their own plans) has gone from under
fifteen to close to fifty.
Coincident with the growth of coverage,
of course,
has been a huge
expansion of the fiscal operations of industrial pension plans.
a)
Contributions
b)
The number of beneficiaries
in
1975
ran at
100
Thus:
times their 19^0 level.
increased more than 35 times over this
period and benefit payments were ]k8 times higher.
c)
With contributions exceeding benefit payments
(and previous accumu-
lations earning interest, dividends, and capital gains
the pension funds hcive continued to grow.
net balances,
Indeed, by 1975 the fundr.
themselves had Increased to over 100 times their
value terms.
In
19''0
level
In
book
Much the same story, with different orders of magnitude, of course,
can be told for state-local
we forego
pension funds.
But
in
the
interests of brevity,
here, except to note that there are suggestions that these funds
it
are accelerating vis-a-vis those of industrial
pension plans.
It
is
these
latter that receive prime attention when the impact of pension fund accumulation
is
discussed.
But
the plans
have great accumulative potential
_/
for employees of state and
Pensions:
A Summary Repor t l^lational
and Daniel
M.
of Economic Research,
F.
—
Murray, Economic Aspects of
Bureau of Economic Research,
Private Pension Funds:
Inc.,
governments
that has not been properly appreciated,
For a discussion of this matter, wee Roger
Holland,
local
Inc.,
1968)
Projected Growth (National Bureau
1966).
Summary data appear
in
Table
What we have
in
essence, then, wi tnessea over the last 35 years, has
been the emergence of
a
2.
private pension structure of imposing magnitude.
Underlying this development are
and demographic factors.
a
number of broad soc iologi cal -economi
c
changes
Particular influences such as income tax encourage-
ments and advantages have also been important.
The introduction of high and sliarply progressive rates of personal
income
tax made compensation arrangements on which tax payments could be deferred more
attractive.
For they provided a way of injecting some long-run averaging
taxation of the income stream, and,
liability.
for the usual
During the war, more particularly,
case,
a
in
the
lower life-time tax
interest of workers in pensions
was enhanced because the stabilization policy limited direct wage increases.
For the same reason employers utilized pension arrangements to a greater extent
to attract or hold scarce workers.
Quite likely, although perhaps not
rationally, the high corporate rates levied at that time made pensions seem
relatively "costless" and so hastened their adoption or extension to more workers
"perhaps not rationally" because wage increases would also be
_/
a
deductible
business expense, and because high wartime rates might well not be expected
to prevail
peacetime.
in
as did the provision of the
Internal
Revenue Code of
19'<2
required that
that
for a pension plan to be non-discriminatory, and hence have the employer's
contributions to
it
free of corporate tax,
to top officers or selected small
it
groups of employees.
the spread of pension programs came from the
of the coal strike settlement
Relations Board
in
IS'iS
in
must not narrowly be directed
19'^6,
Further impetus to
inclusion of
a
pension as a part
the decision by the National
(upheld by the Supreme Court
in
IS'iS)
Inland
the
in
Steel Case that pensions are a bargainable issue, and the Steel
Labor
Industry
Fact Finding Board's recommendation of a pension program in ]3hS.
But all
Is
The growth of pension programs
these are immediate factors.
also a response to deeper underlying needs and changes
structure and economy.
An
than
7
times,
the country's
important demographic basis for the grov/th
pension programs has been the steady and substantial
population.
in
increase
in
in
the older
Between 1900 and 1975, those 65 and over increased almost more
from 3-1 million to 22.
h
million,
a
rate of increase considerably
greater than that for the total population.
Thus, persons 65 and over repre-
sented A.l percent of the U.S. population
1900,
9.5 by 1965, and 10.5 percent by 1975/
For the 1900 data see Historical
—
in
5-3
in
1929, 7-5 by ig'iS,
These figures reflect increasing
Statis tics of the United States
,
p.
7;
for the data for other years see Economic Report of the President, January,
1978.
p.
287.
life expectancy,
but an even stronger thrust
for pensions comes
from the
steady decrease in the number of years those who survive to say, 60, can be
expected to spend
_/
in
—
employment.
So while there has been a rather mild
This trend predates the rapid growth of pensions, but could very well also
have been accelerated by their development.
So to some extent
there
is
a
chicken-egg argument here.
increase
in
life expectancies,
of non-working old age.
_/
—
there has been
which 11.5 years would be spent
counterpart
in
in
life expectancy of
the labor force and 2.8
]ShO had only a slightly increased
but expected time in the
the years
l4.3 years of
in
retirement.
life expectancy--l 5.
1
His
years--
labor force had declined to 9.2 years and expected
retirement increased to 5-9 years.
This sharp change in years of non-working
old age predates the groundsv^ell of pensions.
not much change since 19'^0,
as
in
The movement of population from the countryside
1900 a white male at age 60 had a
In
sharp increase
a
for similar data
the data of I96O refers to all
A more recent estimate shows
(although not strictly comparable
male workers),
indicate that a male worker
who survived to 60, could expect to live 15-8 more years of which 8.5 would
be
in
the labor force and 7-3
in
retirement.
(The data for 1900 and
19^0
are from United States Department of Labor, Bureau of Labor Statistics, Tables
of Working Life; Length of Working Life for Men
Washington,
1950,
of Working Life,
p.
hi.
for Men,
The
,
BSL Bulletin No.
I96O estimate appears
in
I96O," Monthly Labor Review
,
1001,
Stuart Garfinkle, "Tables
July,
19&3,
p.
822.)
to the city and from agriculture to manufacturing and service trades also
pushes
support
in
in
the direction of the replacement of informal
family-oriented income
old age to more formal work-related arrangements.
The growth of pensions,
then, appears to be a response to underlying
structural changes in the economy as well as to particular tax advantages.
And this growth also suggests that
income support
old age may be a "commodity"
in
with an income elasticity greater than one; as income expands the portion of
it
devoted to providing for old age increases more than proportionately.
Perliaps,
it
not coincidence that
is
change
last 25 years or so there
the
in
the way workers prefer to receive their
appears to have been
a
productivity gains.
Traditionally, the fruits of economic progress have been
realized
hours worked.
Since
IS'^O,
_/
In
in
It
in
income and some decline in
real
manufacturing have,
prior years there was
a
if
anything,
risen somewhat,
tendency for the average work week to decline.
I9A0 the average weekly hours worked for production workers
facturing was 38.1,
was
in
with the except im of the war years, average weekly
hours of production workers
whereas
increase
the form of some
in
in
39.'*,
in
in
1950
it
was AO.5,
1965
in
it
was
A 1.2,
in
manu-
and in 1975 it
1977 it was A0.3.
was this observation that led to Roger Murray's suggestive conclusion that
"employee preferences have shifted to some extent, from the desire for more
leisure during their v;orking years to
a
greater degree of financial
during retirement or to earlier retirement."
_/
Roger
F.
—
Murray, Economic Aspects of Pensions:
Bureau of Economic Research,
Inc.,
1968, p.
independence
5.
A Summary Report
,
National
Murray notes, however, that
vacations and paid holidays did increase over the period during which average
hours worked remained ostensibly unchanged (or even increased slightly).
the point
study.
is
not unambiguous.
V/e
leave
it
as suggestive and deserving
So
further
Coverage
Private industrial pension plans, despite their extraordinary growth
in
last 30 years,
the
still
cover less than half the work force that seems
eligible for them--employee5
non-agricultural establishments, excluding
in
government employees who are covered by plans of their own.
is
shown
Table
in
1,
the rate of growth of the coverage
appears to have tailed off over
last decade.
the
And, as
percentage's
So private
industrial
pension plans may, very well, never cover much more than half the seemingly
eligible working population.
To some observers this
writing
in
1971,
is
a
being voluntary they now cover
And he added, "
only about half of all employees
pensions are ever tofijlfull
_/
Gilbert Burck,
considered this to be "....the most important deficiency
of private pensions."
cover nearly all
very serious defect.
a
in
the private economy.
social
private
they must be expanded to
role,
the working population."
if
—
Gilbert Burck, "That Ever Expand ng .Pens ion Balloon," Fort vine
i
1971,
p.
October,
13'4
Private pensions, are,
Workers participating
tnconie
,
tax is due
in
in
fact,
not a strictly private affair.
them receive preferred tax
t
reatment ,--no
from the covered worker at the time the contributions
are made into the fund by the employer on the worker's behalf, and fund
earnings accumulate free of tax.
he
receives the benefits paid out
taxed when
In
retirement the worker
if
the employer's contributions and
is
earnings on both the employee's and employer's contributions.
of the extent of the "public" interest
aggregate of income tax postponed for
is
a
the
long period tiiis way.
One measure
In
Fiscal
1976,
the foregone tax revenue was about $6 billion,
for the Tax Expenditure Budget
States Government,
_/
individual worker the
For the
$1,000
pp.
.
pre
108-109).
-
ference can be illustrated as follows:
year earning interest at Gt would,
a
estimated
Special Analyses, Budget of the United
in
Year 1976
Fiscal
(as
in
the absence of taxes,
accumulate to $79,058 over 30 years.
For a worker in the
bracket who could invest the interest free of
20'^
tax but had to pay tax on the
And for a
be $63,2'47.
If
initial
contribution, the accumulation will
bracket worker
502;
it
would be $39,529-
neither the interest nor tha annual contribution were free of personal
Income tax, the accumulations would be $51,361
(20% bracket)
and $23,782
(50% bracket)
While,
in
the workers are
retirement,
likely to be in lower tax brackets,
even were that not the case, substantial savings would result.
Paying personal
tax on his pension fund accumulation as it
would leave the worker
•$51|36l
in
is
paid out
the 20<^ bracket with $63,2'i6 as compared with
he'd have left after current tax on annual
principal and interest.
For the 50% bracket worker the after-tax amount would be $39,529 compared
with S23,782.
Therefore, given the public investment
a
legitimate concern with how well
to expect substant ial ly full
all workers
Is
in
unrealistic.
in
they achieve
coverage
in
a
social
purpose.
is
But
of the apparently eligible population--
non-agricultural establishments
For
private pensions, there
(except government employees)--
the underlying potentially eligible population are
10
two large groups of workers who,
in
a
realistic sense are not "suitable"
for employment- rel ated pension plan coverage.
in
It
is
not the industries
which they work that account for this characteristic; rather
their conditions of employrrent.
Thus,
it
is
people who characteristically are
employed part-time do not really constitute grist for the pension mill
in
the sense that they do not have a strong enough association with the
employer, and would be inconvenient to service.
A similar conclusion
applies for young workers who are still exploring, seeking to discover
their employment and occupational niche, and
highly mobile.
"Young"
in
this connection
v^ho,
is,
consequently, are
of course, arbitrary.
But twenty-five or under might be a suitable cut-off point.
If
adjustments both on the score of young workers and part-time
workers were made, we would have
a
more relevant base against which to
assess pension plan coverage.
I
have made these two adjustments on an earlier occasion, and found
these two groups of workers to represent about 30 percent of the underlying
—
population.
_/
Daniel
M.
Assuming this percentage to apply
Holland,
Private Pens ion Funds:
Bureau of Economic Research,
1966,
p.
in
more recent years
Projected Grov/th, National
21.
percent
as well, we can estimate that private pensions now cover about 70
of those we can realistically expect them to cover.
Thus, although the
extent of coverage seems to be increasing only slowly over time,
currently
is
it
quite high--in the aggregate about 70 percent of those who
realistically could be expected to participate
in
a
pension plan.
Closely related to the criticism that the coverage of private pension
plans
is
unduly low, however,
is
the criticism that coverage is very
diverse as between different industries and among firms
in
Information on country-wide industry coverage appears
Estimated coverage
of Table
1
in
this Table
is
1
do^
given industry.
Table
in
3.
not comparable with the magnitudes
above, being defined differrntly.
data of Table
a
and those of Table
3
do not
Among other things, the
workers under
include:
profit-sharing plans with retirement features and other plans that make
lump-sum payments, those
in
plans that cover less than 26 vvorkers, and
covered employees of non-profit ornan zat ions (other than labor unions).
i
Despite these differences, Table
Indeed,
the data therin are,
to
3
tiie
is
a
primary body of information.
best of my knowledge,
the most recent
and thorough summary of pension covereage by industry sector.
From Table
3
we get an interesting and essentially valid picutre of coverage differentials among industries, albeit
/
a
general
understatement of coverage.
Because of the exclusion alreay noted.
Finally,
j-espect
—
/
the Table also provides a sense of how coverage has grown with
to the
As to the
labor force.
latter point first, mild headway seems to have been made in
the extension of coverage.
government workers) as
a
For the employed civilian
in
1967,
(excluding
whole, coverage has grown more rapidly than
employment; the coverage percentage increasing from
of the total
labor force
and to hS percent by 1972.
3'*^
in
1S61
to 36^
13
Manufacturing, Mining and Transportation and Public Utilities are
industries of "high" coverage (from 60%
70^); Construction, Wholesale
-
and Retail Trade, and Service are industries of "low" coverage (30^
These averages reflect the varying proportions
(typically high coverage) and small
firms
coverage).—
_/
firms
Insurance and Real
low coverage reflects
a
Estate
(which tend to have
multiplicity of small
from other sources suggests that coverage
is
institutions, but this information
a
sector whose
firms and agencies.
very high
is
low
in
mixed bag of
a
is
35?).
each industry of large
The average picture has not changed much
Finance,
financial
in
-
in
Evidence
larger
the
swamped by the predominance
of smaller firms characterized by low coverage.
years
recent
(from I96I
through 196?)
.
Relative rankings among industries
remained the same, and except for Construction and Transportation and Public
Utilities,
the degree of coverage did not change much.
So it appears from this review of the
1.
In
a
industry
data that:
number of industries, private pension plans cover
a
minority
fraction of employed workers.
2.
This condition
is
not tending to improve substantially over time.
These data on differences
in
the degree of coverage reflect,
extent, particular characteristics of the various industries.
In
to some
a
number
of industries, Construction for example, the employee's association with
a particular employer tends
to be transient, and pension coverage on this
score could be expected to be lower than in industries where workers have
a
long-lived relationship with
a
single rMiiployer, since mul
t
i
-employer
pension arrangements are more difficult to work out than plans tor individual
firms.
But
industry variations
in
coverage ratios also reflect
]k
differences between industries tn levels of pay, degree of unionization
—
and size of employer establishment.
_/
Of course,
these characteristics,
too, are correlated.
Larger employers
tend to pay higher v^ages and be more unionized.
Evidence to this effect appears
of
tiie
previous paragraph for
a
in
Tables
h
and 5. To pursue the point
moment, note in Table
5
that the fact that
Transportation, Construction and Utilities are high coverage industries
(expenditure for retirement plans are made on behalf of 77 percent of
employees) and
Trade and Service,
Is'iundoubtedly explained
in
industries of low coverage (38 percent),
part, at
least, by the fact
that the
latter
coverage
is
are characterized by smaller "establishments."
The two Tables
{k
and 5)
confirm that,
general
in
more
likely among higher paid workders than those who are lower paid, workers
who belong to unions than those who don't, and workers
ments than those in smaller places of work.
which these Tables are taken puts
Or as
in
larger establish-
the brief report
from
it:
"In summary, workers who are employed
in
small, non-union establishments
at
relatively low levels of pay are the least likely to be participating
in
a
retirement plan.
Rapid improvement
poor economic position of many small
is
firms.
effectively blocked by the
This situation
is
a
matter of serious concern because these workers are among those least
able to provide for themselves
_/
,
July 1971
later years."
—
"Incidence of Private Retirement Pla'is," Monthly Labor
Errerson Bier,
Review
in
,
p.
^0.
The phrase "poor economic position of many small
firms" directs attention to
17
the margi nal
f
i
rm
,
the
Similarly, the phrase
worke r,
firm that apparently cannot afford pensions.
"relatively low levels of pay" suggests the marginal
the worker who cannot afford to defer current
receipt of any portion
of what the employer can afford to pay him.
These suggestions are set forth more explicitly
in
another section
of this same report:
"While the economic fortunes of
a
funds available for its employees,
company, or industry,
limit the
their preferences largely dictate
how these funds are allocated among wages and other forms of comWages are
pensation.
a
very large part of the total at
of compensation, and must be the employees'
low levels
first concern.
Unless
workers' earnings are sufficient to meet their ordinary living
expenses, many of the common employee benefits are viewed as frills,
particularly those payable
in
the distant
future.
Consequently,
In
establishments with average compensation below $2.50 an hour, 8 out
of 10 employees were in groups without
Ibid.,
_/
In
a
p.
—
retirement expenditures."
38.
nutshell, the root causes of low private pension plan coverage are
marginal
and marginal workers.
firm.s
If
many small
firms lack the financir;!
strength to take on pension cists and many workers, receiving low pay, cannot
afford to "save" any of it, should we expect private pension plans to encompass
them?
role"
In
In
my judgement,
requiring private pension plans to "fulfill
respect to these employers and employees
objective for
a
is
work-related private sector program.
never pay workers more
ttian
to set an
a
social
inappropriate
For employers can
they are vjorth, and workers whose marginal
net
18
low cannot afford to accept any of it as deferred pay,
revenue product
is
but need
to live on now.
it
all
19
Benefits and Other Plan Provisions
The 35-year record of the
capture the real
past summarized in Tables
1
and
2
does not
importance of the benefit payments to be expected from
private pension plans.
Only
sizable numbers of those
in
the most recent years of the Tables have
were
viho
in
the work
were instituted gone into retirement.
force when pension plans
From now on the growth
in
beneficiary
numbers and amounts will be more substantial.
Even so,
the dimensions
reached in the Table are not
the number of old-age beneficiaries under private
By 1975,
state-local government plans came to ^3
(Those receiving private pensionsvere
OASI.)
in
percent of OASI
inconsequential.
industrial and
recipients.
very large part also recipients of
in
While the average benefit payment of private industrial plans grew four-fo'
current dollars from IS'^O-IS/S,
in
real
terms average benefits grew very slight
Incorporating the older provisions of plans, subsequently substantially
modified, or the more modest pensions paid to workers who had been under
the plan for a relatively short period,
is
.
not
this behavior of benefit payments
likely to characterize the future.
Most workers are covered by "conventional" plans whose benefit formulae
are related to salary
(as
contrasted with pattern plans which cover the
workers who are members of the major international unions and under which
benefits are set as a given amount per month per year of credited service).
And most conventional plans are now based on final pay.
_/
—
Therefore if wages
This represents a major transformation in private pension plan benefit
formulae.
In
1955 only 38 percent of plans surveyed by the Bankers Trust
had final-pay formulas.
Dy
197'*,
78 percent of conventional
Survey had benefits based on final pay.
plans in the
See summary of Bankers Trust studies
20
Alfred
In
Bui letin
,
Skolnik, "Private Perjsion Plans,
M.
June 1976, p.
1950-7'*," Social
13-
and salaries continue their long-run trend of
participants
private plans are promised
in
indexed for actual
be
essence
in
growth,
benefit that will
a
With inflation running at
benefit payments could decline
Understandably, therefore,
purchasing power during returement.
in
real
inflation over their working life.
the brisk pace of the last five years,
seriously
to 3-5 percent
3
indexing ceases on retirement.
But this
Security
there has been increaded
interest
in
cost of living adjustments for retirees.
To date, however, such arrangerrtents are not characteristic of private pensions.
The Bankers Trust Survey reported adjustments of this kind
percent of the conventional plans
in
its
197'*
for about 6
Some un ions--al umi num,
Survey.
container and auto, for example, have bargained for cost-of-living increases
for ret rees
i
.
—
Skolnik, p.
_/
\k.
Some companies have sought to achieve the same objective with a variable
annuity option under the basic plan.
About
11
percent
Trust sample of conventional plans provided this option
companies adjustments to pension benefits of retirees
in
197'*.
In
other
recognition of cost-
in
of-living increases, have been made on an ad hoc basis.
—
Of
the companies
Ibid.
_/
surveyed by the National
Industrial
Conference Board
a
variable annuity system was available, and
al
ready ret red.
/
p.
the Bankers
in
Z'l
in
1975,
in
2k percent
percent increased pensions to tho;
/
i
The Conference Board, Compensating Employees:
5k.
Lessons of the 1970s
,
1976
This sample was made up predominantly of large, established companies,
21
The
Internal
discriminate
in
Revenue Code requires that pension benefit formulae not
favor of higher paid workers, but defines discrimination
with respect to social security plus private plan benefits combined.
rule here is that the benefit formula
a
not discriminating if the sum of
is
security benefit plus the private plan's benefit
of the primary sociol
constitute
flat or declining fraction
workers earnings.
reading from low to high levels of
Since the replacement
as average earnings
The
rise,
rate of Social
Security declines
that of the private plan alone could
rise
v;i
th
average earnings and still not be discriminatory.
In
fact,
for conventional
plans this
is
the typical
pattern.
median wage replacement for final-pay plans ranged from 33 to
average compensation of $8,000- $^40,000
for workers with final
and the higher the salary, the higher the replacement ratio."
_/
h]
in
"The
percent
ig^^
—
Ibid.
These numbers are generated by
service at the same salary.
earnings increase at
final
a
calculation which assumes 30 years of
By an alternative method which assumes
percent, an employee wi th
5
years of service and
3D
a
salary of $9,000 would have received from the 197^ median private
plan a benefit equal
to 29 percent of his
final
who earned $20,000, the private plan benefit
Adding the primary OAS
equal
that
I
for the employee
have come to 35 percent.
v/ould
tenefit to both, gives them
to 68 percent and 50 percent,
the extent thct
salary;
a
combined benefit
respectively, of their final pay.
they received additional
OAS
I
To
benefits because of their
spouse, the replacement rate would be higher.
Whether 68 percent or 50 percent
is
enough,
i.e.
whether they represent
about as much in the way of retirement income as workers would like to have
we have not way of telling.
There has been
plan provisions over the post-war years.
In the
a
But
targeted benefit replacenxsnt rates as
great
liberalization of private
this has not shown up as much
in
other features of the plans.
22
From initial pre-occupat ion with retirement benefits per se, the plans
have branched out to death benefits, disability retirement, and early
retirement options.
They have also liberalized vesting provisions substan-
tially, over most of the period under their own steam,
(since 197^) as
and more recently
required by ERISA.
Vesting provisions have been
design, until quite recently.
In
a
major sore point of private pension plan
principle employers and workers had
opposing preferences, with employers preferring late vesting
in
order to
"tie" workers to the firm and keep pension costs down and workers pre-
ferring early vesting which would give them the freedom to change employers
witliout penalty.
Within the working force delayed vesting benefitted steady,
long-term workers at the expense of transient and short-term workers.
Vesting provisions of private plans were liberalized somewhat over
time.
Generally based on age and/or years of service, the typical
ments as summarized
younger age and/or
In
in
a
the Bankers Trust Surveys tended to move to a
shorter term of service.
\S7^ by ERISA which required all
three specified formulae which
in
plans to provide vesting under one of
fully vested after 15.
The private pension plan structure,
in
This development was accelerated
effect would have all workers at least
50 percent vested after 10 years and
evolve
require-
the direction of improvement,
then, has shown an ability to
liberalization and
a
wider range
of options.
The current
level
of replacement rates provided by
private pension
plans and social security, would appear to compare favorably with programs in
other countries.
International comparisons
are
inherently ambiguous, and
more than usually so for social security arrangements, especially combinations
23
of public and private plans.
Haanes-01 sen
however, has developed
,
careful set of estimates for the countries listed
in
Table
6.
a
For the
single worker Social Security alone, there are three countries higher than
the U.S., and another as high.
combined,
if
For Social
Security and Private Plans
the example cited above the the U.S. worker with final
pay
of $9,000 is representative of the average v^orker in manufacturing for
1975,
the 68 percent
replacement rate
Average gross weekly earnings
$9,000
is
a
reasonable figure.
Report of the President,
1978
Over the longer pull
,
the highest of all
countries with
—
the possible exception of Germany.
_/
is
in
manufacturing
in
1975 were $I89.51,
so
(For manufacturing earnings see Economic
p.
299.)
there are good reasons to expect programmed
pension benefits and/or associated features of private pension plans to
continue to become more generous
the past suggests.
scholars
(l8
in
And
all),
it
vjho
is
in
the future.
the judgement,
participated
in
This is what the record of
too, of an expert group of
the "Delphi" study of the
for the Future whose results were published
in
1969-
It
v;as
Institute
their consensus
that the following features of private pension plans would " increase
greatly":
a.
"Higher prospective pension incomes for future retirees permitting
them to enjoy higher standard of living relative to their final
pay than do current pensioners."
b.
"Periodic increases to pensioners income during retirement to
reflect cost of living increases."
25
"Pension portability (i.e., on changing jobs employee's vested
c.
retirement equity transfers to new employee's fund or becomes his
to do with as he chooses."
_/
T.J.
—
Gordon, A Study of Potential Changes
Summary and Conclusions
,
in
Employee Benefits. Volume
I
Institute for the Future, Middletown, Connecticut,
27.
1969, p.
The majority of them also expected the following new retirement benefits
to emerge
in
the next
(counting from 1968)
a.
"Optimal retirement at age 55 with full
b.
"Pension plans extended to include benefits for retirees other
than money, such as:
communities, Legal services."
Ibid.
,
p.
Finally,
benefits for all employees."
Recreation, Education, Medical services.
Residential Costs, Apartment
_/
ten years:
in
Company-sponsored Retirement
—
29.
they also expected a number of other retirement benefits
to "increase slightly."
'
a.
Less stringent conditions for medical
b.
Survivor benefits.
c.
Deferred compensation for
a
significant fraction of workers to
to supplement pensions.
d.
/
Earlier or imnx^diate vesting.
Ibid., p.
27.
retirement.
—
26
An additional
reason for expecting private pension plans to grow
that they have been considered a useful
lever in an
t
i
-
i
is
nf lat ionary
policy, particularly for programs that rely primarily on persuasion.
Such policies sometimes encourage higher pensions
compensation for lower current wage increases.
for example,
in one of
his annual
wage negotiators to "keep
a
(deferred pay) as a
Thus President Eisenhower,
reports exhorted business and labor
statesman-
1
i
ke eye on the
future," and set
lower pay increases currently but with higher deferred pay, such as pensions,
as part of the package.
It
is
interesting to note that considerable liberalization and
enrichment of private pension plan terms and characteristics appear
to have been accomplished over most of the period with a constant ratio of
pension cost to payroll.
(See Table 7.)
28
Pensions and Savings
By a homey analogy,
the fiscal
operations of pension funds can be
likened to a bathtub with contributions plus fund earnings running from the
faucet and benefit payments flowing out the drain.
been (and will continue for many years to be)
every year the level of the water
in
the
tub
The rate of outflow has
less rapid than the
is
intake.
So
higher than the year before.
Indeed these annual accumulations by pension funds are a major component of the
household sector's net asset acquisitions
1975,
for example pension
fund asset growth
for employees of state and
as measured by
oi:
its measured savings.
In
(private industrial plans plus those
governments) came to h8% of personal saving
local
the National
of
Income Accounts; or,
to use a variant measure,
more In keeping with the Flow of Funds Accounts, to }2% of household financial
asset acquisition.
On either measure,
then,
for pension plans--the
long
income tax liability on contributions to the fund
of personal
term deferral
the tax incentives
made by the employer on behalf of his employees and on the earnings of the fund
(plus, of course, all other conditioning factors) --seem to have been an effective
and important
incentive.
Pension fund accumulations are an important fraction
of personal saving and, of course,
the more meaningful
that industrial
in
a
definitional sense they are.
relates to how "real" the saving is.
question
pension funds accumulated $32.5 billion
in
But
The fact
1975 does not
establish that because of their increased funding an additional $32.5 billion
that would not otherwise have been forthcoming is available for capital
formation.
For
it
could be that compensatory offsets to savings that might
otherwise have been made occur
Suppose the price of Scotch
stantially.
in
response to pension fund accumulation.
alone of all spirituous liquors
v^as
cut sub-
Sales of Scotch would go up, and sales of competing beverjges
29
Sales of spirituous liquors as
would go down.
probably by nothing like the increase
class might well
a
go up, but
the sales of Scotch alone.
in
Preferred tax treatment for pension plans vis-a-vis other forms of saving
is
equivalent to
a
cut
in
their price
return they provide to savers).
but savings
in
other forms,
(really it
is
a
relative increase
So pension saving should go up, which
(i.e., made via other arrangements)
the
in
has,
it
insofar as they
could serve some of the purposes that pension savings are made for, may be
expected to decline.
On net balance,
considerably less than the increase
then,
in
saving may rise, but by
total
This
pension fund saving per se.
is
where we would come out by deduction. The argument says something about the
nature of the effect, but has nothing to say about its magnitude.
depend importantly on the degree of subst tutabi
i
1
i
That vjould
ty between pension fund
accumulation and other forms of saving.
•
Additional
theoretical support for the view that pension saving
is
made primarily at the expense of other saving comes from the life-cycle
They assume
hypothesis developed by Modigliani and others working with him.
that "consumption and saving decisions of households at each point of time
reflect a more or less conscious attempt at achieving the preferred dis-
tributionof consumption over
life cycle," but do not distinguish among
the
different assets accumulated for this purpose.
_/
—
Implicitly,
to the
Franco Modigliani, "The Life Cycle Hypotliesis of Saving, the Demand for
Wealth and the Supply of Capital," Social
Research
,
Summer,
1966,
p.
162.
extent that the desired total of saving contains pension fund accumulations,
it
will
incorporate commensuratel
There
is
y
less
in
other forms.
also some crude empirical evidence that seems to support the
30
expectation that pension fund savings will not,
to the flow of personal
saving, but will
in
the main,
be net new additions
rather be made at the exnense of
Over half a century--'f rom 1897 through \Sh3--
alternative forms of saving.
Raymond Goldsmith
in
his study of saving
the American economy discovered
in
saving (including consumer durables)
that the ratio of personal
to personal
income appears to have been substantially constant, as have been proportions
of total savings made by the personal, corporate, and government sectors,
respectively (if saving ttirough social security
funds
_/
is
included
Raymond W.
in
personal
saving).
—
other governmental trust
This outcome over along period of
Goldsmith, A Study of Savings
Princeton University Press,
and
in
the United States
,
Volume
I,
1955, pages 6-9.
years encompassing significant institutional changes, particularly the rapid
development of financial
saving (for example,
intermediaries and
life
insurance and,
a
relative growth
in
"contractual"
later, amortization of mortgage
payments as general practice), suggests that there were powerful and persistent
forces making for constancy
hand" of empirical
in
savings shares.
A resort to the "invisible
regularities never proves anytliing, of course, but the
record of the past suggests that accompanying the growth of pension funds as
with the rapidly burgeoning financial
will
be adjustnients
in
institutions of early periods, there
choice and use of savings media that would keep
aggregate savings ratios approximately unchanged.
Both the deductive and the empirical argument just presented have
assumed,
Implicitly, that pension plan savings are reasonably good substitutes
for other forms of saving.
If
not, the offsets need not be substantial.
fact, there are good reasons to consider that accumulation in this
form is
In
31
quite an imperfect substitute for other household
_/
See:
Roger
Murray, Economic Aspects of Pensions:
F.
National Bureau of Economi
a)
Equity
saving.—
in
c Ffesearch,
Inc.,
New York,
1968,
A Sumn.ary Report
p.
,
55.
pension fund cannot be drawn on to meet emergency needs
a
or for purpose other than retirement income.
They lend little liquidity
to the household's asset portfolio.
b)
Pension plan participation
is
compulsory.
Some accumulations are made
on behalf of workers v/ho would not otherwise save or would not save
as much.
c)
There appears to be considerable ignorance and uncertainty among employee:
of the facts of pension plan coverage and accumulation.
that,
the deductive argument
In
addition,
in
common with most economic reasoning, consumer preferences viere assumed
to be fixed and
invariant.
But,
is
to revert- to Scotch once
people attracted by its lower price had tried
recognized that
it
filled
a
subject to the further qualification
it
more, suppose some
and grown to like
need no other liquor quite served.
it,
had
Then sales
of Scotch would increase but not primarily at the expense of other hard liquors.
These qualifications suggest that the savings effect of private pension
fund accumulation
is
a
more open question than the earlier arguments would
hold, an argument to be settled by recourse to the evidence rather than deductively.
Early evidence (from the late 50s and early 50s) garnered
Investigations indicated that personal saving
cet.
par.
in
all
in
two separate
other forms was as great
for those covered by pension plans as those not covered, pointing to
32
the conclusion that pension plan saving was a net addition to total
saving.
One study drev^ on a large but not very "representative" sample, and the other
on a smaller sample structured to be representative of the United States
population as
a
whole.
Phillip Cagan summarized his primary finding and his explanation for it,
this way:
"...covered households make no net reduction
in
all
likelihood make a net increase though
it
in
other forms of saving and
may be less on the average than
one percentage point.
"Pens ion coverage draws attention to the problems of providing for retirenen;
and goes
a
long way in helping to solve them.
"It facilitates the rapidly spreading shift to financial means of providing
for retirement from the older reliance on family,
family farm or business.
But by
itself it
the average household supplements
it
is
rental
property, and the small
apparently found inadequate;
by additional
accumulations, mostly
in
bank accounts and government bonds, at the expense of consumer durable
purchases.
I
shall
better term."
a
/
call
this a " recognition" effect of coverage for want of
—
Phillip Cagan, The Effect of Pension Plans on Aggregate Saving
Bureau of Economic Psesearch,
sample consisted
Inc.,
New York,
I965,
pp.
28 and 53-
,
National
Cagan's
of about 11,000 households who subscribed to Consumer's Union
and answered a mail questionnaire distributed to them.
2.
From another inquiry with the same purpose, but using
a
very different
sample and methodology, George Katona concluded "that coverage by private
pensions makes a difference
in
explaining saving performance, encouraging
33
more saving rather than inhibiting it."
_/
George Katona.
Center,
Pr
i
—
The explanation, as he sees
vate Pension Plans and
Individual Saving
it
Survey Research
,
Institute for Social Research, University of Michigan,
I965,
pp.
81-82.
runs as fol lows
"The prospect of receiving pr vate pens ions made many people feel
i
that
they were close enough to their goal of achieving a satisfactory income
after retirement so that with some additional saving effort they would
be able to reach that goal.
pension plans often felt that
a
the other hand,
On
it
people without private
was not possible to assure themselves
satisfactory standard of living during their old age; whatever they
could save would not suffice, so that they were not greatly motivated
Aspirations appear to rise with accomplishment, and attainable
to save.
rewards have
as well
__/
a
stimulating effect not only on spending but on saving
because both arc held to be of great value."
George Katona, Burkhard Strumpel, and Ernest Zahn, Aspi rat o ns and
i
Influence:
Comparative Studies
McGraw-Hill Book Company,
Some years
concluded that
later
"iiis
1971,
(197'0
in
the United States and \/estern Europe
p.
103.
,
Munnell analyzing the data used by Cagan,
published results seem less convincing, and the conclusion
that pensions stimulate saving more questionable."
_/
—
—
And on closer study
Alicia Haydock Munnell, The Effect of Social Security on Personal Saving
Bal linger Publisliing Company,
of a sub-sample of Cagan'
s
Cambrdige, MA,
data (and
soir.e
to his study from the same population)
197^*,
P-
98.
collected immediately subseque.Uly
she obtained "results that directly
,
3^
contradicted his original conclusion. Coverage by either
security was accompanied by
social
rate among persons aged 55"65,
Ibid.,
p.
a
3
a
pension plan or
percentage point decline
the retirement savers.
in
the savings
—
98.
Even more recently working with a new body of data (covering the saving
behavior for the years 1966~71 of about 5,000 men who were between kS and 59
in
1966), Munnell
concludes on the basis of
coverage reduces savings efforts
for whom retirement
_/
Alicia
of Pol
i
t
H.
ical
in
a
very careful analysis that "pension
other areas--at least for those older men
—
the primary savings motive."
is
Munnell, "Private Pensions and Saving:
Economy
,
October,
197^,
p.
New Evidence," Journal
lOl't.
Because the sample was representative of the country as
a
whole, Munnell
was able to estimate the aggregate effect on saving of annual provision of
private pensions.
in
_/
In
all, "....the reduction in
.
ann'jal
aggregate saving
1973 due to future private pension benefits would be about $13 billion.—
Ibid.,
p.
1031.
It
v;ould be
less than this "if younger persons are very
insensitive to pension coverage."
However contributions
(employer and employee combined)
plan funds were $21.1 billion in 1973,
so pension savings are only partially
In
1973,
for example, with
running about 60 percent of contributions,
total
personal savings
offset by displacement of savings
the offset
to private pension
in
other forms.
were higher by about $8 billion because of i.idustrial pension plans.
cites as the reason for
a
partial
Munnell
(rather than full) offset "the fact that
the substitution of pension saving for other saving is
imperfect probably
35
because of lack of vesting and other uncertainties."
/
—
If
lack of vesting
is
Ibid.
important
in
expla
i
n
i
ng the
partial offset, we can expect over time a higher
percentage tradeoff of other savings against pension fund saving, because
vesting conditions have improved substantially
in
recent years as required
by ERISA.
Substantially, the same story on savings effects should apply for those
employees covered by pension plans of state and local governments, which
characteristically have had more generous vesting provisions than private
industrial plans.
Munnell concludes, "On the basis of these rough calculations, private
pension plans appear to increase aggregate savings on net and contribute
to capital
_/
accumulation."
—
Ibid.
The effect of OAS
I
on saving appears
to be quite different.
Essentially
a pay-as-you-go operation, with a pension promise validated by the full
faith
and credit and the taxing power of the federal government, the social security
system provides promises which those presently working can "count on," but
does not build up a fund.
effect of OAS
I
on saving and reached the conclusion from analyzing a variety
of evidence that the social
on saving.
Both Feldstein and Munnell have researched the
security system exercises
a
negative effect
36
Feldstein,
too,
is
currently studying the effect of private pension
The most recent paper of his that
plans on saving.
I
have seen on the
subject notes first in its conclusions that:
"The analysis and statistical estimates presented
an
important difference
the aggregate economic
in
in
impact of private pension
programs and unfunded public social security programs.
social
this paper point to
Although such public
security programs are likely to reduce national savings by acting
as a substitute for household retirement saving,
this tendency is offset
funding
private pension programs by the combination of the companies' partial
response to unfunded liabilities."
and the shareholders'
Martin Feldstein, "Do Private Pensions
_/
1957, Harvard
Discussion Paper 553, May,
May,
1977,
much more study
Increase National Saving?",
Institute of Economic Research,
is
is,
in
needed before solid estimates can be made.
an important question"
is
,
pp.
33 and
that
His own
first attempt to answer
a
"that the $11.5 billion growth of pension reserves
197^ added $^.0 billion to private saving...."
Ibid.
He emphasizes
theory, unambiguous.
preliminary estimate, which he describes "as only
_/
He goes on, however,
33.
p.
to note that neither effect
in
—
in
—
32.
With respect to saving,
then,
there is
a
real
complementarity between
the two major systems that have emerged in the last 30 years for income
support in retirement.
there
is
a
lower than
(OASI) und "private" plans
(in
which
public interest) dovetail neatly-the one tending to make saving
it
aggregate.
otherwise would be, the other tending to raise saving
a
in
the
lesser "distortion"
Reliance on both systems,
in
effect, means
than would occur
if
we relied on either one alone, and
of the savings level
argues for
The "public" plan
continuance of
a
a
major role for the combination public and
private system for income sufiport
in
retirement.
,
37
Additional Saving for How Long ?
During working life, then,
plans tend to save more,
in
appears that those covered by pension
the aggregate
partially offset by less saving
support consumption
life will
it
(because pension plan saving
other forms).
in
retirement.
in
plan fiscal operations on saving
in
But their saving
Hence,
is
only
working
in
the net effect of pension
the aggregate depends on the net balance
of contributions plus fund earnings on the one hand and benefit payments on
Reverting to the bathtub analogy, at the outset of
the other.
a
plan, with many
workers and few retirees, contributions geared to expected future benefit
payments run at much higher levels than current benefits.
the bathtup more rapidly than
drains out.
it
each year as the fund accumulates assets.
added to
and
vs/ater
ttie
Aid
V/ater
flows
into
"The "dirt rings" will be higher
the earnings on those reserves are
But over time the number of retirees picks up momentum
inflow.
starts to flow out at
more rapid rate.
a
This tends to tone down on
annual accumulation.
If
we were dealing with
a
steady-state population of workers and retirees,
the fund would level off at the point where contributions plus earnings just
equalled benefits.
As
long,
however, as the working population grew and/or
plan provisions were liberalized, contributions plus earnings would tend to
exceed benefits.
Casual
Pension funds would be net accumulators.
inspection of Table
1
and
2
indicates that while benefits are
catching up on contributions plus earnings there
is
some way to go.
Private
industrial plans are still accumulating assets, and so are state-local
government employee plans.
Indeed,
the asset accumulation propensities of
the latter are impressively large given that they cover a much smaller
38
population than industrial plans,
are notoriously underfunded.
a/id
accumulation continue, i.e.
For how long will
for how long can we
expect private pension plans to continue to add to aggregate saving on net
balance?
A dozen years ago
than, as
/
turns out,
it
Daniel
made
correct.
stab at this question which was more courageous
a
—
Projecting for the period 1965 through 1975,
Holland, Private Pension Funds:
M.
of Economic Research,
for example,
I
I
inc..
New York,
Projected Growth
National
Bureau
19^6
seriously underestimated the level of annual
and prematurely projected a tendency for the annual
I
,
fund accumulation,
accumulation to level off.
did capture the surprising vigor of state-local plan fiscal operations
relative to private industrial plans, however.
Nothing
is
from my guesses
in
lost,
and something
is
likely to be gained,
if
we turn
to more recent conjectures and estimates.
discussing the future pattern of private industrial pension fund
accumulations William Hsiao identified two Important forces that would tend
to keep contributions plus earnings
kO years,
"at
tiie
larger than benefit payments over the next
end of which time, pension plan funds migiit well become net
sellers of financial assets.
1.
EPxISA starting
In
197^,
This will cause a "....bulge
requires funding of past service credits.
in'
funding patterns
[that] may continue for
the next 20-30 years."
2.
He also expects that when the baby-boom of the Post-War period fully
impacts the labor force between
contributions will
1975 and
1985,
covered workers and
increase substantially, and fund assets will continue
39 6 'O
to grow.
_/
in
—/
William C.L. Hsiao, "Demographic Changes and Funding for Pension Plans,"
Funding Pensions:
Series No.
16,
Reserve Bank of Boston
Federal
Held in October,
Implications for Financial Markets
issues and
1976)
,
(Proceedings of
a
,
Conference
Conference
2k.
p.
"Yet as the age cohort groups born between 1950-60 reach retirement
year 2015 and after, the pension funds
benefits.
pay out the accumulated funds as
Meanwhile, with the expectation that the lower fertility rate
we have experienced will continue,
the proportion of active workers will
Accordingly the aggregate contributions are likely to decrease.
decline.
It
v;ill
in
seems highly probable therefore, that the balance of pension funds will
be depressed because
the net cash
f
low--cont
ri
but ons minus benefit payments-i
may be negative."
_/
Ibid.
As a comparison of Tables
and
1
suggests, state-local employee plans
2
are accumulating reserves at a faster rate than private industrial
And this despite the fact that they tend to fund
moreover Munnel 1-Connol
ly
'
s
and would extend over
Alicia
Local,
it),
To the extent that their degree of
stepped up, perhaps because of the danger that they might otherwise
be put under ERISA or a similar
_/
(or lack of
projections suggest very impressive annual additions
to fund assets over the next 20 years.
is
lower proportion of their
Even at their current rate of funding
pension liabijities.
funding
a
plans.
H.
Civil
for Financial
Munnel
a
M.
Connolly, "funding Government Pensions:
Service and Military,"
Markets
,
the accumulations would be heavier,
—
longer period
and Ann
1
regimen,
in
Conference Series No.
(Proceedings of a Conference Held
in
Issues and
Funding Pensions:
October,
16,
Federal
1976)
pp.
State-
Implications
Reserve Bank of Boston
9^*
^nd 117-
41
Meeting Pension Promises
receive a defined benefit
Most workers covered by Pension plans will
which
is
the pension under this plan shall
the following:
annual compensation in the
retirement
2/3 percent of average
1
highest consecutive years of the 10 preceding
5
had been under the plan for 30 years would,
to 50 percent of his
equal
be
Thus a worker who
years of service (not exceeding 30 years).
X
for example is
A characteristic formula,
related to final pay.
(a
retirement, receive a benefit
pay.
final
The record of the past
in
long extensive record)
suggests that real
say,
wages tend to grow at about the same rate as productivity,
Nominal wages then tend to grow at
3.5 percent.
rate of inflation.
the actual
If
effective and more pronounced
in
thei
r
work
1
i
to
to 3-5 percent plus
3
anything, de facto indexing has become more
recent years, and therefore tins relation-
ship of the past can be expected to persist over the future.
therefore characteristically have
3
a
Workers
pension claim indcKed to wages over
fe.
As a hedge to meet workers'
pension claims, pension funds cannot find
precisely similar market investment instrument.
Were they to invest
expect, over
solely in short-term debt, say Treasury Bills, they could
long periods a return equal
a
real
to about the average rate of
inflation,
i.e.
variability
return of close to zero over the long pull, with some
around the average, of course.
term expectation
is
a
real
long-term corporate bonds the long-
On
return of about 1.5 percent, with greater
variability around the expectation.
over the long pull
percent real
(on
On common stock,
they could expect
the basis of the record of the past)
about 6
greater annual
return, but, of course, with a substantially
variability than on bonds.—
/
for Treasury Bilk, Corporate
The annual standard deviations of return
.
-
,.
.
__.-
_.:..„,,
L
A-?'
7
7;.
nnd 23.5'i.
The
i!<ila
a
1^2
on historically-experienced returns and their variability are fron:
Roger
Ibbotson and Rex A.
G.
Year-by-Year Historical
Inflation:
Business
1976,
Returns (1926-7^)," Journal of
11-^47.
pp.
pension plans have had
the past, as Treynor notes,—
In
/
January,
,
Sinquefield, "Stocks, Bonds, Bills, and
Jack
L.
a
"something
Treynor, "The Principles of Corporate Pension Finance," The
Journal of Finance, May,
1977,
for nothing" aspect about
P-
them,
627.
the sense that they represented
in
claims that potential beneficiaries took seriously, but corporate managers,
actuaries,
accountants, creditors and stockholders "have regarded
pension funds lightly."
More recently, however, they have and will
continue to be taken more seriously on the corporate side.
One reason for this lies in the poor stock market results of recent
years.
The precipitous decline
for many corporations
a
stock value destroyed
in
v/hat
had been
comfortable cover for pension plan obligations.
Another was the introduction of more stringent funding requirements,
regulations and potential sanctions under ERISA.
Legally, at least,
up to 30 percent of corporate net worth is attachable to meet pension
plan obligations should accumulated funds fail
when
a
good number of firms were
in
to be sufficient.
this condition,
the PBGC
(Whether,
(Pension
Benefit Guaranty Corporation), established by ERISA as the insurer" of
plan obligations, would choose to exercise this power,
is,
of course,
highly debatable.)
And there are intimations that
investors have now become keenly
aware of the obi iyat ions that pension plan promises represent.
Oldficld found that "Although
not directly reported
in
a
Thus
[unfunded pension] obligations are
firm's balance sheet,
thoy influence the
A3
value of a firm's common stock."
_/
George
/
Oldfield, "Financial Aspects of the Private Pension System,"
S.
Journal of Money, Credit and Banking
degree of influence
is
debatable.
,
February,
1S77 Part 1, p.
The
S**.
Oldfield concluded that for each dollar
of unfunded vested liability the share price was reduced by about one dollar.
Since aggregate unfunded liabilities exceed unfunded vested liabilities,
Feldstein (op.
cit., p.
23)
concludes that "this implies that share prices
are depressed by substantially less than the total
And he adduces additional evidence to support the same
pension benefits."
belief, viz.
the failure of corporations to float debt to fund pension
obligations even though there would be
(op.
cit.
p.
unfunded expected future
a
tax advantage
in
so doing because
"the interest cost incurred by the firm on its new debt
23)
would be tax deductible while the
nev^
interest income received by the pension
fund v;ould not be taxes" has as "one plausible explanation" the possibility
that "the market would recognize the explicit debt
accordingly)
in
recognized."
(p.
a
(and depress share prices
way that the implicit "debt" to the pension fund
2^4)
Is
not
^li
In
Bull
the pre-ERISA,
market years, a heavy reliance by pension
funds on stock which could be expected to generate on average over the
long pull
to meet
the
real
returns sufficient
pension plan obligations was the "natural" way to go.
real
regulatory environment and market "tone" suggests
market tone
_/
record of the past persisted)
the
(if
is
a
transient phenomenon,
—
it
is
Now both
The negative
not.
but the regulatory environment
is
Waiving aside the question, addressed below, as to whether the
generosity of the pension plan promise and the increased certainty
of its excercise have brought about a permanent decline in profit
prospects, and, therefore,
in
the price level of common stocks.
here to stay.
Literally and narrov^ly interpreted the regulatory requirements of
ERISA do not seem particularly irksome.
standards of prudent
i
al
i
They substitute for the older
that a more sophisticated "rule" that
ty
requires
diversification to avoid undue risk. These and other features of ERISA
some commentators believe have tended to push pension fund management
toivard more
_/
Cf.
conservative investment.
Randall
D.
Issues and
Conference held
in
But on an
Weiss, "Private Pensions:
Growth of Retirement Funds,"
Pen sions:
~
Impl
i
in
Federal
firm basis
Impact of ERISA on the
Reserve Bank of Boston, Fund ing
cat ons for Capital
i
The
individual
Markets
,
Proceedings of a
October, 1976.
Weiss concluded that "The fiduciary responsibility rules will
increase
iA
investments and thus lower their overall
the conservatism of plan
of return."
(August ^,
(p.
And he cites
I'lS).
1976) that
in
"indicates that overall
(p.
a
survey of pension fund managers
investment strategy has become more conservative."
\h8).
the push
the
reporting on
Street Journal article
Wall
a
rate
in
this direction would not be as severe as
it
would be if PGBC took
aggregate pension fund structure as the relevant domain for standards
of prudent
i
al
i
ty
An individual
firm or two that sought a high return/risk position
for its pension fund position could be handled by PGBC.
too heavy
an equity position,
occurrence of an event
it
If all
firms took
however, PGBC would be faced with the possible
could not insure against.
So, depending on how ERISA and PGBC
requirements are interpreted
have one of several emerging possibilities.
the pension promise
If
vie
is
not adjusted downward,
1.
Firmscould impose
a
more conservative funding regimen on their
pension funds than they have
in
the past.
This means that the
old level of "foregone cash wage payments" that employer contri-
butions to the pension fund represent will not be sufficiently
large, given the lower rate of fund earnings,
to aggregate to
the promised pension. Either workers will have to "save" more
(or,
as they would see
it,-
take a first derivative cut in cash
wage increases); or companies will hove to contribute additionally
to pension funds out of profits,
of shareholders.
/
and,
therefore, at the expense
—
Whether this implies further deterioration
in
equity prices depends
k5
on whether we think that recognition of this contingency
already depressing stocPc prices.
effect, "taken their lumps"oi
hovN/ever,
2.
a
If
is,
it
a
factor
then, equities have,
once and for all basis.
further cause for (historically)
a
is
If
In
not,
is
it
low stock prices exists.
Again with the pension promises unchanged, firms could seek
higher return for their pension funds, with
of stock,
a
heavy loading
thus creating the possibility for a substantial
number
of plans that at some time stock values could be so low that
current
flov-vs
could not meet promised benefits and pension fund
resources would have to be called on.—
_/
Given the Ponzi -game-
force
is
1
But while PBGC appears
ike nature of pension plans as
growing relative to retirees
(or more
long as the v/ork
strictly as long as con-
tributions on behalf of work force plus earnings are greater than benefit
payments) there could be
a
sufficient cash flow to mset current obli-
gations to retirees, but at the expense of funding future obligations
incurred on behalf of current workers.
generational conflict.
50 years
from now,
Thus the prospect of inter-
And this becomes more severe if, as may well happen
the working force growth rate
declines.
td have the power to take up to 30 percent of net worth of
firms
In
this position to meet
be folly to exercise
this power at such a time.
back up pension plan promises?
last resort,
their pension promises,
One possibility
they would be made good by the full
and credit of the Federal Government.
Then what would
is
that,
this a likely story?
is
tiie
Private pensioners,
taxpayer.
No one knows, of course, but It
clear that the contingency
in
faith
the retired elite, would be supported by the general
Is
would
it
is
fraught with the possibility of deep
^40-
H
community
3.
di vis
I
veness.
A third possibility
is
that
full
recognition of the range of
responsibilities assured by ERISA and PGBC, involves these
entities with
a
concern not only with the funding process, but
with the nature and size of the pension promise.
present
graded.
levels,
Compared to
future pension promises would have to be down-
This would
involve conflict between younger workers, who
would bear the brunt of this adjustment, and older v/orkers and
retirees who had more generous promises.
k.
The possibilities discussed to this point all
way or other
in
represent some
which workers v/ould shift the risk of their
pensions as presently designed to some other group.
there
is
another possibility.
But
Pension design could be changed.
We could switch from defined benefit plans to defined contribu-
tion plans, v;ith the pension being the annuity that can be
purchased with the sum accumulated for each worker.
Company
obligation and risk would be limited to the contribution promised
by the plan formula.
Workers would bear the risks associated
with their pension plan portfolios.
TABLE
1
GROWTH OF PRIVATE INDUSTRIAL PENSION
AND DEFERRED PROFIT-SHARING PLANS IN
THE UNITED STATES, 19^0-1975
19^0
]3h5
1950
1
955
19^0
1965
1970
1975
^.1
G.k
9.8
1
^f
2
18.7
21.8
26.3
30.3
Civilian Employees (Excluding
Government and Agriculture)
(Millions)
28.2
3'^.^
39-2
'(3.8
'iS.g
50.7
58. A
62.3
Percentdgc Covered
]h.5
Coverage (Kill ions)
Contributions
($
billions)
Contributions per Covered
Worker (S)
Beneficiaries
(Millions)
Benefit Payments
($
Billions)
Per Beneficiary
(Current
Per Beneficiary
(Constant
Assets
($
$)
a
$)
Bill ions)
Assest per Covered Worker ($)
.
TABLE
2
GROWTH OF STATE AND LOCAL GOVERNMENT
EKPLOYt:E PENSION PLAND AND FUNDS,
19'<0-75
19^0
[950
I960
1970
1975
Coverage (Millions)
l.^
2.6
h.S
7.3
9.5
Employees of State and
Local Governments
(Millions)
3.3
0.59
1.17
1-59
9'<0
2,905
6,300
Percentage Covered
Contributions
($
Billions)
Per Covered Employee
($)
Retired Beneficiaries
(Millions)
0.13
0.25
Benefit Payments
($ Millions)
113
27^
869
1,096
1,593
2,^)83
3,962
2,069
1,520
1,796
2,135
2,^68
Per Retiree
(Current
Per Retiree (Constant
$)
$)
12
15
TABLE
^
Prevalence of Expenditm^es for Retirement Plans
in the Nonfarm Economy, by Selected Characteristics,
1968
(in percent)
16
TABLE
5
Prevalence of Expenditures for Retirement Plans
in the Private Nonfarm Economy and Industry Di^dsion,
by Selected Characterictics, 1968
(in percent)
Tola! privals
nonfarm
Mjnuf jctur
Ing
All nonmanufacturing
Mining
Construction
TrsnEportatlon,
F Iri.Tnce,
Trade and
communication
and ullllllea
Insurance, and
rcdl estate
service
El-
Noei-
E>-
Ei-
Noei-
fi-
Er-
Ho ea-
Ei-
No oa-
pend-
fend-
pend-
fiend-
pend-
tips nd-
No rr-
ftend-
prnd-
pcnd-
fi^nd-
ppnd-
pend-
pend-
pr:iid-
pend-
nend*
turea
Ituree
lures
Ituree
lurea
Ituree
lluret
Ituret
ituree
tures
Iturel
tturea
Itures
ituret
Iturot
Iturea
Ci-
No
ei-
\
No
ei-
No
ei-
I
Type of
employee
employees
All
NonofTice
Olfice
55
50
67
45
50
33
73
70
27
30
S3
17
80
49
2S
19
;3
59
86
93
47
53
40
62
f.O
18
20
38
52
54
28
48
46
77
74
39
38
(2
19
70
81
31
84
23
26
16
61
72
30
56
69
44
Avertje Hourly
Compensetlon
lesslhiin{;sa
...
n
J?50lo J349
$3 50 10 1199
51
74
is end over.
tl
72
18
41
47
66
72
14
7
82
53
(')
('1
34
28
(5
83
34
33
85
98
21
63
15
17
13
8
87
92
16
45
84
54
23
77
18
4 'J
51
82
52
49
5!
81
19
21
59
41
71
29
97
3
79
75
25
61
39
20
4
70
24
47
59
56
41
53
13
67
61
33
39
70
44
34
30
66
79
71
15
20
52
80
48
2
9i
Av6r*ge Annual
Cirnlnci
Undei J5 000
JSCM Sid undet
JO
»IO,000
(10,000 and o«ei...
74
36
^6
29
ft
16
90
10
66
67
56
S4
62
16
38
79
37
73
26
60
93
71
59
29
57
70
43
30
fO
9«
80
73
15
20
85
87
56
42
72
58
28
44
75
92
J5
15
2
44
employee
Organlzillon
Umon
18
82
44
Nonunion
Number
of
Cmploycet
in
Cttabllshmenti
Under 100
100 lo *j9
500 and over
*
?7
_.
S3
32
38
M
68
36
7
93
7
•
74
40
71
7
98
^
12
56
25
9f
Insufficient data to warrant publication.
Source:
Bnerson Bier, "Incidence of Private Retirement
Plans", Monthly Labor Review July 1971, p. 38.
.
2^
TABLE
6
REPLACEMENT RATE OF SOCIAL SECURITY ALONE AND
SOCIAL SECURITY AMD PRIVATE PENSI0M3 COMBUJED,
FOR AVERAGE WORKER IN MANUFACTURING, SIX COUNTRIES, 1975
Pension as % of Earninqs
in
Year Prior to Ret
i
reiT'snt
27
TABLE
7
AVERAGE EMPLOYER PENSION PLAN COSTS
AS A PERCENT OF PAYROLL, 1953-75
Year
1953
1955
1957
1959
IS61
1963
1965
1967
1969
1971
1973
1975
Al
1
Indust ries
TABLE 8
NET PURCHASES OF CORPORATE AND FOREIGN BONDS
BY STATE AND LOCAL GOVERNMENT
RETIREMENT FUNDS AND PRIVATE PENSION FUNDS, 1966-75
(1)
Net Purchases
by State
and Local Funds
(2)
(3)
Net Purchases
by Private
Pension Funds
Total
=
1+?.
(M
(5)
Total Net
Total Punch;
Bond Issues of Pension "
as Percent
Net Is:
- Zlk
Total
Year
1966
2.9
2.5
5.^
11.2
kZ
67
3.7
1.1
't.B
16.6
29
68
2.6
0.6
3.2
1^.'4
22
69
^.0
0.6
h^
13.8
33
70
^.5
2.1
6.6
23.3
28
71
3.9
-0.7
3.2
23.5
\^
72
ii.S
-0.8
3.7
18.
if
20
73
6.0
2.1
8.1
13.6
60
Ik
6.^
^.7
11.1
23.9
A6
75
5.0
2.8
7.8
36.3
21
Source:
Flow of Funds Accounts
TABLE 9
NET PURCHASES OF CORPORATE EQUITIES BY STATE AND LOCAL GOVERNMENT
RETIREMENT FUNDS AND PRIVATE PENSION FUNDS, 1966-75
(1)
k
(2)
(3)
W
(5)
TABLE 10
ASSET HOLDINGS OF STATE AND LOCAL GOVERNMENT
EMPLOYEE RETIREMENT FUNDS, 1950-75, SELECTED YEARS
Un
Bi 11
ions of Del lars)
I960
1950
^ of
% of
Asset
Demand Deposits +
Currency
Amounts
Total
1970
Amounts
Total
1975
%
of
Amounts Total
^ of
Amounts
Total
119
2.5
?-^3
1-2
601
1.0
1,700
1.6
29
0.6
600
3-0
10,100
16.7
25,800
l^^.h
Government
Securities
2,')98
51.^
5,907
29-9
6,595
10.9
6,812
S.k
State C Local
Obligations
1,5^9
31.9
^,^06
11. 1
2,032
}>.k
2,500
l.k
585
12.0
7. 12^4
36.1
35,055
58.1
60,889
57-5
75
1-5
1,^50
7-3
5,920
9-8
8,250
7-8
Corporate Equities
U.S.
Corporate Bonds
Mortgages
Total
Financial
i»,855
Assets
Source:
19,730
60,303
105,95!
Board of Governors of the Federal Reserve System, Flow of Funds
Accounts, 19'i6-1975
TABLE
11
SHARE OF PRIVATE PENSION FUNDS AND STATE AND LOCAL GOVERNMENT
EMPLOYEE RETIREMENT FUNDS IN TOTAL OF
BONDS AND CORPORATE EQUITY OUTSTANDING,
SELECTED DATES, 1950-75
1$ Amounts in Billions)
950
Asset
2.8
I960
1970
1975
TABLE
12
HOLDINGS OF PRIVATE INDUSTRIAL PENSION FUNDS
1950-75, Selected Years
(In Bill ions of Do] lars)
i960
1950
Asset
Demand Deposits +
Time Depos ts +
Currency
i
Amounts
^ of
Total
% of
Amounts Total
1970
1975
of
% of
Amounts Total
Amounts Total
?;
JY
2
"87
DC 2 S '89
MAR
3 1 1990
SEP 27
1991
ACME
BOOKBIfDiNG
SEP 6
CO., INC.
1983
100 CAMBRIDGE STREET
CHARLESTOWN, MASS.
mil
3
Hill
lull
I
III
III
III
I
Hill
TDflD ODM
I
I
I
II
III
nil
S2M
II
Mfll
Ml' .IBftflRlES
)0V-7?
3
TDfiD DOM
52M SD7
H028.IV1414 no.l032- 78
Kalwani, Manoh/Structure of
73626J.
D*.BKS
.
repeat buy
Q,(JP6,95U
TOaO DOl EMD bSl
3
HD28-M414
Kalwani.
no. 103278
Manoh/Structure
737571
D»BKS
|i!ni lip |!!i
III
I ill
II
11
1
I:
TDflQ
3
II
1979
repeat buy
of
00137169
flip
'hfmk
M
IILJiiil
002 OSD
Olfl
HD28.WI414 no.l033-
78
Baldwin. Carli/Liquidity preference un
73601.1^
D*BKS
_
00OS6623
Plipi
TOflO
3
001 20T
HD28.IVI414 no 1034- 78
Van Breda, Mic/Bayes vs.
736013
.
TOfiO
3
the
D»BKS.
_
3fl3
lens
0006.6.6.2
001 20T M33
H028.M414 no.l035- 78
§l!!!?,",v
of
^^'"'^^""^'^y^^'^'cs
736015
D»BKS
price
ioao 001 20^
3
elast
QlineRB."
MSfl
no.1036- 78
Schmalensee. R/A simple model of
HD28.I\/I414
n»BKS
736017
llllllll
llllilllll
TOflO
3
001 20T
HD28.M414 no.1038Choflray,
73601?
m
79
Jean/Methodology for seqment
.D»BKJ.. 1)0066621
.
iillliilili
llllllll
TOaO 001 20
3
risk
00066625
3ME
HD28.M414 no 1039- 79
^.r'5™n. Jeffr/Decision support system
827646
3
0«BKS
M.
TOflO
L
00069325
001 2M
fl?S
HD28.M414 no.1040- 79
Kaplan. Nancy /A computer code for
73602.1
D»BKS
00066622
3
TOaO 001 2DT
sol
3t,?