View original document

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

MONETARY POLICY--THE POSSIBLE AND THE IMPOSSIBLE
Address
ROBERT
President,

Federal

by

P. BLACK

Reserve Bank oj

Richmond

to the
West Virginia

Bankers Association

Eighty-Eighth

Annual Convention

July 24, 1981

It is a pleasure and an honor to participate in your
program today.
In running
over in my mind the
many possible topics I might try to tackle, I decided
that perhaps I should say a few things about Federal
Reserve monetary policy.
Fed policy, of course, is
much in the news these days. But aside from whatever immediate impact the System’s actions may be
having on conditions in the financial markets at the
present moment, it seems to me that this is a good
time to talk about monetary policy for a couple of
reasons.
First,

our

quandary

country

has been

in a real

economic

for some time now. To cite just a few bits

of evidence,

real GNP

grew at a 4.2 percent

annual

rate in the 1960s, but it grew at only a 3.1 percent
rate in the 1970s.
Employment,
in contrast, grew
more rapidly in the 1970s than it did in the 1960s,
which

means

that

we experienced

some decline

in

productivity
in the 1970s. On top of this, economic
activity fluctuated more widely in the 1970s than in
the 1950s and the 1960s.

We suffered

three

reces-

sions in the 1970s, and the one that began in 1973
and ended in 1975 was the most severe recession
since the 1930s. But the most pervasive and in my
view the most dangerous problem we face currently
is inflation.
In the 1960s prices rose roughly 25
percent; in the 1970s prices rose almost 100 percent.
Moreover,
as you are all well aware, inflationary
expectations have risen rapidly along with actual inflation.
These expectations,
in turn, have been incorporated
into long-term
interest rates and have
helped propel these rates to unprecedented
heights.
Many economists
believe that monetary
policy can
help the nation achieve a better economic performance in the 1980s.
2

ECONOMIC

REVIEW,

The second reason that this is a good time to talk
about monetary policy is that the Administration
is
trying to engineer a substantial
shift in both emphasis and direction in the formulation
of national economic policy.
This shift has important
implications
for monetary policy.
Accordingly,
I want to make four main points this
morning, and I’ll flag them so that you won’t miss
them.
First, I’ll touch lightly on what might be
called the traditional post-war view of how monetary
policy works.
Second, I’ll describe how I think this
view is changing.
Third, I’ll share with you my
personal view regarding the proper role of monetary
policy, and, finally, I’ll give you my assessment
of
the chances that the policies we are now following
will contribute
to an actual improvement
in the
nation’s economic performance.
Let me move then to my first point:
the traditional post-war view of monetary policy. As you all
know, there are several competing views about how
the Federal Reserve should conduct monetary policy.
I think it is fair to say that over the post-war period
as a whole, the majority view among those with an
interest in public policy has been that the Fed should
conduct monetary policy by trying to manage interest
rates.
Some people have even argued that the Fed
should try to conduct monetary policy with the express purpose of keeping interest rates low in order
to stimulate
borrowing,
investment,
aggregate
demand and economic growth.
Others have held the
somewhat
more moderate
view that the System
should adjust interest rates up or down as necessary
to smooth out swings in the business cycle. In any
case, this traditional
view of monetary
policy has
clearly been reflected in the actual conduct of policy
over most of the last 25 years.
SEPTEMBER/OCTOBER

1981

My second point has to do with recent changes
attitudes

toward

creasingly

monetary

recognized

mists and by the general

policy.

both

to anticipate

in

I think

by professional

ations

econoRe-

and

short-run

activity
with

growth

it is in-

public that the Federal

relatively

nomic

offset

of the aggregates.

discretionary

serve cannot

that no one can forecast

period

confidence,

control interest rates for any extended
with
of time, especially in an environment

in eco-

anticipated

fluctu-

changes

in

I’m increasingly
the business
think

the -rate

of

convinced

cycle with much

anyone

understands

Finally,
with my preceding
remarks
as background, let me say a little about present and prospective policy.
I’ll divide this part of my discussion
into two subparts.
First, why have we at the Fed
sometimes failed to achieve our policy objectives in
recent years?
Second, what efforts are we making
to improve
the techniques
we use in conducting
policy and what results can be expected from these
efforts ?

Now for my third point. If the Fed cannot effectively manage interest rates with monetary
policy,
what then can it do ? The answer to this question
should surprise
no one:
We can and we should
control the rate of growth in the money supply.
I
reach this conclusion by a fairly simple route.
Most
of the historical evidence economists have developed
suggests that our nation’s real output can grow by
about 3 percent or so per year. It seems intuitively
clear that the means of financing this physical growth
should grow at about the same rate if we are to
avoid the development
of inflationary
pressures
or
recessions.
This logic leads me to conclude that the
money supply should increase at about 3 percent or
so per year over the long run, with an adjustment
to
reflect the trend rate of growth in the velocity of
money.

Why has our performance
at the Fed fallen short
of what we wanted to achieve?
I think there are
First, it seems to me that some
several reasons.
economists
and some policymakers
have been excessively pessimistic regarding the relative costs and
benefits of reducing the long-run
rate of growth in
the money supply.
There seems to be a relatively
widespread
belief in some circles that one cannot
affect the rate of inflation
significantly
in a short
period of time without inducing a severe recession.
This view can be summed up in the rule of thumb,
derived from conventional
econometric
models, that
one must give up 10 percent of the potential growth
in gross national product in any year in order to
achieve a 1 percent reduction in the rate of inflation.
Two of our economists at the Richmond
Fed, Roy
Webb and Tom Humphrey,
applied that rule to the
German
hyperinflation
of the early 1920s.
They
estimated that, according to this rule, it would have
taken a 50 percent GNP gap maintained
over 600
centuries to eliminate the 300,000 percent inflation
rate witnessed in Germany from mid-1922 through

As you know, the Federal Reserve is now committed to controlling the money supply. The System
annually sets targets for the growth of the monetary
aggregates in the year ahead in accordance with the
Full Employment
and Balanced Growth Act of 1978
-the
so-called Humphrey-Hawkins
Act. The main
dispute in this area these days is over the extent to
which the Fed, in setting these objectives, should try
RESERVE

I don’t

fluctuations

very well the short-run
impact of the growth of the
Therefore,
I
money supply on economic activity.
favor reasonably steady, nondiscretionary
growth in
the money supply in both the short run and the long
run with appropriate adjustments
for long-run trends
in the velocity of money.
Under such an approach
money would serve as an automatic economic stabilizer, exerting a brake on inflationary
pressures and
serving as a cushion against recession.
My own
feeling is that what we now call M-1B (currency,
coin, and transactions
balances held by the public) is
the best of the existing concepts of the money supply
to use for this purpose, but the choice of a particular
monetary aggregate is a secondary matter. The main
thing is to choose one definition of the money supply
and then stick to it.

deeply embedded inflationary
expectations
such as
If the Fed tries to control interest
we face now.
rates, it inevitably
winds up trying to move them
away from the levels determined
by the natural
forces of demand and supply in the financial markets.
Attempts
to do this, however, set off a chain of
reactive forces in the markets that
eventually
drive
rates back towards their original levels and beyond
them. To illustrate,
if the Fed tries to push rates
below market levels, it has to supply additional
reserves to the banking system. These added reserves
Increased
cause the money supply to grow faster.
money supply growth, however, raises both actual
and anticipated
inflation,
and this increase in expected inflation
puts upward pressure
on interest
rates as both borrowers and lenders build this revised
As a result, rates
expectation
into nominal rates.
eventually
move back toward their original
levels
and then beyond them.
In short, efforts to manage
interest rates are counterproductive.
It seems to me
that recent monetary
history provides
compelling
evidence that this is the way the world really works.

FEDERAL

and

these

BANK

OF

RICHMOND

3

late 1923. Actually, as you know, the German inflation was virtually eliminated in early 1924 at an estimated loss of only 10 percent of potential GNP. Now
this example is probably not entirely fair since by
early 1924 the German public knew that the German
government
was serious about dealing with inflation,
whereas the American public was never completely
confident that policy was on an unambiguously
antiinflationary
course during the period on which many
of our econometric
models are based.
But the example does underline the truth that a serious commitment to monetary control can help reduce inflation
in a short period of time without imposing unacceptable social costs in terms of lost output. Incidentally,
I think that the precise speed with which we reduce
the rate of expansion
in the money supply is less
important than removing all doubts about our ability
to hit our monetary targets and our firm intention
to do so.
A second and perhaps more important
I believe we have not achieved better

reason why
results with

monetary
policy in recent years is what might be
called bad engineering.
There is absolutely no doubt
in my mind that the System’s policy objectives have
been reasonable
and attainable.
Our execution
of
policy, however, has not been as effective as we in
the Fed would have liked it to be. In particular, we
have been hampered by some institutional
flaws in
the apparatus that we use to control the growth of
the money supply.
In addition, I think that some of
our procedures have had some shortcomings.
Let me explain what I mean by engineering
flaws.
As many of you undoubtedly
know, prior to October 6, 1979, the Fed tried to govern the rate of
growth of the money supply by controlling
the Federal funds rate.
Specifically, we tried to determine
the level of the Federal funds rate that was consistent
with our objectives for the growth of the monetary
aggregates,
and then we supplied the quantity
of
reserves necessary to hit that level of the funds rate.
But we were fooled repeatedly
because we simply
did not have the empirical information
we needed to
select a pattern of Federal funds rates that would
produce the desired growth in the money supply
consistently
over time.
Since that date we have been trying to control the
money supply by controlling the supply of reservesspecifically,
the supply of nonborrowed
reserves.
This procedural
change was important
and was, in
my judgment, definitely a step in the right direction.
As I indicated a moment ago, however, our execution hasn’t always been as effective as might be
hoped.
I would make several observations
in this
4

ECONOMIC

REVIEW,

regard.
First, the new procedure implied that we
would have to allow more short-run movement in the
Federal funds rate than in the past.
Although we
have certainly let the rate move more freely than in
years past, it seems clear in retrospect that we have
not allowed it to vary as flexibly as required to hold
the growth
of the money supply under control.
Second, I do not think we have always adjusted our
nonborrowed
reserves targets as quickly and strongly
as we should have. Third, the discount rate weapon
has not been used as aggressively
as it might have
been to supplement
the other tools for controlling
money growth.
Fourth, at times we have tendedoften mistakenly
as it has turned out-to
assume
that short-run movements of the money supply away
from their target paths would be self-reversing.
Last year, for example, we did not react very quickly
to the weakening in the growth of M-1B in March,
and we certainly
did not respond nearly strongly
enough to the upsurge of growth that began in June.
Finally,
creates

our system of lagged reserve accounting
technical difficulties under our new control

procedures
that were not present when we were
trying to control the money supply using the Federal
funds

rate as the operating

instrument.

Fortunately,
there is a definite realization
within
the System that these engineering
problems exist,
and we are taking actions to correct them. In recent
months we have loosened the constraint
on movements in the Federal funds rate, and at some point
in the future I hope that this constraint will be eliminated altogether.
Further, we are currently adjusting
our nonborrowed
reserve instrument
much more
rapidly than earlier, and there is also a possibility
that the discount rate will be used more actively in
the future as a tool for monetary control.
Finally, I
think there is a good chance that we may soon move
over to some form of contemporaneous
reserve accounting.
I recognize that there are some disadvantages to contemporaneous
accounting both from your
standpoint
in the banking
industry
and from our
standpoint
operationally
at the Fed. Nonetheless,
it
seems increasingly
clear that some shift back toward
contemporaneous
accounting-with
some modifications to the old pre-1968 system to make it more
palatable
to you-would
reinforce
our efforts to
control monetary growth more effectively.
Taking all of these considerations
am rather optimistic.
I don’t want
the conduct of monetary policy has
far in 1981, but on balance I think
improvement
over events in 1980.
watching or participating
in Federal
SEPTEMBER/OCTOBER

1981

into account, I
to suggest that
been perfect so
it represents an
I’ve been either
Open Market

Committee meetings for a long time, and I’ve never
seen the Committee
more serious about hitting its
long-run targets for money growth.
So I think we
are on the right road, and if we can hold to our
course, I believe that most people will probably be
surprised by the speed with which inflation can be
Since I consider inflation
brought under control.
the root cause of most of our other economic problems, I believe that any progress we make on the

INSTRUMENTS

inflation front. will yield lower unemployment,
higher
productivity
and growth, a stronger balance of payments, and lower interest rates. The trick now will
be to hold firm, even if we run into some stormy
economic weather in the months immediately
ahead.
I can assure you that we in the Federal Reserve will
hold firm, and I believe strongly that the nation will
reap the benefits of these policies sooner than most
people seem to expect.

OF THE

MONEY

MARKET

The Federal Reserve Bank of Richmond is pleased to announce the publication of the
fifth edition of Instruments of the Money Market. This book describes the major money
market instruments
and the institutional
arrangements
of the markets in which these
instruments
are traded. Domestic money market instruments
discussed include Treasury
bills, Federal agency securities, Federal funds, repurchase agreements,
CDs, commercial
paper, and bankers’ acceptances.
There are also chapters on Eurodollars,
the Federal
In
Reserve discount window, and the dealer market for U. S. government
securities.
addition, there is a chapter on short-term
investment
pools, e.g., money market mutual
funds, which purchase large amounts of money market instruments.
The book begins
Virtually
the entire fifth edition
with an introductory
chapter on the money market.
(1981) is new.

FEDERAL

RESERVE

BANK

OF

RICHMOND

5

UNEMPLOYMENT ITS MEASUREMENT:
AND
IMPLICATIONS
FROMA SURVEY LONG-TERM
OF
UNEMPLOYMENTBALTIMORE
IN
CITY*
The Survey
proposal

of Unemployed

that this Bank undertake
thereby

and “Discouraged”

by James F. Tucker,

to provide

Workers

in Baltimore

of the Federal

a survey to collect information

the general

Mr. Tucker

Vice President

on discouraged

public with a better understanding

then contacted

Professor

Mr. Ayele and William E. Cullison, Research
Mr. Ayele and Ms. Yvette Armstead

workers in Baltimore

of their unemployment

Moges Ayele of Morgan

and the Federal Reserve Bank of Richmond

evolved from a

Reserve Bank of Richmond,

State University,

contracted with him to perform

Off icer of this Bank, developed

compiled

and bears sole responsibility

Over the past thirty-five
years, the one economic
statistic most often chosen as a measure of economic
welfare and capacity utilization
has been the unemployment rate. Because of its wide-publicity
and its
unfavorable
straightforward
nature,
seemingly
changes in that statistic can have serious repercussions for incumbent politicians and high-level bureaucrats.
Despite its widespread
use, however, many
economists
believe that the unemployment
rate as
currently
computed is a relatively poor statistic for
measuring
either economic welfare or labor market
capacity utilization.
Geoffrey Moore, former Commissioner
of the Bureau of Labor Statistics and Director of the National
Bureau for Economic
Research,
for example, has

the results,

for any errors

ECONOMIC

REVIEW,

the survey.

Mr. Cullison wrote this article
or omissions in it.

argued for some time that the unemployment
rate has
been overemphasized
as a target variable for economic policy.1
He has suggested, instead, that an
employment/population
ratio might be a preferable
economic indicator.
Moore was critical of both the
definitional
concept of the unemployment
rate2 and
the accuracy of the unemployment
data, which are
subject to relatively more sampling error than the
comparable employment
data.
Moore criticized the
definition
of unemployment
because of its subjectivity. Unemployment
is defined as it is because one
may be out of work voluntarily

or involuntarily,

and

the concept of unemployment
excludes those who are
voluntarily
unemployed.
This subjective characteristic of the unemployment

data caused Moore to con-

clude that they were “softer”
* We wish to thank the members
of our advisory
committee, Pearl C. Brackett,
Deputy Manager
of the Baltimore Regional
Chapter
of the American
Red Cross;
Catherine
P. Doehler, Director
of Financial
Development
of the Baltimore
Regional
Chapter of the American
Red
Cross;
Mary
Garland,
Coalition
of Peninsula
Organizations;
Robert
Hearn,
Assistant
Provost,
The Johns
Hopkins
University;
William
Hoffman,
Jr., Education
and Career
Specialist,
Greater
Homewood
Manpower
Center;
Steve Horwitz,
Director
of Public Information,
Maryland
Motor Vehicle Administration;
Lenwood
Ivey,
Executive
Director,
Urban
Services
Agency;
Nat Jackson, Director,
Greater
Homewood
Manpower
Center;
Michael
Jans,
Director,
Northeast
Community
Organization;
Dea Anderson
Kline,
Director
of Community
Affairs,
The Johns
Hopkins
University;
Larry
Pencak,
Director,
Southeast
Community
Organization;
Louis
Schreiber,
Executive
Director
of the Northwest
Baltiand Marie
Washington,
Director,
more
Corporation;
East Baltimore
Community
Corporation,
for their assistance with this project.
Special thanks are due the Northwest
Baltimore
Corporation
for
allowing
the
field
workers
to use its Northwest
Pimlico
Multipurpose
Center as a base of operations.

6

the survey.

the questionnaire.

of Baltimore, led the team of field workers conducting

Ms. Donna Howell of Richmond

and

problems.

ployment

statistics

than the em-

data.

1 See Geoffrey Moore, “Employment,
the Inflation-Recession
Dilemma,”
in
nomic Problems,
1976, (Washington
prise Institute,
1976), pp. 163-82, for
of those views.

Unemployment
and
Contemporary
Eco: American
Entera sample exposition

2 The unemployment
rate. is simply the percentage
of
unemployed
persons
in the civilian
labor force.
The
latter is defined as the sum total of employed and unemployed persons. Persons are considered
to be unemployed
if (1) they did not work at all for pay during the survey
week or if they worked less than 15 hours for no pay in a
family enterprise,
(2) they made at least one specific
effort to find a job within the past four weeks, and (3)
they were available for work and willing to work during
the survey week.
Persons
are also considered
to be
unemployed
if they are on temporary
layoff awaiting
recall.
Employed
persons
are defined as persons
who
worked at least one hour for pay or at least 15 hours for
no pay in a family enterprise
during the survey week.

SEPTEMBER/OCTOBER

1981

Moore’s criticism was not the only one. The unemployment
rate also came under suspicion from a
number of quarters
in the late sixties and seventies not only because it did not appear to measure
capacity utilization in the economy in a manner consistent with other economic indicators
but also because rather large increases
in employment
were
having little effect upon reducing the unemployment
rate. The latter reason for suspecting the data was
directly related to the voluntary-involuntary
nature
of unemployment,
as large numbers of females who
had previously
been voluntarily
out-of-work
were
entering the labor market to look for jobs, and thus
being counted among the unemployed.
In addition,
large numbers
of young workers,
who had never
worked regularly
before, were entering
the labor
market, thus changing from being voluntarily
out of
work to being involuntarilyunemployed.
A number
of economists recognized the problem of interpreting
changes in the unemployment
rate over time and
attempted to adjust it for the demographic changes.3

natural
rate of unemployment
to which the economy gravitates.
In terms of measuring labor market
capacity utilization,
however, the two terms have
similar implications.
Robert J. Gordon, for example,
defines the natural rate as
. . . the economy’s

long-run
equilibrium
level of
unemployment
that occurs when output equals its
long-run
natural
level and is a situation
in which
the actual
inflation
rate turns
out to be exactly
what people anticipate.4

Thus defined, the natural
the rate at which inflation
accelerates nor decelerates.

The natural
rate concept focused attention
on
the meaning of particular levels of the unemployment
rate rather than changes in the rate, for it became
increasingly
apparent that policies that lowered the
unemployment
rate below its natural
level meant
Empirical
estimation
of the
accelerating
inflation.
natural rate, however, was made difficult because of
its tendency to change over time. Robert J. Gordon
has estimated it at 5.2 percent in 1964, 5.6 percent in
5.4 percent in the mid-1970s.
1972, and around
Phillip Cagan has argued that the “noninflationary
unemployment
rate (which in his
full employment”
framework is the unemployment
rate at which inflation is stable, i.e., the natural rate) rose from 4.7
percent in 1956 to between 5.9 percent and 6.3 percent in 1977.5 Cagan has noted a number of reasons
for the change in the natural rate over time.
His
reasons included such structural changes as shifts in
the composition
of the labor force, extended unemployment insurance
coverage, liberalized
unemployment insurance
benefits, increased minimum
wage,
increased work registration
requirements,
and other
manpower programs.

Other economists criticized the unemployment
statistics in another way, arguing that unemployment
These
was understated
in the official statistics.
critics noted that part-time workers were considered
to be employed even if they wanted full-time work
and were only taking odd jobs to finance their search
for a full-time job. Moreover, they argued that the
once-a-month
job search criterion was too restrictive,
since it excluded a number of workers who wanted
jobs but were too “discouraged”
with their job prospects to search that often.
The relevance
of conventional
employment
and
unemployment
data for macro policy has been debated for several decades. During the sixties, it was
taken for granted that government
policy could reduce unemployment
at the cost of slightly higher
As actual events and economic
rates of inflation.
analysis began to discredit the notion of a stable
inflation-unemployment
trade-off during the seventies, attention began to focus on other concepts, the
“natural”
and “noninflationary”
rates of unemployment.
The difference in terminology
depends upon
whether one accepts the hypothesis that there is a

George Perry, Michael Wachter, Robert Hall, and
France Modigliani
and Lucas Papademos estimated
a related concept, the “noninflationary”
unemployment rate. They seemed to reach a consensus that
the rate had risen to around 5.5 percent in the mid1970s compared to a 4.0 percent-4.5 percent level in
the 1950s.
It will be the contention
of this paper that the
determination
of the natural and/or
the noninflationary rate of unemployment
is made unnecessarily
difficult because of a flawed definition of unemploy-

3 See, for example,
George
Perry,
“Changing
Labor
Markets
and Inflation,”
Brookings
Papers on Economic
Activity,
3:1970, pp. 411-41; Robert
Hall, “The Process
of Inflation
in the Labor Markets,”
Brookings
Papers on
Economic
Activity,
2:1974, pp. 342-93; Michael Wachter,
“The Changing
Cyclical
Responsiveness
of Wage Inflation,” Brookings
Papers on Economic
Activity,
1:1976,
pp. 115-59; and Modigliani
and Papademos,
“Monetary
Policy
for the Coming Quarters,”
New England
Economic Review, Federal Reserve
Bank of Boston, March/
April 1976, pp. Z-35.
FEDERAL

RESERVE

rate of unemployment
is
is stable and thus neither

4 Robert
Gordon,
Macroeconomics
Brown and Company,
1978), p. 212.

(Boston:

Little,

of Inflation
and the
5 Phillip
Cagan, “The Reduction
Magnitude
of Unemployment,”
in Contemporary
Economic Problems,
1977, (Washington:
American
Enterprise Institute,
1977), p. 40.

BANK

OF

RICHMOND

7

ment. To develop a framework for subsequent analysis, assume that there exists a definition
of unemployment that includes only those persons who (1)
honestly want full-time work (i.e., would accept any
kind of regular job), (2) have adequate job skills
(a set of attributes that would make them desirable
employees from the viewpoint of an employer),
(3)
are earnestly searching for full-time work, and (4)
have a relatively low reservation
wage (would work
for any wage that would provide a net addition of
household income).
Unemployment
so defined might
be called, for want of a better term, hard unemployment.
Assume further that accurate data could be
collected on this type of unemployment.
The level of
hard unemployment
would always be greater than
zero because of the existence of frictional unemployment, but it would not normally be far above the
frictional level. Hard unemployment
would exclude a
number of persons who are currently included in the
unemployment
statistics.
By contrast, assume that unemployment
could also
be defined to include everyone who is presently classified as unemployed plus “discouraged”
worker6 and
any others who would accept the offer of an appropriate job if they did not have to expend any effort
in searching for it. This set of unemployment
data
might be labeled soft unemployment.
Now, full employment in the labor market has often
been defined to be whatever the unemployment
rate
would be at its frictional level. By this definition it
is clear that measures of full employment
may differ
markedly depending
upon their corresponding
conHard unemployment,
for
cepts of unemployment.
example, might be maintained
close to its frictional
level without increasing
inflation.
Soft unemployment probably could not be reduced near its frictional level without exacerbating
inflation.
The key difference between the two concepts is
that hard unemployment
excludes individuals
who
either do not have an intense desire to find a job or
do not have job skills and personality
characteristics
that are attractive to employers.
It is of course true
that it would be difficult in practice to measure hard
unemployment,
but data could be published that are
closer to the conceit.

6 Discouraged
workers
are presently
defined as workers
who want work, but have not looked for work within the
latest four weeks because:
(1) they believe no work is
available
in their line of work or area; (2) they could
not find work: (3) they lack the necessary
skills, schooling, training or experience;
(4) employers
think they are
too young or too old: or (5) they have other personal
handicaps
in finding work.
Workers
giving additional
reasons,
such as family responsibilities,
school
attendance, or ill health, are excluded.

8

ECONOMIC

REVIEW,

The Bureau
of Labor
Statistics,
for instance,
might publish data on individuals
who search for
work frequently
(at least twice per week) and intensely (take substantial
overt actions, not just peruse the want ads in the newspaper).
In addition,
statistics could be provided on intense searchers who
had been unemployed
for more than, say, 5 or 6
weeks. The search criterion is crucial to the concept
of hard unemployment
because of an assumed connection between intensity of job search and a pronounced (or strong) desire to work. The length of
time of unemployment
is important
to reduce frictional unemployment.
In any event, this operational
definition
of hard unemployment
should make the
level of the natural rate easier to find (and close to
zero) for (1) much of the frictional unemployment
would be excluded and (2) workers with insufficient
job skills would probably
have stopped searching
intensely.
Such a measure would not of course be useful as a
measure of economic welfare (or social progress,
etc.).
To evaluate
these broader
social goals a
broader measure would be called for.
But if the
assumption
that intensity
of job search is an important characteristic
of all workers who are truly
involuntarily
unemployed
is correct, the narrower
measure
should provide researchers
with a better
and more stable measure of the natural rate of unemployment.
This

paper

will

investigate

the

claim

that

fre-

quency and intensity of job search are crucial measures of the degree of an individual’s
attachment
to
the labor

market

using

long-term

unemployed

the results
workers

of a survey

in Baltimore

of

City.

I.
AN OVERVIEW

OF THE

SURVEY

The Federal
Reserve
Bank of Richmond
contracted with Dr. Moges Ayele of Morgan
State
University
in Baltimore,
Maryland,
to survey longterm unemployed
and/or
discouraged
workers
in
order to determine:
(1) general characteristics
of
long-term unemployed workers in Baltimore, (2) the
nature of the economic distress faced by them and
their families, (3) what alternative income was available to them and their families, and (4) the differences, if any, between unemployed
and discouraged
workers.
Two hundred and thirteen persons were surveyed
during the summer of 1980. Persons eligible for the
survey (1) were between the ages of 18 and 64, (2)
SEPTEMBER/OCTOBER

1981

breakdown of the responses to the survey.
As noted
above, the original intent of the study was to survey
a larger proportion
of discouraged
workers, but the
field crew found it difficult to find respondents
who
would admit to not having looked for a job within
the past four weeks. According to the national statistics, the ratio of unemployed to discouraged workers was 7.6 to 1 during the summer of 1980. The
ratio in this survey, 6.0 to 1, was in comparison
weighted slightly toward discouraged workers.

had not worked regularly for six months, (3) wanted
a regular job at the time of the interview, and (4)
either would have looked for work within the previous four weeks or would not have looked for one of
the following reasons : (a) they thought that no jobs
were available, (b) they lacked necessary schooling,
training,
skills, or experience,
(c) experience
indicated to them that employers thought they were too
young or too old, or (d) there were personal reasons
(such as a criminal record, perceived racial discrimination, etc.) leading them to think that they could
not find a job. Potential
respondents
who did not
want a job or whose home responsibilities,
school
attendance,
or physical or mental disabilities
prevented them from accepting a job were to be excluded
from the survey.

II.
THE

GENERAL

195 POTENTIAL

CHARACTERISTICS
WORKERS

IN BALTIMORE

Locating
long-term
unemployed
persons proved
unexpectedly
difficult.
The initial plan was to (a)
locate areas of the city with high concentrations
of
unemployment,
(b)
identify
community
groups,
churches, and social welfare organizations
operating
within these areas, and (c) work through
these
organizations
to locate the specific persons to be
It was not difficult to identify the
interviewed.
appropriate community groups, but enlisting their aid
in locating candidates to be surveyed was not easy.
Even so, a majority
of the candidates
were found
through the community
organizations.

OF

RESIDING

CITY

The respondents to the survey were predominately
female, black, and young (57 percent were in the
18-24 age groups).
A majority
(52 percent)
had
earned a high school diploma, and only slightly more
than one-fourth
had some formal training over and
above high school.
About one-third of the respondents surveyed were household heads.
The respondents generally were members of larger households,
only 2 percent lived alone, and 57.3 percent lived in
larger-than-four-person
households.
Fifty-two percent of those surveyed had been out
of work for more than a year. Almost 9 percent had
Most had held their last
never worked full-time.
job for a relatively short interval, 54.1 percent for
less than six months.
Forty-seven
percent left their
job because they were laid-off, terminated,
because
the job was temporary,
or because their employer
went out of business.
Almost the same number, 44.6
percent, left their last job for medical or personal
reasons.
A great majority of the respondents,
97.4
percent, had searched for a job within the past year,
although only 86 percent had searched within the
last four weeks.

The Baltimore Mayor’s Office of Manpower
Resources had a list of unemployed
persons that was
originally expected to be a primary source of candidates for interviews, but it turned out to be relatively
useless. Out of 500 names taken from the list, only
30 qualified;
most were no longer unemployed.
A
large list of names from which a sample could be
selected in a statistically
meaningful
way, therefore,
was not available.
As a result, the survey results
may not be representative
of the unemployed worker
in Baltimore.
Budget limitations
restricted the survey to 217 families, which is too small for many
statistical procedures.
Also, the study contains far
fewer discouraged
workers than was first planned,
and therefore few meaningful inferences can be drawn
from the comparisons
between discouraged
and unemployed
workers.
Given these limitations,
the
survey can best be interpreted
as a pilot project,
with the results suggesting
further research.

When those respondents
who were not looking for
work were asked when they would resume their
search, 47 percent responded that they expected to
begin within the next several months.
Several were
waiting for school to reopen either because they
thought more jobs would be available or because they
would not need to make child-care arrangements.

The survey furnished
195 usable interviews;
24
respondents
could be classified as workers discouraged for job market reasons, 4 could be classified as
workers discouraged for other reasons, and 167 were
classified as unemployed
(without a job but actively
seeking work).
Appendix Table A-l gives a detailed

The responses
to one question
were especially
noteworthy.
When asked whether they thought they
could find a job if they sought one actively, 62.6 percent of the respondents thought either that they could
find a job or that they would have a fair chance.
Since the individuals
were not employed, a “yes” or

FEDERAL

RESERVE

BANK

OF

RICHMOND

9

“fair chance” response to the question implied that
they were not “actively”
seeking work.
Many admitted that they were not seeking work as actively
as possible because of the expense of job search,
mainly for carfare and clothes. Others gave no such
reason.
Almost half (43.4 percent)
of the respondents acknowledged
that they would not accept some
of the jobs available to them.
Although a number of the respondents
(61.3 percent) were living at or below the poverty level, few
(12.7 percent) were in danger of losing their homes
or other property.
Consistent
with this finding, a
majority
of respondents
(52.5 percent)
indicated
that they either had a reservation
wage higher than
the minimum wage or that there were jobs that they
could get but would not want (reservation
working
conditions).
The questionnaire
included the question,
“What
is the lowest pay that you would accept?”
The responses to that question,
however, were not very
useful because the interviewers
influenced the respondents’ answers.
Respondents
who did not have a
reservation
wage rate in mind were told that the
minimum
wage was $3.10.
Consequently
the responses to the question had a floor of $3.10 per hour.
The reservation wage data, however, is not totally
useless.
A test was performed to analyze how well
the data conformed to a priori expectations,
and the
results were encouraging.
A priori, one would have
expected,
(1) that respondents
who thought jobs
were available that they could get but would not
want would have had the higher reservation
wages,
(2) that household heads, faced with more pressing
needs to find jobs, would have had lower reservation
wages, and (3) that females, who earn lower wages
on average than males, would have had lower reservation wages. The statistical analysis confirmed the
a priori expectations.7
The respondents
received approximately
40 percent of all household income from income support
As might be expected, when the unemprograms.
ployed respondent
was the head of household, the
proportion
of household income from support payments was higher, averaging 88.4 percent for the 39
families in which a female respondent was the house7 RESWAGE

=

35.97 +
(3.35)

- 23.58 • SEX
( - 2.96)

22.35 • JBGT
(2.92)
-

1.82 • HHD,
(-1.82)

where RESWAGE
is the reservation
wage; JBGT
is 1
if the respondent
thought
he could get a job, 0 if not:
SEX
equals 1 for female, 0 for male; and HHD is 1 if
The figures
in
the respondent
is a household
head.
parentheses
are “t” statistics.

10

ECONOMIC

REVIEW,

hold head.
By comparison,
an average
of 39.2
percent of household income came from support payments for the 80 families
in which the female
respondent
was not the household head. The corresponding percentages for male respondents
were 41.6
percent for nine household heads and 15.4 percent for
54 who were not household heads.
The income data show that the respondents
earned
$560 per month on average at their last job (in
dollars of summer 1980 purchasing
power), with a
standard

deviation

household
month,
holds.

of approximately

disposable

income

$300.

averaged

but it was often for support
Per capita household

with a standard

deviation

Current

$862.11

per

of large house-

income averaged

$165.56

of $122.37.

The potential workers surveyed had generally held
low-skilled jobs in their previous employment.
Only
approximately
required
searching

one-fourth

of them had held jobs that

skills or training.
They were generally
for jobs that were similar to those last held.

A breakdown
Table A-2.

of jobs held last is shown in Appendix

III.
IMPLICATIONS
INTERPRETING
AND

OF SAMPLE

EMPLOYMENT,

DISCOURAGED

RESULTS

FOR

UNEMPLOYMENT,

WORKER STATISTICS

As a result of Congressional
concern about labor
market data, Congress created a National Commission on Employment
and Unemployment
Statistics
that published its final report on Labor Day 1979.
Several issues recurred
frequently
in its deliberations:
(1) the relationship
of unemployment
to
economic hardship,
(2) the usefulness of the unemployment
rate as an indicator of labor market capacity utilization,
(3) the advisability
of including
discouraged
workers
in the unemployment
count,
and (4) the desirability
of instigating
research on
wages and working conditions sought by unemployed
workers.
These issues will be discussed in turn.
Unemployment

and

Economic

Hardship-

When present
concepts
and definitions
of unemployment
were developed during the 1930’s, unemployment and severe economic hardship
were closely
related.
The labor force was
mostly made up of
adult males
and unmarried
women.
. . .
With
limited family savings and few government
support
unemployment
most often
resulted
in
programs,
poverty.
Subsequent
developments
[such
as the
change in age distribution
of the labor force, the
growth of income support programs,
and the rising

SEPTEMBER/OCTOBER

1981

numbers of multi-earner
families] however, have
substantially
weakened the links between unemployment and economic hardship.”

clined a job because of low pay, and 35 percent of the
respondents
thought that they could get a job but
wouldn’t want it because it either paid too little, was
menial, was too demanding
physically, or put more
simply, that the respondent
would not like the job.
In answer to the question, “Do you think you will
be successful in finding a job if you actively look for
one?” 48.9 percent said yes and an additional
15.3
percent thought that they would have a fair chance.
Although this last finding may be somewhat ambiguous because of the use of the phrase “actively
look for one,” the responses generally indicate that a
number of unemployed
workers could get some sort
of job, yet they chose instead to remain unemployed.
The link between unemployment
and economic hardship indicated by our survey, therefore, seems to be
rather loosely drawn.

The survey of unemployed
workers in Baltimore
illustrates
the link between unemployment
and economic hardship.
A majority
of persons surveyed,
67.5 percent, received household
incomes that fell
below the poverty threshold.”
Those respondents
who were household heads were worse off, as might
be expected, and 84 percent of them had household
The implication
incomes below the poverty level.
here is quite clear (and hardly surprising)-unemployment is more directly associated with hardship
if it is the primary breadwinner
who is unemployed.
But the survey also included two questions designed
to show acute economic distress, “Is there any danger
of repossession
or eviction (from your home), because of your being unemployed?”
and “Have any
possessions had to be sold or repossessed because of
your being unemployed?”
Only 13 percent of the
respondents were in any such danger. There were 19
positive responses to the home eviction or repossession question out of 211, and 14 positive responses
to the other repossession
question.

The Unemployment
Capacity
Utilization

8 National
Commission
on Employment
and Unemployment Statistics,
Counting the Labor Force (Washington:
U. S. Government
Printing
Office, 1979), p. 38.
9 In autumn 1979, the annual cost of living for a lower
income family of four in the Baltimore,
Maryland,
SMSA
was $12,772 according to the U. S. Department
of Labor,
[News, USDL
80-278, April 30, 1980.] $187 or 1.5 percent higher
than the equivalent
U. S. urban average
budget.
The poverty
threshold
in 1978 for an average
U. S. nonfarm family of four (the latest data available)
was $6,662 according
to the U. S. Bureau
of Census,
[U. S. Department
of Commerce,
Bureau of the Census,
Current
Population
Reports,
Series
P-60,
No.
124,
Characteristics
of the Population
Below
the Poverty
Level:
1978 (U. S. Government
Printing
Office, Washington, D. C., 1980), p. 208.1 which was 67.4 percent of
Thus the
the 1978 U. S. nonfarm
low income budget.
poverty threshold for 1979 in Baltimore
was estimated
to
be 67.4 percent of the 1979 low income budget figure, or
$8,608.
Since the survey was taken in summer 1980, and
prices in Baltimore
rose 12.2 percent
between
autumn
1979 and summer 1980, the estimated
poverty
threshold
for a family of four living in Baltimore
in summer 1980
was adjusted to $9,659.
The Census published
a breakdown of poverty level thresholds
by family sizes for U. S.
urban families in 1978, so the poverty level for families
of different sizes was found by adjusting
the figures in
the Census’s breakdown
to the estimated
Baltimore
level.
Accordingly,
they were all increased
45 percent,
the
percent that $9,659 is larger than $6,662.
The resulting
estimated
poverty
levels are shown in Appendix
Table
A-3.

RESERVE

of

The current
definition
[of unemployment]
gives
equal weight
to both a family
head looking
for
full-time
work and a teenager
looking for a Saturday afternoon
job.
Some observers
feel that [the
statistics
thus]
. . . overstate
excess labor supply
and understate
labor
market
tightness.
Others
maintain
that the unemployment
figures
understate
the excess
labor
supply
because
the labor
force definition
excludes individuals
. . . who would
take a job if one were available
but, at present,
are
not searching.
[Furthermore,]
. . . (m) any analysts
[argue]
that
the unemployment
rate
is higher
today than in the past . . . because there are more
young persons
. . . and more adult women in the
labor force.
It has [also] been suggested
that [the
extensions
in duration
and coverage
of unemployment benefits
during
the 1970’s]
. . . may have
[increased]
. . . unemployment
by subsidizing
seasonal and casual
workers.
. . .
Observers
have
suggested
that [registration
requirements
for] . . .
food stamps
or aid to families
with dependent
children
has increased
the unemployment
rate. . . .
On the other hand, some analysts
have suggested
that
the growth
in [income
support]
programs
might
have reduced
unemployment
by providing
income incentives
for individuals
to withdraw
from
the labor force. . . . Several
analysts
have [also]
observed that the growth of . . . training
programs
may have lowered
the unemployment
rate.
. . .
other issues,
not yet fully explored,
are how the
growth of multi-earner
families
affects
unemployment, . . . and whether
the presence
of a second
earner leads to longer job-seeking
by other household members.10

Finally, if economic hardship were automatically
associated with. long-term unemployment,
one might
assume that a worker so affected would lower his
expectations and take a job that he previously would
not have taken.
In the survey, however, 11.8 percent of the respondents
indicated that they had de-

FEDERAL

Rate
as an Indicator
in Labor Markets

The quote from the National
Commission
includes
most of the major criticisms of the use of the unemployment rate as a policy target [their term] or an
economic indicator.
The criticisms revolve around
10 National Commission
on Employment
ment Statistics,
op. cit., pp. 36-37.

BANK

OF

RICHMOND

and Unemploy-

11

the issue of finding a natural rate of unemployment,
which was discussed at the outset of the paper.
Recall that at that time the concepts of hard and soft
unemployment
were developed to provide a framework for the search for a natural rate. The information from the survey on how closely currently defined
unemployed
and discouraged
workers conform
to
hard unemployment
is thus important to an analysis
of the unemployment
rate as a target for aggregate
economic policy in the sense meant by the National
Commission.
The survey questioned respondents
about the frequency with which they looked for a job, the methods
used in searching for a job, whether they had refused
a. job because of low pay; whether they thought that
jobs were available that they could get but would not
want; their reservation
wage; their household
inincome.
come; and the sources of the household
Statistical
analysis indicated that search frequency
waned as time passed since the respondent
held his
last job. Also, older workers tended to search less
frequently and to be less optimistic about being able
to land a job. In another vein, it was found that the
more income earned on the last job and the higher
the education level of the worker, the shorter was the
duration of unemployment.
Household
income per
capita, on the other hand, had no significant
effect
on either the frequency of job search or the length
There was a significant relationof unemployment.
ship between per capita household income and the
unemployed worker’s expectations of landing a jobthe higher the per capita household
income, the
greater the chance that the worker thought he could
get a job if he searched actively.
These findings indicate that if the unemployment
rate is to be used as a measure of capacity utilization,
unemployment
should have a search criterion associated with it, for search frequency indicates the degree
of attachment to the labor force. In devising an ideal
index of labor market capacity, therefore, more attention probably should be given to frequency of search
and type of search. An individual, for example, who
reads newspaper
ads once a month certainly has a
weaker attachment
to the labor force than one who
actively looks for a job.
In the survey,
reportedly
five days,

60 persons

(1) searched
(2) searched

ployment service, employment
direct contact with employers,
within the survey week.
considered
to have been
intensely
12

than the others

were

interviewed

who

for jobs at least once every
either through a public emagencies, or through
and (3) had searched

These persons could be
searching
relatively
more

These intense searchers were significantly
different
from the other respondents
in only two ways: they
were generally younger,11 and less time had elapsed
since their last job. l2 Compared to 107 unemployed
workers who were not intense searchers, the same
general pattern held.
The intense searchers were
younger,13 and less time had elapsed since their last
job.14
Over half, 37 of 60, of the intense searchers had
been out of work for 6-12 months.
Almost one-third
more, 18 of 60, had been out of work for l-3 years.
Only one (18 years old) had never held a job. The
intense seekers also earned somewhat higher wages at
their last job than did the other respondents.
The
sixty averaged $603 per month (in dollars of the purchasing power of summer 1980) whereas the others
earned only $560 on average.
The other 107 unemployed workers averaged earnings of $555 on their
last job. Neither of these average earnings was (statistically)
significantly
different
from that of the
intense searchers.
These differences
illustrate
the
potential importance of the search criterion, however,
for they suggest that the searchers
may possess
greater job skills since they seem to have earned
relatively higher wages in their previous employment.
(A larger sample is necessary to come to any definite
conclusion on this issue, however.)
Note, however,
that the responses
of the two
groups did not differ appreciably on any other quesHousehold
incomes were not significantly
tions.
different between the two groups, nor were the numbers of families in danger of repossession or eviction.
These similarities
of response indicate that if the
search intensity criterion in the unemployment
statistics were to be tightened,
no particular
income
group would be systematically
excluded from the
study. Rather, the persons excluded would be those
who had been out of work for long periods and were
These types of persons probably
somewhat older.
should be excluded from the observed unemployed if
one is searching for a measure of hard unemployment
that is useful as an indicator of labor market capacity
utilization.
Discouraged
and
pared
Discouraged

11t = 2.75 for the difference
200 degrees of freedom.
12t

ECONOMIC

REVIEW,

=

=

between

sample

means

with

3.42.

13t = 1.95 for the difference
and 165 degrees of freedom.
14t

interviewed.

Unemployed
Workers
Comworkers are defined as work-

2.22.

SEPTEMBER/OCTOBER

1981

between

the sample

means

only 40 percent of the unemployed workers
last jobs for medical or personal reasons.

ers who want a job, and who are available for a job,
but who have not searched within the past month
(and hence are not counted as unemployed).
The
search criterion thus is used as an indication
of an
individual’s degree of commitment
to the labor force.
The National
Commission
studied the question of
whether the definition
of unemployment
should be
broadened
to include discouraged
workers,
finally
deciding against that course of action.
Instead, they
recommended publishing data on persons “marginally
attached to the labor force,”15 that is, workers who
had searched within six months and who meet other
criteria discussed below.

In answering the question related to resuming their
job searches, 53 percent of the discouraged workers
indicated that they had no plans to resume their
Fourteen percent indicated that they would
search.
resume their search when their personal
situation
changed (e.g., after school reopened and the children
entered school, when the person’s health improved,
etc.).
The remaining 32 percent indicated that they
would resume their job search within several months.
The two groups of respondents had roughly similar
assessments of their job prospects.
When they were
asked whether they thought they could find a job if
they actively sought one, 65 percent of the discouraged workers and 62 percent of the unemployed
thought that they would have at least a fair chance.
Somewhat inconsistently,
only 55 percent of the discouraged workers and 41 percent of the unemployed
thought that jobs were available that they could get
but would not want.

The Baltimore survey allowed a loose comparison
of discouraged and unemployed
workers in order to
determine
the nature of their respective unemployment experiences.
The results should be interpreted
with caution, however, because the sample of discouraged workers may not have been representative:
26 of the 28 discouraged
workers
surveyed
were
female, compared
to 97 of the 167 unemployed.
Some of the survey responses
that are ostensibly
characteristic
of discouraged workers, therefore, may
simply represent responses characteristic
of females.
However,
70 percent of the national
discouraged
worker count is female, so sex differences may be an
integral part of the data. Another important difference in the respondents
that might impart bias to the
survey results was the difference in household status
between the respondents
of the two groups. Only 31
percent of the unemployed
respondents
were household heads compared to 58 percent of the discouraged
respondents.

In spite of the fact that the respondents
had all
been unemployed for six months or longer, few were
in danger of being evicted from their home or having
Only one discouraged
their mortgage
foreclosed.
worker and 17 unemployed workers (3 percent and
10 percent respectively)
were in such distress.
Also,
only one discouraged
worker and 13 unemployed
workers were in danger of having other property
repossessed.
The discouraged worker households seemed to rely
more heavily upon support payments (58 percent of
household income) than did the households of unemployed workers (39.2 percent of household income),
but this difference was a matter of sex and household
status rather than a distinguishable
characteristic
of
Respondents
who were
the discouraged
worker.
female unemployed
household heads, for example,
averaged 92.2 percent of household income from support payments, compared to an average of 78.9 percent for respondents
who were female discouraged
household heads.

The discouraged
workers
surveyed
had usually
been out of work longer than the unemployed workers. Fifty-two
percent of the unemployed
workers
had been out of work for 12 or fewer months, but
only 27 percent of the discouraged workers.
Consistently, one-third of the discouraged workers had been
out of work for five or more years, but only 9 percent
of the unemployed workers.
Only 50 percent of the discouraged
workers had
held their last job longer than three months, compared to 63 percent of the unemployed workers.
The
two groups also differed in their reasons for leaving
Thirty-three
percent of the unemtheir last job.
ployed but only 20 percent of the discouraged workers left their last job because of factors related to the
labor markets,
i.e., being laid-off or terminated.
Comparatively,
half of the discouraged
workers but

15 See National Commission
on Employment
ployment
Statistics,
op. cit., p. 3.

The unemployed
workers
surveyed
were better
Only 13.8
trained than the discouraged
workers.
percent of the discouraged
workers had training,
either technical or general, above the high school
diploma level, but 27.5 percent of the unemployed
workers
had training
in addition
to high school.
Consistently,
unemployed
workers seemed to have
held better jobs, on average, than the discouraged
The last job held by 27 percent of the
workers.
unemployed
workers
required
skills or training
whereas only 7.4 percent of the discouraged workers

and Unem-

FEDERAL

RESERVE

quit their
-

BANK

OF

RICHMOND

13

previously
held skilled jobs.
The majority
(56.7
percent)
of discouraged
workers had held jobs as
laborers,
cleaning
and food service workers,
and
health and school aides; 13 percent had held sales
and clerical jobs; and 10 percent had never held a
regular job.
Considerably
fewer, 38.5 percent, of
the unemployed workers had last worked as a laborer,
cleaning or food service worker. More, however, (25
percent)
had previously
held relatively
unskilled
sales and clerical (and stock and shipping clerks)
jobs. As noted earlier, a detailed breakdown
of last
jobs held is shown in Appendix
Table A-2.
The
types of jobs sought by the unemployed workers also
generally required more skill.
Eighteen percent of
the unemployed
workers were searching for skilled
jobs.16
The income data also differed between the two
groups.
The difference, however, showed up in the
variances rather than the means of the data.
Adjusted into dollars of constant
purchasing
power
(summer 1980 consumer prices) the monthly income
earned at the last job held for discouraged
workers
averaged $550, whereas income last earned by unemployed workers averaged $567. The standard deviations were $264.52 and $329.29 for the discouraged
and unemployed workers, respectively.
This relation
also showed up in the household incomes of unemployed and discouraged workers.
Household income
per person averaged $168.80 for discouraged
workers’ households,
$175.30 for unemployed
workers’
households.
The standard deviations, however, were
$88.70 for the discouraged
workers and $127.60 for
unemployed
workers.
Using the standard statistical
F test for analysis of variance to analyze the differences in the standard deviations,
the probability
is
less than 0.025 that the two samples were drawn
from the same population.
In essence, this means
that the unemployed group was drawnfrom
a larger
cross-section
of income classes than the discouraged
group.
To analyze the responses to the question, “If you
seek actively, will you be successful?” in more depth,
a dummy variable taking values of 1 for yes, 0 for
fair chance, and -1 for no was created. A regression
equation estimated with the search success dummy
as the dependent variable and time elapsed since last
job and age as the two independent
variables showed
that both age and time elapsed since last job were
significant at the 10 percent level (t = -2.18
and
respectively)
for unemployed
workers,
t = -1.8,
16 Twenty-nine
percent of the unemployed
workers failed
to respond to this question.
Of those who responded
to
the question, 25 percent were seeking skilled jobs.

14

ECONOMIC

REVIEW,

but only time elapsed was significant
in the case of
discouraged
workers
(t = -4.5).
The negative
signs indicated that the older the respondent and the
longer it had been since he had held a job, the less
likely he was to think that he could find a job if he
searched harder.
The survey results for reservation
wages were
disappointing
as noted earlier, which may account
for the insignificance
of the differences in reservation
wages of discouraged and unemployed workers. Only
22 percent of the discouraged workers and 20 percent
of the unemployed
group said that they would require a higher wage than the minimum wage. The
results of various simple regressions
between characteristics of unemployed versus discouraged workers
are shown in Appendix Table A-4.
As noted earlier, the National
Commission
concluded that discouraged
workers should not be defined as unemployed
and recommended
that the
Labor Department
should begin publishing
a data
series on workers with a marginal attachment to the
labor force. The new statistic would encompass persons who were not presently in the labor force, who
were currently available for work, who had actively
sought work within the last six months, and who
wanted a job at the time of the survey.
The reasons
for seeking work that would have excluded workers
from the discouraged
worker ranks, e.g., child care
and home responsibilities
would not be used in defining whether a person was marginally
attached to
the labor force.
Almost 31 percent of the discouraged
workers in
our survey would have been excluded from the marginally attached labor force-they
had not searched
within the past six months.
On the other hand, over
half of the unusable
interviews
(unusable
because
personal
reasons surfaced)
in the survey were of
persons who would have been defined as marginally
attached workers.
A comparison
of those responses
to the responses of the group of discouraged
workers that would have been excluded by the six-month
limit lends support to the National
Commission’s
recommendation.
None
of
the
discouraged-but-not-marginallyattached workers was in economic distress as measured by danger of eviction or property repossession,
and generally those discouraged
workers (who had
not searched for a job in over six months)
seemed
to demonstrate
a modest (at best) ‘desire for employment.
The persons interviewed
who were disqualified as discouraged
workers for citing personal
reasons for not seeking work but who had looked for
work within six months, i.e., were marginally
atSEPTEMBER/OCTOBER

1981

includes persons unemployed
15 weeks or longer,
maintains
the lowest average level.
The long-term
nature of U1 removes unemployment
of a frictional
nature along with that of individuals
who have relatively little difficulty in finding jobs.
It therefore
resembles
hard unemployment,
mentioned
earlier,
most closely but it treats full-time and part-time job
seekers equally.
A better measure of labor market
capacity utilization
might include only relatively intense seekers of full-time jobs who had been unemployed for several weeks. Five to six weeks would
probably be sufficient to remove the frictionally
unemployed.

tached, seemed to have a relatively stronger attachment to the labor force. The number of such persons
interviewed
was quite small, however, because those
who cited reasons for not seeking work that would
have excluded them from the Labor Department’s
discouraged worker rolls were not supposed to have
been interviewed.
Of the 18 that were inadvertently
interviewed,
nine had looked for work within six
months. Six of the nine marginally attached workers
needed day care for their children and thought that
they could not earn high enough wages to justify
paying a baby sitter. Three of these respondents and
two others, however, planned to resume their job
search in September after the children went to school.
They gave two reasons for this. First, they thought
that jobs would open up as students left the labor
market and second, that school would relieve them of
their day-care needs.
Another
of the respondents
wanted a job but had a disabled child (heart disease)
for whom she had to care. That respondent was in
economic distress and was in danger of having her
furniture
repossessed,
although
her attachment
to
the labor force was relatively weak.

As a broader measure, the “marginally
attached”
data appear preferable
to the discouraged
worker
statistics.
According to the admittedly small sample,
workers who had not searched for work in more than
six months were not as eager to find a job as those
who had. In addition, the workers who planned to
resume their job search in September
seemed to
deserve some mention in the labor statistics.

Figure

Of the three remaining
marginally-attached-butnot-discouraged
workers, one was quite particular
about the type of job that he would accept and was
apparently
not very eager to have a job at present.
Another had an attachment
to the labor force, but
was waiting until September, when she was to enter a
social services job training program.
After that she
indicated that she would resume her job search.

Range of Unemployment
(Seasonally

U1
U2
U3
U4
U5

Unemployment
Measures
as Indicators
of Labor
Market Capacity
Utilization
The various unem-

U6

ployment rate measures published by the Bureau of
Labor Statistics,
U1 through
U7, are charted
in
the figure for the 1970-1978 time period.
As the

U7

1969-78

averages)

Persons unemployed 15 weeks or longer as a percent of civilian
labor force
Job losers as a percent of civilian labor force
Unemployed persons 25 years and over as a percent of civilian
labor force 25 years and over
Unemployed full-time jobseekers as a percent of full-time labor
force
Official unemployment rate-persons 16 years and over as a
percent of civilian labor force 16 years and over
Full-time jobseekers plus ½ part-time jobseekers plus ½ total on
part-time for economic reasons as a percent of civilian labor
force less ½ of part-time labor force
Numerator of U6 plus discouraged workers’ as a percent of
denominator of U6 plus discouraged workers

Source:

chart shows, they have similar cyclical fluctuations
and differ mostly with respect to the levels. U1, which
RESERVE

Measures,*
quarterly

Percent

In sum, the survey results suggest that the “marginally attached” concept might convey more information than the “discouraged”
worker concept. This
is particularly
true if the example of persons dropping out of the labor force during the summer because of day-care needs is not capricious.
If such
behavior is representative
of a large group of workers, as logically it might seem to be, the seasonal
variation of the sum of the labor force and marginally
attached workers should be less than the seasonal
variation in the labor force alone. This reduction in
seasonality would lessen the difficulties that are often
encountered
in interpreting
June and September employment and labor force data.

FEDERAL

adjusted

BANK

OF

U. S. Department of Labor Statistics. Reprinted in National
Commission on Employment and Unemployment Statistics,
Counting the Labor Force.

RICHMOND

15

APPENDIX

Table

RESPONSES

TO THE

SURVEY

DISCOURAGED

16

ECONOMIC

REVIEW,

A-l

OF UNEMPLOYED

WORKERS,

AND

SEPTEMBER/OCTOBER

OTHERS

1981

WORKERS,

Table

FEDERAL

A-l

RESERVE

(continued)

BANK

OF

RICHMOND

17

Table

A-l

(continued)

Table

Percent
Common
1.

Remarks

Personal

by

abilitiy

is constrained
-

to

compete

of

LAST

all

inadequate

for

education

costs

of

or

skills

2.6

experience

18.0

searching

5.1

Category

discrimination

in

job

1

Research

Worker

race

9.2

-

1
Worker

age

6.7

Experienced

transportation

and

local

ineffective

1
0

Miscellaneous
Shipping

problems

10.3

Bill

or

Students

return

improve

chances

aid

programs

ore

Clerical

2

Clerk

2

Collector

1

Aide

2
2

Nurse

to

school

Feels

in the

in finding

fall

Brick

should

a job

5.6

Mason

Mechanic

depressed/frustrated/helpless

16.9

0

1

or

0

1

Former

Ironer

or

Laundry
Packer

or

Printer,
Stitcher

or

Welder
Misc.

Operative

Deliveryman
Fork

Lift

Truck

Tow

Truck

Operator

Driver

0
2

4
2
12

Cleaning

Service

Service

Health

Aide,

Personel

Dental
or

3

17

2

Assistant,

Nurse’s

Aide

Services:

Child

Core

4

School

3

1

Monitor

1

1

Worker

Housekeeper

1

2

Usher

Protective

2

12

Worker

4

13

Worker

Service;

Health

Private

0

1
or

Laborer

Food

0

4

Sewer

0

2

Mfg.

0

1

Wrapper

0

3

Worker

0

1

Presser

0

1

Assembler

0

1

Misc.

Serviceman

0

3

Craftsman,

Misc.

0

2

Apprentice,

1

2

Croft

REVIEW,

0

1

Worker

0

1

Upholsterer

ECONOMIC

0

1

Apprentice

Metal

0

9

Repairman

Apprentice

Pressman
Sheet

18

0

10.8

Plumber
6.

0

1

Registered

either

inadequate

20

Clerk

Baker
5.

2

9

Teacher
Federal

8
2

Stock

4.

1

Teacher

Typist

search

of:

-

3.

Number

Service

Household

SEPTEMBER/OCTOBER

Worker

Worker

1981

- Guard

1

1

of

Discouraged

Receptionist

Encountered
because

of

Unemployed

Dancer

Soles
2.

BY RESPONDENTS
Number

jobs
Job

job

JOB HELD

Respondents

by:

insufficient

- high

Respondents

A-2

Table

ESTlMATED

MONTHLY

THRESHOLD

FOR

Table

A-3

INCOME

AT THE

BALTIMORE,

REGRESSION

POVERTY

SUMMER

THOSE
Children
Family
Male

0

Size

COMPARING

OF UNEMPLOYED

OF DISCOURAGED

TO

WORKERS

18

Under

2

1

RESULTS

CHARACTERISTICS

1980

A-4

4

3

5

6+

Discouraged

Head

Unemployed
Worker

Worker
Dependent

1 (under

65)
65)

531

“t”

Statistics

Statistics

$425

2 (under

“t”

Independent

TIMGON

DJBGT

618

638

$675

4

815

827

799

$839

5

984

996

964

939

$960

6

1128

1132

1108

1084

1052

7+

1421

1433

1405

1381

1349

AGE
AGE

TIMGON

3

AGE

DUMSK

$595

ED

1.85

-2.44

-2.89

ED

.67

DJBGT

ED

.45

DUMSK

TIMGON

SRHFRQ

last

1.06
.98

-4.57

-2.19

TIMGON

TIMGON

1301

1.11

.04
-.67

DUMSK

$1068

1.59

-.94

$1289

1.40

Y

-

.70

-

1.78

TIMGON

JOBDUR

.73

-1.15

Family

Size

DUMSK

JOBDUR

.44

-.89

Female

Head

RESWGE

Last

Y

SRHFRQ

Last

Y

SRHFRQ

YPC

-1.11

DUMSK

YPC

-

DJBGT

YPC

.69

-

.47

TIMGON

YPC

.00

-

.48

RESWGE

YPC

.37

1 (under

65)

$393

2 (under

65)

491

$536

3

598

570

$630

4

783

811

807

$798

5

939

968

964

956

1096

1116

1108

1100

1064

.28
-1.09

$923

6

-

Note:

$1032

TIMGON
last

is the

job;

1

1397

1381

1393

1345

1317

had

$1253

never

ent

1377

7+

thought

if

he

Census

Derived
and

from
U.

data

published

S. Deportment

of

by

U.

S.

Bureau

of

get

the

a

held

less),

6

search
job

or more.

FEDERAL

RESERVE

BANK

OF

the

than

YPC

RICHMOND

of

not

last

RESWGE

lowest

1 equals

job;

equals
household

and

3

0

is

reservation
income

he

school

earned

power.

or

less;

could
equals
or

SRHFRQ

equals

income

0

thought

ED

(elementary
7

respond-

he

graduate).

wage
per

if

his’
who

actively,

thought

purchasing
months

1 if
one

-1

search;
Y

Those

otherwise.

level

daily

1980
=

sought

(college

Last
1

it;

held

years.

1 if individual

level

summer

respondent
equals

he

chance,

wont

1 is the

.99

10

DUMSK
if

1.74

the

over

job

equals

highest

to

equals

a

a fair

bimonthly.

converted

duration
and

would

frequency;

quently

had

DJBGT
level;

is

=

excluded.
find

he

but

since

5

could

not.

educational

Labor.

elapsed

months;

a job

he

job

time

6-12

thought

he could
Sources:

=

.81

.72

6

is

less

fre-

on

last

JOBDUR

is

=

10

years

in cents

per

hour,

capita.

19