The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
)B VACANCY STATISTICS
HEARINGS
B EFO R E T H E
SUBCOMMITTEE ON ECONOMIC STATISTICS
OF T H E
JOINT ECONOMIC COMMITTEE
CONGRESS OE THE UNITED STATES
EIGHTY-NINTH CONGRESS
SECOND SESSION
M AY 17 AND 18, 1966
Printed for the use o f the Joint Econom ic Committee
U.S. G O VER NM EN T P R IN T IN G O FFICE
63-947
W A SH IN G T O N : 1966
For sale by the Superintendent of Documents, U.S. Government Printing
Washington, D .C ., 20402 - Price 60 cents
Office
JO IN T E C O N O M IC C O M M IT T E E
(Created pursuant to sec. 5(a) of Public Law 304, 79th Cong.)
W R IG H T P A T M A N , Texas, Chairman
PAUL H. DOUGLAS, Illinois, Vice Chairman
SEN ATE
HOUSE OF R EPR ESENTATIVES
R IC H AR D BOLLIN G , Missouri
H A L E BOGGS, Louisiana
H E N R Y S. REUSS, Wisconsin
M A R T H A W . G RIFFITH S, Michigan
T H O M AS B. CURTIS, Missouri
W IL L IA M B. W ID N A L L , New Jersey
R O B E R T F. EL LSW O R T H , Kansas
JOHN S P A R K M A N , Alabama
J. W . F U LB R IG H T, Arkansas
W IL L IA M P R O X M IR E , Wisconsin
H E R M A N E. T A L M A D G E , Georgia
JACOB K . JAVITS, New York
JACK M IL L E R , Iowa
LEN B. JO R DA N, Idaho
J a m e s W . K n o w l e s , Executive Director
J o h n R. S t a r k , Deputy Director
M a r ia n T . T r a c y , Financial Clerk
H a m il t o n D. G e w e h r , Administrative Clerk
E
c o n o m is t s
W il l ia m H . M o o r e
N e ls o n D . M cC ltjng
G e o r g e R. I d e n
D o n a ld A. W e b s t e r (M in o rity )
S u b c o m m it t e e
on
E
c o n o m ic
S t a t is t ic s
W IL L IA M P R O X M IR E , Wisconsin, Chairman
SEN AT E
HOUSE OF R EP R ESE N TA TIV E S
PAUL H . D OUG LAS, Illinois
J. W . FU LB R IG H T , Arkansas
H E R M A N E. T A L M A D G E , Georgia
JACK M IL L E R , Iowa
R ICH AR D BO LLIN G , Missouri
TH OM AS B. CU RTIS, Missouri
R OBER T F. E L LSW O R T H , Kansas
G e o r g e R. I d e n , Economist
n
CONTENTS
OPENING STA T EM E N T S
Proxmire, Hon. William, chairman of the Subcomm ittee on Econom ic
Statistics--------------------------------------------------------------------------------------------------Curtis, Hon. Thomas B., member, Subcommittee on E conom ic Statistics.
Pag*
1
14
C H R O N O L O G IC A L L IS T OF W IT N E SS E S
Cassell, Frank H., Director, U.S. Em ploym ent Service___________________
Ross, Arthur M ., Commissioner of Labor Statistics, U.S. Departm ent of
Labor____________________________________________________________________
Chavrid, Vladimir D ., Director, Office of M anpower Analysis and Utiliza
tion, U.S. Em ploym ent Service__________________________________________
Goldfinger, Nathaniel, director, Research Department, A F L -C I O _______
Creamer, Daniel, manager, Special Projects Department, National In
dustrial Conference Board; accompanied by John G. Myers, senior
economist, Special Projects Department, National Industrial Conference
Board____________________________________________________________________
2
22
59
111
120
A D D IT IO N A L M A T E R IA L S
Cassell, Frank H., Director, U.S. Em ploym ent Service: Prepared state
m ent_____________________________________________________________________
Curtis, Hon. Thomas B., member, Subcommittee on E conom ic Statistics:
Remarks of Arthur F. Burns, president, National Bureau of Econom ic
Research, Inc., and professor of economics, Colum bia U niversity. _
Ross, Arthur M ., Commissioner of Labor Statistics, U.S. Departm ent of
L ab or:
Prepared statem ent___________________________________________________
Selected bibliography on job vacancies_______________________________
The M anpower Situation: Emerging Shortages and Residual Sur
pluses, statement issued with release of “ Summary Em ploym ent
and Unemploym ent Estimates: Decem ber 1965” __________________
Report on M anpower Shortages and Reserves________________________
Report on M anpower Requirements and Supply, M ay 1966__________
Chavrid, Vladimir D ., Director, Office of Manpower Analysis and
Utilization, U.S. Em ploym ent Service:
Prepared statem ent___________________________________________________
Job Vacancy Surveys: Job Vacancy D ata for an A ctive M anpower
Policy, by Louis Levine, Director, U.S. Em ploym ent Service______
Em ploym ent Service Operating Data as a Measure of Job Vacancies,
by Vladimir D. Chavrid and Harold K uptzin ______________________
Indiana Job Vacancy Surveys, by Emily Hawk and Evelyn Toliver__
Portland Studies Job Vacancies, by Wesley E. Zellner_______________
Milwaukee Em ployers Cooperate in Job Survey, by H . J. Jackson.
Progress Report— Job Vacancy Studies, by Norman M edvin and
James H iggins______________________________________________________
Aiken, John H., executive director, Federal Statistics Users’ Conference:
Letter to Senator Proxm ire___________________________________________
Statement submitted by the Federal Statistics Users’ Conference__
Proxmire, H on. William, chairman, Subcommittee on Econom ic Statistics:
Article: Experimental Job Vacancy Survey Program o f the U.S.
Department of Labor, by Irwin F. O. Wingeard, Bureau of Labor
Statistics____________________________________________________________
Letter: Chairman Proxmire to Frank Cassell, Director, U.S. E m ploy
ment Service, requesting additional information for the record____
hi
6
15
27
37
40
42
48
65
77
80
84
86
88
89
97
97
99
109
Myers, John G., senior econom ist, Special Projects Departm ent, National
Industrial Conference B oard: Expansion of statement presented for the
record in conjunction with Daniel Creamer_____________________________
Creamer, Daniel, manager, Special Projects Department, National
Industrial Conference Board: Summary of Opinions of Em ployers and
Com m unity Organizations on Local Uses of Job Vacancy Data—
R ep ort___________________________________________________________________
Article: H elp-W anted Advertising, by Richard Tow ber, N IC B ____
Additional docum entation: Job Orders Outstanding W ith the Em
ploym ent Service___________________________________________________
Additional information supplied by Department of Labor in response to
request of Chairman Proxm ire__________________________________________
Page-
130
145
154
160
172
JOB VACANCY STATISTICS
T U E S D A Y , M A Y 17, 1966
C o n g r e s s o f t h e U n it e d S t a t e s ,
S u b c o m m it t e e o n E c o n o m ic S t a t is t ic s o f t h e
J o in t E c o n o m ic C o m m i t t e e ,
,
Washington D.C.
The subcommittee met, pursuant to notice, at 10 a.m., in room
S-407, the Capitol, Hon. William Proxmire (chairman of the sub
committee) presiding.
Present: Senator Proxmire and Representative Curtis.
Also present: James W. Knowles, executive director; William H.
Moore, senior economist; George R. Iden, economist; Donald A.
Webster, minority economist; and Hamilton D. Gewehr, administra
tive clerk.
Chairman P r o x m i r e . Mr. Cassell, I understand that Mr. Ruttenberg is unable to be here today. We are sorry that he is not here,
but we are glad that he has such able replacements.
I also understand that Mr. Cassell is in one of those unfortunate
emergency situations where he also has a conflict and will have to
leave promptly. That distresses us, because we would like to have
him in on the entire discussion which is often the most fruitful part
of a presentation. But under the circumstances, Mr. Cassell, you
may go ahead.
I see all these statements are long. I think it would be very
helpful if they could be summarized, if possible. If not, perhaps part
of the statements can be read and other parts made a part of the
record. I will try and set a good example by asking unanimous
consent that the opening statement by Senator Proxmire be placed
in the record at this time. It will not be read. It is available to
the press or anyone else who wants to read it.
(The opening statement is as follows:)
OPENING STATEMENT BY SENATOR PROXMIBE, CHAIRMAN OF
THE SUBCOMMITTEE ON ECONOMIC STATISTICS OF THE JOINT
ECONOMIC COMMITTEE
Senator P r o x m i r e . The subcommittee's interest in data on job
vacancies—the subject matter of today’s hearing—goes back at least
4 or 5 years. These hearings are essentially a followup of a recom
mendation made by this subcommittee in our report on “ Employment
and Unemployment” of January 1962. That recommendation reads:
Research should be undertaken directed toward developm ent o f a regular
m onthly survey of jo b opportunities or vacancies to illuminate the demand side
of the labor market in the way the present series measures the supply o f labor.
Experience here and abroad indicates that substantial difficulties must be over
1
2
JOB VACANCY STATISTICS
com e before a statistical series on vacant jobs can become operational but past
success in using survey techniques to solve some other difficult data-gathering
problem s suggests that a useful program may be practical. In any event, the
data from such a survey would be so useful in analyzing labor markets, in operating
^employment services, and developing practical worker training and retraining
programs, that expenditure of some funds on research into this problem would
be warranted.1
The subcommittee is far from being alone in making such a proposal.
As a part of our continuing program for improved statistics for
economic growth, the subcommittee last year asked individual econo
mists, representatives of interested private organizations, and statistics
users for suggestions as to the improvement of Government statistical
programs generally. A number of these people, in responding to the
subcommittee’s request, mentioned that one of the major opportuni
ties for improving our economic knowledge lies in providing statistics
which would show, as currently as possible, the number and types of
job vacancies. Some correspondents cited the usefulness of job
vacancy data in guiding public and private training and retraining
programs; while others suggested that job opening data would give
us a better picture of current opportunities in the labor market.
The point was made that even when unemployment is high, certain
jobs go begging; but we do not have reliable information as to where
they are, what they are, or how many.
The comments received from the individuals and statistics users
were later submitted to the Office of Statistical Standards for comment
and they, in turn, obtained the comments of the Government agencies
concerned with the several suggestions.
After studying these proposals, the Bureau of Labor Statistics last
fall summarized its conclusions about job vacancy data in these words
which should be made a part of the record of these hearings at this
point.
The Departm ent has made pilot studies of the feasibility of collecting job
vacancy inform ation from employers and of the questionnaires and survey designs
■most appropriate for doing so. These studies clearly indicated that such a pro
gram is feasible. The Departm ent is continuing to work on a number of problem
areas, including the task of getting accurate occupational data for the vacancies,
evaluating the data to determine whether jobs are vacant because of the wage
rates or conditions of work offered, and the general problem of developing a survey
system to provide the data at minimum cost.2
There thus seem to be at least three areas of usefulness for such
statistics: (1) in matching men and jobs; (2) in local manpower plan
ning; and (3) in manpower planning at the national level. We will
try to learn something about each of these in the progress of these
hearings.
Chairman P r o x m i r e . Mr. Cassell, you may proceed.
TESTIMONY OF FRANK H. CASSELL, DIRECTOR, U.S. EMPLOYMENT
SERVICE
Mr. C a s s e l l . I would rather paraphrase my statement.
I want to express my appreciation to the committee for the oppor
tunity of appearing here today and having a chance to talk about the
1“ Employment and UnemDloyment,” report of the Subcommittee on Economic Statistics, January
1962, p. 6.
2 “ Improved Statistics for Economic Growth— Comments by Government Agencies on Views Submitted
to the Subcommittee on Economic Statistics,” March 1966, p. 44.
JOB VACANCY STATISTICS
3
valuation and formulation of manpower policy, and the relationship
of job vacancy data to it.
I thought I would go back in time a bit and indicate how my
particular point of view has developed on this subject.
As Director of the Employment Service, I feel a special respon
sibility for the dissemination of information about job openings, and
about the availability of workers which will contribute to the full
employment of our people, and which will help meet the manpower
needs of the country.
I have some previous experiences in this area which I should like
to relate.
In 1961-63, I was chairman of the Governor’s Committee on Un
employment in Illinois. In our report to the Governor we recom
mended the development of job vacancy data on a systematic basis,
because the educators on our committee continually pressed for such
information-----Chairman P r o x m i r e . What w a s the period?
Mr. C a s s e l l . 1961-63.
Chairman P r o x m i r e . That w^as under Governor Kerner?
Mr. C a s s e l l . Yes. And when we recommended the job vacancy
data studies this was almost coincidental with the issuance of the
Gordon report on the same subject.
Our interest came primarily from the educators on our committee,
who persistently asked the questions. What do we train these people
for? How are we going to spend the money? How can we spend it
wisely? Into what kind of occupations should we direct these young
people?
They wanted information that would make their programs more
productive and economical.
This was the focus of our particular interest there. And my present
interest is that we simply will have to have such information for
vocational people, for high school administrators, and for anyone
who is involved in counseling and guidance.
The Governor’s committee felt, furthermore, that we needed the
job vacancy information to enable us to provide the data to facilitate
mobility from one part of the State to another, from one place where
there were no jobs to places where there were jobs. And as you will
recall, in southern Illinois we have had very heavy unemployment.
In northern Illinois we had a shortage of people.
In addition to that, I have been chairman of the Chicago Urban
League’s Education and Guidance Committee. And one of our
jobs there was to work with young people, to try to tell them some
thing about where the jobs were, and where opportunity was.
Without job vacancy data we could not do this job properly. We
simply did not have the job information to give to these young people.
Now, as a former executive of Inland Steel Co., I would have found
job vacancy data especially helpful in planning ahead doing corporate
manpower planning to relieve labor shortages. If I can digress just a
moment, last fall I wrote a paper for the Stanford Business Review
in which I was somewhat critical of corporate manpower planning
because in a large part of American industry manpower planning
simply is very underdeveloped. And it is underdeveloped for the reason
that there are not enough good data upon which industry can base its
planning. One of the most important pieces of information corporate
4
JOB VACANCY STATISTICS
planners need for manpower is job vacancy data. And they need it
over a period of time.
In the absence of such data, the typical thing to do is to not train
and develop people, but to wait around until, hopefully, somebody
comes in the door for employment. The consequence is, when we
are suddenly confronted with shortages, everybody starts to do the
same thing at the same time, training and hiring.
I think from a very special point of view, and in view of my own
past experience, that job vacancy data would be extremely helpful,
and would encourage American industry to do more of its own man
power planning to deal with these kinds of problems.
With this in mind, I would like to say that it seems to me that an
adequate job vacancy survey program, which both Arthur Ross and
Bill Chavrid will talk about, will bring the rifle shot to the subject
rather than the blunderbuss. Essentially, in trying to decide what
kind of programs we need, and how we should spend our money, our
present information is not really as accurate and precise as I think
it should be.
Second, this kind of data would certainly aid the Employment Serv
ice in knowing where to go to find jobs for the unemployed. This
kind of data is very helpful when you are trying to decide which
industries and firms may offer the best employment prospects by
area or by occupation.
It certainly would add to our ability—that is, the ability of the
U.S. Employment Service—to raise the efficiency of our placement
activities. And as I said earlier, it would aid manpower planning
during periods of extraordinary labor shortages or excesses, and by
area or by occupation.
I should mention that just this last week we met with the people
from the Employment Service in Milwaukee, where we have some
very substantial occupational shortages. We spent a great deal of
time discussing what our plans should be to answer the problems
of labor shortages there. And I should say the vacancy studies
which have already been done in Milwaukee were extremely helpful
in enabling that group of people from Milwaukee to do some planning.
This kind of information, of course, would also help us on inter
area recruitment.
For example in 1965 Milwaukee was short of welders, and Portland,
Oreg., had an excess of welders. Now, these are the kinds of imbal
ances that show up in a study like this, and it is important for us to
know this.
Taken together as a series of snapshots over a period of time, job
vacancy studies, it seems to me, can help fill in the picture of occupa
tional needs.
In other words, I think what I am saying is that taking a snapshot
at one time is not adequate. You have got to take a series, to see how
over a period of time the vacancy situation develops, how long the
vacancies exist, and to give us a better picture from an operational
and educational viewpoint as to where opportunity lies.
And certainly this kind of data would supplement the knowledge of
guidance people, counselors, and school administrators as they plan
school efforts.
From an operating standpoint, from the standpoint of operating the
U.S. Employment Service, we need specific data about specific occu
JOB VACANCY STATISTICS
pations and about specific local areas. And this is terribly important
to us in trying to develop this rifle-shot attack on the problems of
employment and unemployment.
This is the kind of data which we are talking about. This is the
kind of data which I should like to see developed in such a way that it
can be used by local communities, such as we did in the Milwaukee
situation.
I think it is terribly important that we enable local employers and
local unions and local communities to intelligently plan their own
programs to meet shortages or excesses. And I think these data will
help.
There are two or three key points about these studies I would like
to mention before I close. First I consider it still an experimental
program, a pilot program, even though the areas surveyed covered
about a quarter of the Nation’s nonfarm employment.
Second, I was pleasantly surprised to learn of the 80-percent em
ployer response to the study. I participated in the early phases of
the Chicago feasibility studies, and wondered myself whether we would
get that high a response from employers. And so I am very pleased
to report that we have a very high employer response, and the em
ployers do indeed cooperate. And as a matter of fact, in this year’s
round of studies apparently employers will be cooperating fully as
well as they did in the first round.
Third, the vacancies covered a very wide variety of occupations,
ranging from nurses to machinists; to engineers, machine operators,
common labor and hospital attendants.
And lastly, with respect to leadtime, these data, I am told by our
people, can be made available in each of the labor market areas in
which we do the studies, practically immediately. It would take
some time to summarize nationally. But the data we accumulate
regionally in the areas can be made available to the people in those
areas for their own planning almost immediately.
Now, the details for the fiscal year 1965 program will be described
by Mr. Vladimir Chavrid, Director of the Office of Manpower Analysis
and Utilization for the USES.
As I said, I still consider these programs experimental. I think we
should continue the job vacancy studies in conjunction with our
USES unfilled openings reports. Present indications are that there is
a close correspondence between total job vacancies and the unfilled
openings in the Employment Service.
In general, the types of openings which the Employment Service
cannot fill readily are representative of the jobs which are hard to fill
in the job market as a whole.
In 1966 we plan to try to identify and quantify the bottleneck
occupations which, if filled, would open up additional jobs. By that
I mean a bottleneck occupation could very well be a machinist. Fill
the machinist job, and then you can fill a whole group of supportive
jobs to the machinist. If you can fill an engineer job, you can fill an
awful lot of draftmen’s jobs under the engineer. These are what I
call bottleneck occupations. If you do not have this one key fellow
you cannot employ a lot of others.
As you know, funds for the expansion of this program were denied
for fiscal 1967 by the House Appropriations Committee.
In view of the favorable results of the pilot projects and the impor
tance of job vacancy data for manpower operations, we hope the House
6
JOB VACANCY STATISTICS
as a whole and the Senate will give further consideration to this pro
gram expansion. By helping to keep abreast of current manpower
developments in specific areas and occupations we believe that job
vacancy data can help contribute to shaping and molding our active
manpower policy to respond meaningfully and promptly to the chang
ing job market conditions as they occur.
I thank you very much for this opportunity to appear here.
Chairman P r o x m i r e . Thank you very much, Mr. Cassell. I
understand you have to leave. You tell us when you want to go.
(Mr. Cassell’s complete statement is as follows:)
P repared
S ta te m e n t
of
F ran k
H.
C a s s e ll,
Se r v ic e
D ir e c to r ,
U .S .
E m p lo y m e n t
I would like to express m y appreciation to this subcommittee for the privilege of
appearing before you today in order that I might discuss briefly the Departm ent
of L abor’s experimental program for the collection and use of data on job vacan
cies. I plan to discuss this program primarily from the standpoint of its value in
the form ulation and administration of manpower policy and programs, and the
activities of the Federal-State Em ploym ent Service system. It is m y under
standing that Arthur Ross, Commissioner of Labor Statistics, will cover the subject
from the viewpoint of job vacancy data as an econom ic analysis tool.
M y comments today are quite different from those which might have been
prepared several years ago when work in the area of job vacancy research, at least
in this country, was almost nonexistent. They are also significantly different
from what we might reasonably be expected to say in the future, when we would
hope to have several years of solid experience with the use of such data as a
foundation and demonstration of definitive accomplishments. The history of the
developm ent of job market information— and this is effectively encompassed in the
past 30 years— reveals at least tw o clearcut trends. One is the increasing require
ments for detailed and specific information— specific in terms of particular com
ponents, groups and localities— for the implementation and administration of
m anpower legislation and programs. Second, we find that as previously un
available information comes to light, new uses and needs emerge. These tw o are
closely inter-related.
I should like to illustrate briefly. The deep national concern with the mass
unem ploym ent of the 1930’ s led to the development of a reliable and current
measure of overall em ploym ent and unemployment. There can be no dispute
that the M onthly Report on the Labor Force provides the single most im portant
and indispensable indicator of the N ation’s econom ic health. But it soon becam e
equally obvious that for manpower operations— whether concerned with warinduced shortages, depressed areas, or human resource developm ent— a good deal
more was needed. A national unem ployment rate of 3.7 percent tells us little
about the situation in Los Angeles, for example, and is even less useful if we are
attem pting to deal with the problems of the Watts area. I becam e most keenly
aware of this several years ago in Illinois, when as Director of Personnel Adminis
tration at Inland Steel, I had the honor to serve as Chairman of the G overnor’s
Com mittee on Unemploym ent, which made an exhaustive study of manpower
and unem ploym ent problems in that State.
On the second point, our Dictionary of Occupational Titles furnishes a con
venient illustration. Originally developed by the Em ploym ent Service during
the 1930’s to provide a com m only understood language for local office matching
o f job applicants and employer job orders, it soon acquired a host of other uses.
These ranged from facilitating the transfer of workers from civilian to warrelated production, to rationalizing the job structure and providing career ladders
within em ploying establishments, to developing counseling and guidance tools,
to identifying the changing nature of jobs in a technologically advancing econom y,
to guiding curriculum and course content in vocational schools.
The im plementation of an active manpower policy is a voracious user of data.
Policy objectives can be stated in broad terms, but manpower operations, if they
are to be effective and efficient, require data and tools to create, adapt, and carry
out programs to resolve problems that have been clearly and specifically defined.
The interest in job vacancy data that arose during the late 1950’s was largely
focussed on an overall national statistic. Since we have no experience with such
JOB VACANCY STATISTICS
7
data in this country, I can only speculate that a series of this type might prove
to be a useful addition to our existing store of economic indicators.
M y concern here today, however, is with vacancy data as they relate to man
power operations. For this purpose, we do already have some experience, limited
though it is, as a pilot venture.
W e have now com pleted one full year of research and analysis into the collec
tion of job vacancy information and are mid-way into another year of experi
mentation. This pilot program, conducted in areas which include one-quarter of
the N ation’s nonfarm wage and salary employm ent, clearly demonstrated that a
large-scale job vacancy program is technically feasible, that employers willingly
cooperate in providing valid data, and that the resulting inform ation will be
useful operationally for implementing an active manpower policy.
Inform ation on job vacancies, with industry and occupational detail, in specific
local areas throughout the country is needed for the effective administration of
many of our principal manpower programs. Such data identify jo b opportunities
in our major employm ent centers, pinpoint industry and occupational manpower
shortages, and guide the establishment of training courses under the M anpower
D evelopm ent and Training A ct and the Vocational Education A ct of 1963. Such
inform ation can also be useful in designing programs to stimulate em ploym ent
growth and reduce unemployment, to find jobs for “ poverty-type” workers, and
if necessary to initiate programs for insuring adequate referral of workers to
essential defense production and high-priority nondefense industries.
Recent developments— both on the domestic scene, as well as abroad— have
served to underline the value at this time of a jo b vacancy inform ation program,
responsive to manpower operating requirements:
1. The identification of skill shortages in certain occupations and industries has
becom e extremely im portant in recent months as the trained supply of workers
continues to diminish. B y providing inform ation on the nature of available jo b
opportunities and the imbalances which exist on a local area basis between the
kinds of workers needed and the skills of available workers, the jo b vacancy pro
gram can be of considerable use to the Em ploym ent Service operationally in fill
ing current openings and in alleviating skill shortages through training, restruc
turing jobs, encouraging relaxation of employer specifications, and special recruit
ment campaigns.
2. The manpower legislation of the 1960’ s requires detailed knowledge o f jo b
opportunities in specific labor areas across the country to provide suitable voca
tional objectives in training or retraining portions o f our work force. This
legislation includes, in addition to the M anpower D evelopm ent and Training A ct
and the Vocational Education Act, the E conom ic Opportunity A ct of 1964— and
the array o f anti poverty programs related to it— as well as the Public W orks and
Econom ic Developm ent A ct o f 1965. In conjunction with other occupational
information tools, job vacancy inform ation can help fulfill the requirements o f
some aspects o f this legislation b y assisting in the design o f im proved programs
for the retraining of workers with obsolescent skills, those who wish to upgrade
their skills, or those who have no marketable skills at all.
3. Some desirable jobs at semiskilled and even unskilled levels are currently
hard-to-fill. These could serve as entry level-jobs for “ p overty” group workers
who lack requisite skills and education to meet qualification standards for higher
level jobs. Considerable interest in such jo b opportunities identified b y the
vacancy surveys has com e from the Office of E conom ic Opportunity, the President’ s
Commission of Civil Rights, and Plans for Progress employers, all of whom are
attem pting to fit disadvantaged workers into productive and satisfying em ploy
ment.
4. Comprehensive inform ation on job opportunities, b y occupation and area
can help eliminate pockets of unem ployment which exist because of lack of skills,
géographie isolation, cultural disadvantages, and other obstacles to the matching
of workers and jobs. It provides the raw data on jo b prospects needed by
the Em ploym ent Service to encourage worker mobility, and to provide information
useful in counseling younger workers and students and guiding them toward
occupational choices that provide better prospects for employm ent.
5. The econom y has already entered a period of manpower stringency, although
shortages are still of the “ spot” variety rather than nationwide in scope. If this
trend continues, and if the Viet-Nam conflict makes further demands on the
econom y, we may need to give further emphasis to ways and means of identifying
the industries and areas experiencing the most pressing manpower shortages,
to methods which employers can use to facilitate the elimination of manpower
bottlenecks, and to manpower programs needed to insure the most effective de-
8
JOB VACANCY STATISTICS
velopment and use of human resources. The job vacancy program provides an
important data resource for these purposes.
BACKGROUND OF THE EXPERIMENTAL PROGRAM
The U.S. Em ploym ent Service— and the affiliated State E m ploym ent Services—
has long been aware of the value of job vacancy data as an im portant operating
tool. As long ago as W orld War II and again during the K orean conflict, such
data, by occupation and industry, were collected in the m ajor metropolitan
areas by the State Em ploym ent Services to identify and alleviate manpower
shortages. Although liiost of these programs were discontinued*when hostilities
ended, some job vacancy data have been collected since that time on a voluntary
basis by 15 State agencies, although, in all but one instance, without any occupa
tional breakdown.
The inauguration of our experimental job vacancy program is, to some extent,
a response to the findings of the President’s Com mittee to Appraise Em ploym ent
and Unemploym ent Statistics (the “ G ordon Com m ittee” )- In its report to the
President in September 1962, the Com mittee listed job vacancy data as one of
the most conspicuous gaps in our present system of collecting em ploym ent and
unem ploym ent information.
Alm ost coincidentally with the release of the G ordon Com mittee report, the
report of the Illinois Governor’s Com mittee on Unemploym ent which I chaired
was released in January 1963. One of the recommendations of our report too,
was for the initiation of a program to “ measure volume and com position of job
vacancies and to report such information on a regular basis.” I understand
that this recom mendation was one of the m ajor factors in the selection of the
Chicago area for the first feasibility study made, by the USES to test the
possibility of collecting job vacancy data from employers.
This study was initiated in 1963, with the cooperation of the affiliated Illinois
Em ploym ent Service. Another feasibility study was conducted in the spring of
1964 in Buffalo, sponsored locally by the New Y ork Em ploym ent Service and
the local Chamber of Commerce. The results of these preliminary studies were
highly encouraging. A t the same time, other research by the U.S. Em ploym ent
Service and the Bureau of Labor Statistics helped to define technical concepts
and m ethodology, and investigated the collections and uses of job vacancy
inform ation in other countries.
A t the direction of Secretary of Labor W. Willard Wirtz, and with the endorse
ment of the President, we began an experimental job vacancy program in the
fall of 1964 to test the possibility of collecting such data on a relatively wide
scale, and their usefulness for manpower operations. Primary emphasis was
upon the value of the data for facilitating the operations of the job market, with
a statistical program as a possible by-product. The USES, in cooperation with
its affiliated State Em ploym ent Service agencies and the Bureau of Labor Sta
tistics, conducted two comprehensive job vacancy surveys in each of 16 metro
politan areas representing a broad cross section of American industry in terms
of size, industrial characteristics, geographical location, and nature of unem
ployment. These areas include: Baltimore, Birmingham, Charleston (S.C.),
Charleston (W . Va.), Chicago, Kansas City, Los Angeles, Miami, Milwaukee,
Minneapolis-St. Paul, New Orleans, New York, Philadelphia, Portland (Ore.),
Providence, and Richmond. In the aggregate, these areas account for approxi
mately one-fourth of the N ation’s total nonfarm employment. The first round
of surveys was conducted during the last quarter of calendar year 1964 and the
second survey took place in April 1965.
A detailed description of the fiscal year 1965 program and its findings will be
presented by Mr. Vladimir D. Chavrid, Director of the Office of M anpower
Analysis and Utilization of the USES. M y comments will be confined to
summarizing some of the m ajor findings of this research.
MAJOR FINDINGS
One of the most significant and conclusive results of the experimental program
was the demonstration that valid job vacancy inform ation b y detailed occupation
can be collected from a large sample of employing establishments. Approxi
mately 80 percent of the nearly 20,000 employers sampled responded to the survey
questionnaires. As a result of additional experimentation in some areas, it was
demonstrated that employers would also provide information on wage rates offered
for the vacant jobs and on the number of vacancies for part-time or tem porary
JOB VACANCY STATISTICS
9
jobs. Although sthe higfy degree of em ployer cooperation in the fiscal year 1965
surveys occasioned some surprise, it bore out very closely the pre-test of feasibility
conducted in 1963 for Chicago.
A bout one out of every four employers responding to the 1965 surveys reported
at least one vacancy. As expected, the proportions of employers reporting
vacancies varied by size of establishment. A bout one-half of the large establish
ments had some vacancies, while only 1 in every 10 smaller establishments reported
at least one vacant job.
The 1965 experimental program also revealed that the Em ploym ent Service
plays a far more im portant role in the total jo b market than was previously
believed. Approxim ately 30 percent of the job vacancies, excluding domestic
service, were already listed with the Em ploym ent Service. M oreover, the dis
tribution among broad occupational groups was reasonably similar for total job
vacancies and unfilled Em ploym ent Service job openings. This correspondence
was apparent, though to a lesser extent, in many of the more detailed occupational
categories as well. M ore im portant, however, was the fact that the bulk of the job
vacancies and of the ES unfilled openings were concentrated in the same occupa
tions, except in certain skilled categories where hiring is norm ally done through
union hiring halls or similar channels. This suggests that, in general, the type of
openings which the Em ploym ent Service cannot fill readily are representative of
job s which are hard-to-fill in the job market as a whole.
This finding is being further confirmed by some of the preliminary results now
becoming available in our F Y 1966 program. In this survey, we asked employers
to indicate whether they wished Em ploym ent Service assistance in filling their
jobs. In the first 9 areas reporting on this subject, 60 percent of all employers
answered affirmatively. In none o f the areas did the proportion fall below 50
percent.
The 1965 surveys also demonstrated that with respect to vacancies, as with
other job market measures, the labor demand-supply differences among areas are
very substantial. While the combined data for the surveyed areas provided much
im portant information, the data in this form are not particularly useful for opera
tional purposes because o f these area differences. For instance, the number of
secretarial vacancies in all areas combined was in approximate balance with the
number of applicants available. However, on an area basis, a large shortage o f
secretaries in New York City and a smaller shortage in Richm ond were balanced
by surpluses o f such workers in the remaining areas. For welders, the combined
data indicated a surplus o f workers, although in Birmingham, Charleston (S.C.),
and Milwaukee, these workers were in short supply.
Vacancies in the surveyed areas were found in a broad spectrum of occupations,
although the heaviest demand was concentrated in a relatively narrow band.
For example, openings for nurses accounted for 1 out of every 4 professional va
cancies, while jobs for salespersons, clerks, and for stenographers and typists
accounted for two-thirds of the vacancies in clerical and sales. Over half of the
job vacancies in the service occupations were for waiters and waitresses, kitchen
workers, practical nurses, and hospital attendants. One-fifth of the vacancies for
unskilled workers were in warehousing occupations. Please remember, however,
that this reflects the situation as it existed a year to 15 months ago. Even then,
however, openings for skilled workers were more widely dispersed. Vacancies in
each area tended to reflect its own industrial pattern.
A surprisingly large proportion of the job vacancies a year to 15 months ago—■
about half of the total— were relatively hard to fill, as indicated by the fa ct that
they had been open for one month or longer. The identification of these hard-tofill jobs, by occupation, in relation to total vacancies, promises to be one of the
most significant contributions of the program, particularly to identify occupations
for which training courses should be instituted. Over a period of time, such data
could becom e one of the more valuable operational tools available to the E m ploy
ment Service, in addition to providing vitally needed inform ation for vocational
guidance and vocational education.
In some areas, the survey requested inform ation on wages for the reported
vacancies. Analysis of the reported data indicates some of the vacancies were
for jobs at wage rates below those normally offered for such occupations in the
areas involved. Surprisingly, most of these below average offers were for pro
fessional and managerial occupations and for clerical and sales jobs. While
precise data are not available in the absence of prevailing wage inform ation for
each occupation in each area, comparisons of the jo b vacancy data with wage
offers for Em ploym ent Service jo b openings and similar data suggest that around
10
JOB VACANCY STATISTICS
15 to 20 percent o f the vacancies for which wage data were reported could be
considered as offering substandard wages.
We attem pted to collect inform ation in a few areas on part-time and tem porary
vacancies. However, the inform ation obtained with respect to this group was too
lim ited to support conclusions with respect to the part such vacancies play in the
jo b market. Because there may be considerable econom ic im portance in connec
tion with the part-time and tem porary employm ent opportunities available in an
area, particularly as they apply to older and younger workers and to women, this
aspect of the program is being re-examined in our fiscal year 1966 program cur
rently in progress.
Another significant finding of the fiscal year 1965 surveys was that the level o f
jo b vacancies apparently fell far short of the total number o f unem ployed in the
areas, and was even well below the number of jo b applicants currently registered
with the Em ploym ent Service.
There were very wide fluctuations among the areas, however, in the ratio of
jo b applicants to vacancies, ranging from an approximate balance in Richm ond
and Milwaukee to a 10 to 1 relationship in Charleston, W . Va.,— a long-time area
o f high unemployment. N ot too surprisingly, however, preliminary results from
the F Y 1966 surveys show a sharp increase in the number of vacancies this April
as compared with those in April 1965.
The vacancy surveys also docum ented in greater detail structural unem ployment
problem s in some areas. The types of workers needed were primarily in the higher
skills, while the workers available were largely concentrated in the lower skill
levels. In the professional and managerial classifications, aggregate vacancies
exceeded the number o f available applicants, while in clerical, sales, and skilled
occupations, applicants outnumbered vacancies by about 2 to 1. In the service,
semi-skilled and unskilled groups, the applicant-vacancy ratio was weighted
heavily on the applicant side.
USES OF THE DATA
Probably the most extensive operational use of the information collected on job
vacancies was in connection with manpower training programs. Vacancy in
form ation was instrumental in many of the pilot areas in identifying occupations
for which training could be established under M D T A , OJT, vocational education
and in special programs aimed at disadvantaged youth. The data also proved
useful to the local ES offices for counseling and in job developm ent f o r ViA.rri-toplace applicants.
Another important application of job vacancy information is its use in connec
tion with antipoverty programs through the identification of the hard-to-fill jobs
in the lesser skilled occupations. Of the vacancies for unskilled and service
workers uncovered in the first survey, about half had been available for at least a
month. Im portant among these were vacancies for laborers in the metalworking,
transportation equipment, warehousing, and construction industries, practical
nurses, hospital attendants, porters, waiters and waitresses, and kitchen workers.
Nearly two-thirds of the vacancies for semiskilled workers had been available 30
days or longer. Am ong these were occupations in the machine shop, construction,
and textile manufacturing industries, as well as openings for chauffeurs, drivers,
and routemen. An analysis of the wage data for these lesser skilled occupations
did not indicate any concentration of wage offers below the prevailing wage rates
for the areas.
While job vacancy data proved to have many important operational applica
tions, the first year’s test did not provide any conclusive findings as to the degree
to which the program might help increase the placement potential of the public
Em ploym ent Service. Concentration on the technical and mechanical problems
of getting this comprehensive experimental program underway within a short
period limited the ES staff time available for obtaining jobs orders from employers.
In addition, many of the vacancies reported b y employers were already on file
with the local Em ploym ent Service and placements made to these jobs were not
attributed to the job vacancy program. As I noted earlier, more than 60 percent
of all surveyed employers requested Em ploym ent Service assistance in filling
their job vacancies in those areas where returns are in under area F Y 1966
programs.
FISCAL YEAR 1966 PROGRAM
Since no additional funds were made available for this program in fiscal year
1966, current year plans are designed to im prove the technical aspects of the
program and to give particular emphasis to its operational usefulness. The first
JOB VACANCY STATISTICS
11
survey under this year’s program just took place about a m onth ago, and returns
are now beginning to be tabulated.
In our present program, the sample of surveyed employers is being im proved
by increasing the representation of smaller firms and by im proving the indus
trial distribution. The survey form itself was im proved, and provision was made
for the collection o f wage data in all surveyed areas. As I mentioned, the ques
tionnaire now also asks the employer to indicate whether he desires local
Em ploym ent Service assistance in filling vacancies.
If time and resources are available, w~e hope, under our fiscal year 1966 pro
gram, to attempt to identify and quantify vacancies in “ bottleneck” occupations,
that is, vacancies which , if filled, would open some additional number o f jobs.
This research may be done in tw o labor areas, one having a high vacancy rate
and the other a low vacancy rate. We also plan further studies of the relationship
between jo b vacancies and ES unfilled openings and jo b applicants.
In addition, the possibility of using job vacancy data to increase worker m obility
between various labor areas through the matching of occupational surpluses and
shortages will be explored. Because of current skill shortages, it is im portant to
identify those occupations in which workers are in short supply and the areas in
which a surplus of such workers exists so that critical manpower shortages may be
alleviated by encouraging these workers to move. Finally, we anticipate that we
can achieve some im provements in the use of job vacancy inform ation in our
own manpower training operations, but also in assisting employers in the surveyed
areas in planning their own internal staff training and prom otion policies in the
light of immediate and anticipated occupational shortages.
Funds for program expansion were denied for F Y 1967, b y the House Appro
priations Committee. In view of the favorable results of the pilot projects and
the im portance of vacancy data for manpower operations, we hope that the
House as a whole, and the Senate will give further consideration to this program
expansion. B y helping to keep abreast of current manpower developments in
specific areas and occupations, we believe that job vacancy data can help con
tribute to shaping and molding our active manpower policy to respond meaning
fully and prom ptly to changing job market conditions as they occur.
Thank you for this opportunity to outline m y views on this im portant program.
Chairman P r o x m i r e . One of the questions that has troubled me
concerns the usefulness of job vacancy data over a period of time.
You say that the job vacancy data would help corporate planning as
well as help in many other ways. If data should reveal, for example,
that as of today—mid-May 1966— there are vacancies in certain
areas, how useful would that be in terms of making corporate plans
to train personnel or hire personnel that could not possibly show much
results before 1967? I realize that you say that there are patterns
that develop by studying this situation, but even if we have had data
on such changes in the past I wonder if on the basis of your great
experience you find that this would be useful?
Mr. C a s s e l l . I will just turn it around as if I am doing corporate
manpower planning.
The kind of information I need as a corporate manpower planner
is information on a steady, consistent basis, so that I can watch
trends over a period of time, and so that I can plan ahead to meet
the leadtimes on the various kinds of occupations. In other words,
if these data had been coming along, let’s say, for the last 5 years,
there would have been little excuse for me not anticipating that
there would be a shortage of engineers, or a shortage of welders,
or a shortage of this or that. One would only have to look at the
data over a period of time to start doing his own training planning
ahead.
A continuing picture would enable you to anticipate the shortage.
If you don’t have this information—-and it is kind of a hard thing
to describe—you always think that people are going to turn up.
If you don’t have the data you are sitting there and saying, “ Well,
12
JOB VACANCY STATISTICS
I am not going to spend the money on training, these people are
going to turn up.”
But if the reports keep coming in quarterly, saying, “Look, these
occupations are short, they are very short, and we can’t fill them
in less than 30 days, and then I begin to say, “ Well, I had better do
something about training,” or, “ I had better do something about
changing the nature of the job,” or “ I had better do some upgrading”
or some such thing internally, because it has become clear to me
that this is a shortage occupation. And I think the existence of the
data enables an employer, looking at this, to become very realistic
about what his chances are of going into the market. His only other
alternative is what he is doing today. We bid against each other
for the shortages.
Chairman P r o x m i r e . There is another problem, you bid against
each other for the shortages——
Mr. C a s s e l l . So that you drive up the price, and then very fre
quently—as is the case in Milwaukee and other cities—the people
simply are not there anyway, because we did not do the planning
with the leadtime.
Chairman P r o x m i r e . What is the nature—what is the way, if you
could describe it in a minute or two, that this data is actually secured?
Obviously, if you don’t use want ads and newspapers, you go around
to a certain proportions of corporations, including all the big ones, and
a sample of the small ones?
Mr. C a s s e l l . Are you speaking of this particular study, or the way
I would do it?
Chairman P r o x m i r e . What do you think would be the most effi
cient, the lowest cost way of getting useful and valid sound data?
Mr. C a s s e l l . I think this question is going to be answered later
by both Mr. Ross and Mr. Chavrid.
Chairman P r o x m i r e . I think that was something on which you
could help us, because you have had a great deal of experience in
management.
Mr. C a s s e l l . From a management standpoint a great deal of the
data I have had to work with was inadequate. Some of the analyses
of the newspaper ads are helpful for particular occupational levels. It
is not something we should disregard.
But by and large, in the absence of systematic gathering of job
vacancy data, an employer is pretty largely in the dark. He can call
up another employer but about all they do is commiserate with each
other on their shortages. But he does not know easily what is happen
ing in another region, another area of the country. And one of the
important things about this, too, is that hopefully as we learn how
to deal with the mobility problem-----Chairman P r o x m i r e . But actually, what you would do would be to
go to the employment office of a corporation, and ask how many
people they plan to hire. Do you feel that on the basis of your experi
ence you could get useful statistics that would be valid at a reasonable
cost?
Mr. C a s s e l l . Well, on the basis of the studies as they are being
conducted now, as Dr. Ross and Mr. Chavrid will describe, I think
at this point in time that is a reasonable assumption. I think all the
time in working with data like this, one wants to work always to
simplify, and always to economize. And I would think at this point
JOB VACANCY STATISTICS
13
in time what I have seen is good and useful, but that is not to suggest
that at a later time, as we learn to develop these data and use it, we can
simplify it and hopefully reduce the cost of gathering it.
But the real point is that it simply does not exist for the employer at
this point in time. It is something he does not have.
Chairman P r o x m i r e . Y o u have made a strong case that the em
ployer would want this information and the management would want
it. As I understand, organized labor has not taken a position on this,
and that was one of the reasons why the House has not acted favorably
here? Is there any reason why organized labor would, in your judg
ment and on the basis of your experience, be interested in this? I
should think it would serve their interest as well as management’s.
Mr. C a s s e l l . I am not in a position to judge what the motivations,
the views of organized labor are. But I would say, on the basis of
what I have said here today, that the program is feasible and that the
operational uses of the data to put people to work or to alleviate
labor shortages make the cost well worth incurring. My viewpoint
comes from the fact that I am running an operating organization.
My job is to match people and jobs, get them into jobs, help em
ployers overcome shortages. And I think the unions certainly should
be sympathetic to that particular point of view; namely, eliminating
unemployment wherever we can eliminate it, upgrading people,
training people to fill the shortages. So I cannot understand what
is in their minds. My guess would be that my viewpoint—which is
strictly operating, trying to do the job we have been assigned—
certainly ought to be consonant with the general views of labor people
who ought to be interested in this same thing.
Chairman P r o x m i r e . H o w long would it take to summarize
national data? You said it would take longer than the regional data.
Mr. C a s s e l l . I think I had better leave that to Mr. Chavrid or to
Mr. Ross. Say about 2 or 3 months?
Mr. Ross. Weeks.
Chairman P r o x m i r e . Two or three weeks?
Mr. C a s s e l l . I would like—let me make a guess here—a few weeks
or longer depending on the amount of occupational and area detail
we want. We have not yet gotten to the point where we are able to
determine this. But certainly it would be available very quickly at
the local basis.
Chairman P r o x m i r e . This notion of bottleneck occupations is cer
tainly a very useful concept. Is it sufficiently understood; is there
enough experience with it, so that this can form the basis of valid
data? You said, for example, that if a company succeeds in finding
an engineer you will get five or six other jobs.
Mr. C a s s e l l . Y o u can put draftsmen to work then.
Chairman P r o x m i r e .; tinder the circumstances, I should think you
would not get an adequate response unless you have a skillful inter
rogator. You go to management and ask, “ Do you need draftsmen?”
and they will say “ No” if you don’t discover that they need engineers
first.
Mr. C a s s e l l . Yes. I think this is going to have to be ground into
the program. This is a relatively new concept, and it is one that is
going to be looked at during the coming year.
From a management concept it is an old concept. But the point
is that we have never done anything with it analytically. Instinctively
63 - 9 47— 66 ------------ 2
14
JOB VACANCY STATISTICS
you know in your own organization the bottleneck occupations; you
can name them, and you know why you cannot hire people, because
of the bottlenecks.
Chairman P r o x m i r e . A lot of this depends on judgment, the
psychology at the same time, to that extent it is less firm than the
other statistics where you ask a person if he is looking for work, or
you ask a question that can be determined on the basis of a somewhat
more objective, factual reply?
Mr. C a s s e l l . Well, I happen to have been on the other end of
this when they were attempting to determine what the definitions
would be, and whether all of this was possible. And I, along with a
number of other companies in Chicago, was a guinea pig. And
my feeling from the other side was that these are pretty good data.
They represented what we were thinking and what we felt we wanted.
So I would say, I think by and large, they are pretty reliable.
And it comes from the fact that I had to answer the questions from the
other side.
Chairman P r o x m i r e . Y o u worked with it?
Mr. C a s s e l l . Yes.
Chairman P r o x m i r e . My final question is, could additional funds
be better spent on the present interarea placement program, if you
had to choose between them?
Mr. C a s s e l l . I am not about to choose. I need this—we all
need this kind of data to help the interarea program. I think they
are related and are necessary.
Chairman P r o x m i r e . Congressman Curtis?
Representative C u r t i s . Thank y o u , Mr. Chairman.
And first I would like to have put in the record an opening statement.
Chairman P r o x m i r e . Without objection, it may appear at this
point.
(The opening statement of Representative Curtis is as follows:)
OPENING STATEMENT OF REPRESENTATIVE THOMAS B. CURTIS,
MEMBER OF THE SUBCOMMITTEE ON ECONOMIC STATISTICS
OF THE JOINT ECONOMIC COMMITTEE
Mr. C u r t i s . I want to congratulate the chairman for his initiative
in arranging these hearings on the need for job vacancy statistics and
the progress that has been made in developing them.
As the chairman knows, I have felt for some time that information
on job vacancies is a vitally necessary tool of economic analysis. My
interest goes back to 1961 when, under my direction, the House
Republican policy committee conducted a far-reaching study en
titled “ Employment in the Dynamic American Economy.” That
study clearly indicated the importance of data on unfilled jobs, par
ticularly in determining the proper focus of training and retraining
programs. In remarks on the House floor in connection with that
study, in my book, “ 87 Million Jobs,” and in countless hearings and
reports since then I have tried to do whatever I could to point out the
need and the desirability of quick action by the administration to
develop these data.
It is regrettable that it has taken nearly 4 years to begin the difficult
task of developing this statistical series. Thanks to the work of this
subcommittee and to the pioneering efforts of the National Industrial
JOB VACANCY STATISTICS
15
Conference Board and the National Bureau of Economic Research,
we are finally beginning to move forward. I will be greatly interested
in the progress report of the administration witnesses and their future
plans for developing this series.
These hearings may seem remote from the pressing policy problems
that now confront us. But I believe that the long-run implications of
this work will represent a major step forward in our understanding of
how the economy operates and the appropriate steps required to
maintain maximum employment and growth. The vast influence of
our unemployment and price statistics testify to the importance
which statistical data can have in shaping Government policy. In
my opinion, job vacancy statistics will be no less important.
I am sorry that Prof. Arthur F. Burns, who has been an articulate
and consistent supporter of job vacancy data, is in Europe and unable
to be here for these hearings. No one has done more to clarify the
policy uses which such data would have. As Dr. Burns has pointed
out, what really matters for the purposes of the Employment Act is
how the amount of unemployment actually compares with the number
of job vacancies. If we were able to compare unemployment with
unfilled jobs, we would know with greater certainty than is now
possible whether aggregate demand is deficient or whether structural
correctives are required to deal with unemployment.
In order to make Dr. Burns’ views on this subject available, I ask
that a copy of his speech before the Joint Economic Committee’s
symposium on the twentieth anniversary of the Employment Act of
1946, held on February 23, 1966, be included in the record of these
bearings at the conclusion of these remarks.
Chairman P r o x m i r e . Without objection, it is so ordered.
(The speech referred to follows:)
R e m a r k s 1 o f A r t h u r F. B u r n s , P r e s i d e n t , N a t i o n a l B u r e a u o f E c o n o m i c
R e s e a r c h , I n c ., a n d P r o f e s s o r o f E c o n o m i c s , C o l u m b i a U n i v e r s i t y
aggregate
or
str u ctu ral
approaches
to
a c h ie v in g
em ploym ent
act
o b j e c t iv e s
The Em ploym ent A ct which we celebrate today has had its share of the vicissi
tudes of fortune that go with life. The bill originally proposed by Senator M urray
ran into massive opposition in the House, and many anxious months elapsed
before the Congress hammered out an acceptable compromise. The machinery
established by the A ct has not always functioned sm oothly or as its designers may
liave hoped. A t times, the findings by the Council of Econom ic Advisers have
lacked the detachm ent or the lustre of science. A t times, the pronouncem ents of
the Joint Econom ic Com mittee have suggested excessive partisanship or haste.
In one year the Congress refused to vote a full year’s appropriation for the Coun
cil's activities, and its ability to survive became doubtful. Despite such occa
sional setbacks, the moral authority of the Em ploym ent A ct has grown with the
passage of time. Indeed, in the span of a mere twenty years, the A ct has acquired
the force of an economic constitution. The President, his Council of E conom ic
Advisers, the Congress, in some degree the entire executive and administrative
.establishment, including the Federal Reserve Board, now function under this
“ constitution” when m ajor econom ic policies are developed.
As befits a constitution, the Em ploym ent A ct lays down general principles and
procedures, but gives little guidance on how the Federal government is to dis
charge its new responsibility of prom oting “ maximum employm ent, production,
and purchasing pow er.” T o be sure, the A ct stresses the im portance of proceeding
“ in a manner calculated to foster . . . free com petitive enterprise.” This con
straint reaffirms our nation’ s com m itm ent to the principle of freedom, but it does
i Hearings before the Joint Economic Committee on “ An Economic Symposium on the Occasion of the
Twentieth Anniversary of the Employment Act of 1946,” on Feb. 23,1966.
16
JOB VACANCY STATISTICS
no more than that. The A ct also specifies that the means em ployed in furthering
its objectives must be consistent with "the “ needs and obligations” of the Federal
government, with “ other essential considerations of national p olicy ,” and with
“ the general welfare.” In view of this broad language, our successive presidents
have been able to deal under the umbrella of the Em ploym ent A ct with such
objectives of policy as stability of the general price level, faster im provem ent o f
productivity, equality of opportunity, and equilibrium in the balance of payments.
However, the A ct itself is entirely reticent on these matters, and therefore gives
no clue to the way in which any of these objectives is to be sought, or how the
pursuit of one or another of them m ay aid or limit the achievem ent of “ maximum
employm ent, production, and purchasing pow er.” In short, the A ct practically
leaves the means for dealing with recession, unem ployment, or inflation to ju dg
ment concerning the individual case.
The flexibility inherent in the Em ploym ent A ct has proved very helpful to
governm ent officials charged with its administration. Indeed, econom ic life is
so full o f surprises that it is doubtful if the A ct could have survived if the Congress
had prescribed some formula, whether the one suggested by the Murray bill or
any other, for achieving maximum employment— to say nothing of maximum
production or purchasing power. A t the same time, the sweeping but imprecise
mandate of the A ct has im posed an extremely difficult task on the Council o f
E conom ic Advisers and the Joint Econom ic Committee, and beyond them on
professional economists as a class.
Taking the past twenty years as a whole, the administrators of the Em ploym ent
A ct have concentrated on the maximization of employment, bu t they have n ot
neglected other major objectives of national policy. B y and large, our econom y
has perform ed well during this period. W e have preserved the essentials o f
freedom in a revolutionary age, when many other nations have lost or destroyed
their freedom. Our econom y has continued to grow in size and efficiency. W e
have made great strides in moderating the business cycle, and the fruits of industry
have been widely distributed among our people. The Em ploym ent A ct has
contributed to these achievements by introducing elements o f order into econom ic
policy-m aking and by providing assurance to both businessmen and consumers
that econom ic storms would n ot be left to themselves. W e must not, however,
gloss over the lapses from full employm ent during the post-war period, or the
series of recessions, the deterioration in the value of the dollar, the chronic deficit
in the balance of payments, and the persistence of pockets of poverty in our land
o f plenty. I f the efforts of the administrators of the Em ploym ent A ct have n ot
always been successful, the reason in large part is that they have worked with
tools that are much too crude.
W e need, in particular, better ways of determining whether, when or to w hat
degree unem ploym ent can best be attacked by over-all monetary and fiscal
policies. Our nation has relied preponderantly on such policies, during the past,
few years on the ground that aggregate demand was deficient. This approach
has certainly not been wanting in plausibility. In view of the fact that we ex
perienced a recession in 1957-58, that the recovery which followed was incom plete,
that another recession occurred in 1960-61, and that a good part o f 1962 was
marked b y sluggishness, there can be little doubt that a deficiency of demand was
a major cause of unem ployment during much of the period since 1957. However,
the Council’ s calculations of the gap between actual and potential output, quite
apart from being fragile, cannot be treated as measures of demand shortage. If
aggregate output falls short of its potential, the gap may have nothing to do with
any weakness of demand. It may instead reflect obstacles on the side of supply
or a failure of the constituent parts of demand and supply to adjust sufficiently
to one another. Since the structure of our econom y keeps changing, these changes
as well as difficulties on the demand side must be reckoned with in a scientific
diagnosis.
Let me note briefly a few of the structural factors. First, welfare programs have
grown very rapidly in recent years. A great merit of these social measures is
that they maintain a flow of income during periods of unemployment, so that
even poor men m ay practice some of the discrimination in job choosing that
comes as a matter of course to the well-to-do. Our statistical measures, however,
do not recognize this voluntary aspect of unemployment, nor the fact that our
social legislation together with increasing prosperity have been tending to increase
it. Second, women and teenagers have become a much larger factor in the labor
force since the late fifties. But women are less inclined or less able than men
to end their unem ployment by taking a job in another city. Indeed, they are
JOB VACANCY STATISTICS
17
less prone than men to m ove to another occupation or another firm within the
city of their residence. M oreover, married women com m only seek only part-time
or intermittent work. And since a large proportion of the teenagers in the labor
force are students, they too frequently seek part-time or intermittent work.
B ut a new entrant into the labor force rarely finds or takes a job im mediately; in
other words, he is unemployed for a time. Since these unavoidable intervals of
unem ploym ent are repeated for intermittent workers, the volum e of unem ploy
ment has tended to rise as the intermittent work force has grown. Third, the
obstacles to rapid adjustment in the labor market have lately become larger.
T h e pace of technological change has quickened* The -supply of part-time
workers has increased with stidden rapidity, while the evolution of demand has
been gradual. Also, the legally prescribed minimum wage has risen much faster
than the average wage at the very time when the ranks of unskilled and inexperi
enced workers were swelling. Hence, shortages of some types of labor and in
some communities have coexisted with surpluses in others to a larger extent
than before.
It is developments such as these that the structuralist school has emphasized
rather than any deficiency of aggregate demand. And just as the expansionist
school has sought to fortify its claims by an impressive array of evidence, stressing
in particular the depressed state of business investment in fixed capital between
1957 and 1963, so the structuralist school has marshalled considerable evidence
on the high and rising level of overtime work, on the concentration of unem ploy
ment among less educated workers, on the jum p in the ratio of the unem ploym ent
rate of Negroes to that of whites since the mid-fifties, on the exceptionally high
ratio of the unem ployment rate of teenagers to that of adult males during the
past three years, and so on. M ost structuralists have been entirely ready to
grant that easy money, lower tax rates, and larger Federal, expenditures— that is,
the remedies favored by the expansionists— would reduce unemployment. They
have insisted, however, that more lasting effects would be achieved by attending
to the structural causes of unemployment, and that the risk of inflation would
also be reduced in the process. Although their views were heeded to some
degree, as the M anpower D evelopm ent and Training A ct of 1962 and related
legislation testify, the expansionist theory proved more congenial to the m ood
of our times. Had it done so to a lesser degree, I believe that unemployment
would now be no higher while the danger of inflation would be smaller than it
has become.
M y purpose on this occasion, however, is not to press the relative merits of this
or that school of economic thought. M y basic point is rather that existing in
form ation has prevented econom ic investigators from reaching the precise diag
nosis of the unem ployment problem that the Em ploym ent A ct so plainly requires.
The A ct declares that the Federal government has the responsibility of promoting
“ conditions under which there will be afforded useful em ploym ent opportunities
* * * for those able, willing, and seeking to work.” T o discharge this responsi
bility, statistics are needed to determine to what degree, if any, the aggregate
demand for labor falls short of the number of “ those able, willing, and seeking to
w ork” — that is, of the supply of labor. But the aggregate demand for labor in
cludes the unfilled jobs as well as those that are being manned, just as the aggre
gate supply of labor includes the unem ployed workers as well as those who have
jobs. Hence, to determine the relation between aggregate demand and supply,
information is needed on three magnitudes— employment, unemployment, and
job vacancies. Unhappily, while we have comprehensive statistics on the first
and the second, the data on job vacancies are fragmentary, and it has therefore
been impossible to bring either the expansionist or the structuralist theory to a
decisive test.
If I read the Em ploym ent A ct correctly, its implementation requires con
tinuous, carefully compiled, and comprehensive statistics on job vacancies. It
may be interesting to know whether the existing unem ploym ent rate is above or
below 4 per cent, but neither this conventional figure nor any other can be relied
upon to identify maximum employm ent— or its equivalent in com m on usage, full
employm ent. W hat really matters for the purposes of the Em ploym ent A ct is
not what figure on unem ployment appears to correspond best to the concept of
full employment, but how the am ount of unem ployment that actually exists
compares with the number of job openings. When unem ployment exceeds job
vacancies at prevailing wages, the demand for labor is clearly insufficient to pro
vide em ploym ent for everyone who is able, willing, and seeking to work. A t
such a time, a deficiency of aggregate demand exists, and a governmental policy
that relies on monetary, fiscal, or other devices to expand demand is, in principle,
18
JOB VACANCY STATISTICS
suited to the nation’s needs. On the other hand, when the number of vacant job s
is equal to or larger than the number of the unemployed, there is no deficiency of
demand. A governm ent that is seriously concerned about inflation will not seek
to expand demand at such a time, but will instead concentrate its efforts on se
curing better matching of the men and women who seek work with the jobs that
need to be filled. B y equipping ourselves in the future with more of the informa
tion needed to determine the true state of demand, we should be able to pursue the
objective of full em ploym ent with less danger of causing serious inflation.
This objective will be prom oted by other improvements in econom ic inform a
tion. Our statistical system is the best in the world, bu t it is n ot keeping pace
with the needs of our times. W e learn, for example, that unem ploym ent
am ounted to 3.3 million this January. W hat precisely does this figure tell us?
A short answer is that it reports the number of jobless persons who are able,
willing, and seeking to work. This answer, however, is incom plete and in some
respects misleading. In the first place, the figure includes an unknown number
of individuals who, while they are willing to work and are seeking work, are so
handicapped physically or psychologically that they would be unable to hold
down a job even in a very tight labor market. Second, the unem ployment figure
includes several hundred thousand persons who actually have job s; specifically,
those who are waiting— whether of their own choice or the em ployer’s— to start
work within thirty days, those who are searching for a new job while they are
absent from work, and those who have been temporarily laid off but have definite
instructions to return within thirty days. Third, the unem ploym ent figure
includes an undetermined number who are not looking for work diligently.
A man who applied for a job as much as sixty days ago, but made no other effort
to find a job while waiting for a reply to his application, may still be counted as
unem ployed. Fourth, the unem ployment figure includes a certain number of
persons, again of unknown magnitude, who are not looking for work in any sense,
either because they are temporarily ill, or because they are waiting to be recalled
from an indefinite layoff, or because they believe that no work is available in their
com m unity or trade. On the other hand, the unem ployment figure omits some,
perhaps many, persons who have stopped looking for work because they have
established that acceptable jobs are unavailable within their geographic reach.
Clearly, the unem ployment figures which serve as a basis for much of our policy
making are highly technical and somewhat dubious aggregates. N ot only is it
desirable to refine the concept of unem ploym ent; w e also need to learn how to
assemble and use statistics of unem ployment so that the parts which cannot be
readily influenced by broad fiscal or monetary policies may be approached by
more direct measures.
Other branches of our statistical system also show signs of age and need to be
revitalized— notably, the records of prices and wages. The quotations that enter
into price indexes of industrial commodities at wholesale are largely based on list
prices rather than actual market transactions. But in the course of an econom ic
upsurge, such as we have been experiencing, discounts tend to becom e smaller,
concessions fewer, and premiums more frequent or larger. By neglecting these
changes, our price indexes have understated the advance of the wholesale price
level since mid-1964. If more accurate price indexes had been available, we might
have realized sooner that the remarkable period of general price stability which
began in the late fifties had com e to an end, at least temporarily.
Despite their element of bias, the wholesale price indexes have the merit o f
comprehensiveness— an advantage that our measures of wage changes lack. The
fullest set of figures published by the Bureau of Labor Statistics pertains to hourly
earnings of production workers in manufacturing. These figures represent hours
paid for, not hours worked, and hence do not allow for the increasing number of
hours paid for but not worked. They do not include fringe benefits— a factor
that has becom e of m ajor im portance to employers and employees alike. A
sizable and increasing fraction of employees are classified as “ nonproduction”
workers, and they are not covered at all in the wage statistics. Finally, it is well
to note that employees in the goods producing industries are now outnum bered
by those in the service industries, and that the statistical coverage of wage rates
and earnings in the service industries is meager.
B ut the records that are used most widely and on which businessmen as well as
governm ent officials have come to rely most heavily are the estimates of gross
national product— that is, the nation’ s total output of goods and services. These
figures not only inform us on past and current econom ic conditions, but also serve
as a basis for much of the forecasting in which economists and others necessarily
engage. As is true of so many parts of our statistical system, the gross national
JOB VACANCY STATISTICS
19
product estimates are more dependable than comparable data for most other
nations. They are not, however, as good as they should be. The July 1965
issue of Economic Indicators, for example, reported that the gross national product
in 1964 was 623 billion dollars. The next m onth’ s issue reported the appreciably
higher figure of 629 billion for the same year. In fact, had it not been for certain
changes of definition that accompanied the statistical revisions, the latter figure
would have been 640 billion. Or to cite a more nearly current example, the
increase between the first and second quarters of 1965 was reported in successive
issues of Economic Indicators as 9.2, 9.5, and 11.2 billion dollars, while the increase
between the second and third quarters was reported as 11.0, 11.6, and 12.7 billion.
While I admire the constant striving of statisticians for promptness, precision,
and conceptual relevance, I also suspect that the initial under-estimates of the
growth in our nation’s output last year may have contributed to the somewhat
tardy realization by policy makers that slack in the econom y was vanishing.
I have wondered over the years, and still do, how much might have to be added
to the cost of gathering our statistics so as to reduce, if not eliminate, the need
for sizable revisions in the future, and whether the resulting benefits would not
greatly outweigh the cost. I hope that the Joint Econom ic Committee, which
has often taken the initiative to im prove our statistical system, will seek answers
to these questions.
Let me say, finally, that the implementation of the Em ploym ent A ct requires,
besides better information, more realistic models of the workings of our econom y
than are now current. Contrary to widespread notions, neither the labor force
nor the output per manhour grows steadily and smoothly, year after year. N or
is the gap between actual and potential output like a bathtub that merely needs
turning on of the fiscal faucet to be filled. Experience teaches that productivity
increments tend to decline as full employm ent is approached. If this tendency
is overlooked by the makers of policy, the bathtub may overflow. Experience
also teaches that confidence is a basic factor in econom ic life, and that it therefore
makes a difference, even if we cannot express it in a mathematical equation, how
we seek to fill gaps. Arithmetically, one dollar in the Federal budget is like any
other, but from an econom ic viewpoint the individual dollars differ. The great
success that attended the recent reduction of income tax rates cannot be attributed
solely to the arithmetical magnitude of the fiscal stimulus. It was also due to the
fact that the government took numerous steps to improve confidence after the
unhappy steel price episode of April 1962, that the fiscal stimulus adopted in early
1964 took the form of a tax reduction instead of an increase in expenditures, and
that the tax reduction became effective over the entire range of personal and
corporate incomes instead of being limited, as some well-meaning citizens had
urged, to individuals at the lower end of the income scale. But just as confidence
may be strengthened by creating a better environment for enterprise and invest
ment, so also can it be damaged by imprudent management of governmental
finances or by arbitrary interference with the workings of labor and com m odity
markets.
I wish to congratulate the present Council of E conom ic Advisers and the Joint
Econom ic Com mittee on their efforts to bring our evolving econom ic knowledge
to bear on the nation’ s econom ic condition. They need not be reminded of William
James’ pragmatic maxim that “ we have to live today by what truth we can get
today, and be ready tom orrow to call it falsehood.”
Representative C u r t i s . I would like to pick up just a couple of
points here. How much information can you get from the people in
vocational education as far as the jobs available in the community?
In St. Louis, for example, I think we have a very good public high
school vocational educational program, and our junior colleges are
moving into it. In talking with them, I find they seem to have a
pretty good grasp of job vacancies. Apparently, over a period of
years, they have developed a network of information with employers.
Can you tell me on a national basis, how much value—how much
information—can be gotten from these groups?
Mr. C a s s e l l . I am somewhat familiar with this but not as much as I
should like to be on a national basis. I come from Illinois, and about
the best I can do at the moment is to pick an example from Illinois.
I would have to say that it is very spotty from region to region and
20
JOB VACANCY STATISTICS
area to area as to the contact between vocational people, for example,
and industry people.
I think one of the great values of this kind of study is going to be
to draw the two of them together. You see, there are two sides of
the coin. And here we are producing data which enables both of them
to operate a little bit more intelligently. The job vacancy data we
are developing in the Employment Service with the cooperation of the
BLS are the kind of data that help the vocational man to assess what
the opportunities are. And they help the employer to assess where
his needs are. This, I think, helps bring the two of them together.
And I just finished last night reading a vocational educational con
ference report, which was held in Maryland a week or two ago, in
which the educators kept saying over and over again, “ this is the
kind of data we need to help us do our planning so that we can spend
our money wisely.”
Representative C u r t i s . I think you are quite responsive.
Let me add to your conclusion that the picture is spotty. St. Louis
City is a political unit, and St. Louis County, which is part of the
metroplex, is another political unit. I commented that in St. Louis
City vocational education, at least in the public school sector, seems
to be well advanced. In St. Louis County, which has a larger popula
tion than the city, we are just beginning to move. And this certainly
indicates the spottiness that you mentioned.
And I think the same goes for the private vocational educational
schools, of which there are many. There seems to be no national
organization of vocational educators where they exchange information
on the techniques that they may have developed to try to establish
early warning systems, as I call it, on what new jobs are coming in
demand, and what established skills are becoming obsolescent.
Would that meet with your observations?
Mr. C a s s e l l . I think just before you came I mentioned my own
experience as chairman of the Governor’s committee in Illinois. A
third of that committee were educators. And for 2 years I listened
to what they were saying. They were saying, “ look, when is the
Government going to produce the kind of data that we can use to
do our job?”
I spoke to a group of vocational people here in Washington about a
week ago. And when I had finished I was beseiged by requests
from them, “ will you get information from my town, from my city?”
As a matter of fact, one of the problems here is the fact that I think
the demand for this information is so wide, we are going to be amazed
at the task ahead of us.
Representative C u r t i s . Let me move into another sector of
possible source of information in gathering data. This is the private
employment agencies. In my attempts to find out what they are
doing, I find a very loose national organization. Very few of these
private employment agencies are in touch with each other. There
is a little bit of a network developed. But they certainly do generate
a lot of information on jobs available, because that is their stock in
trade. Would you comment on the extent to which you find that
they are helpful in getting information?
Mr. C a s s e l l . Since I have been on the job here—which is a little
more than 2 months—I have had three meetings with the private
employment agency people to discuss this very subject. And from
JOB VACANCY STATISTICS
21
our standpoint we are very happy to provide the private employment
agency people with the kind of data that we are producing and will be
producing that ought to help match people in jobs.
Secondly, I asked them specifically to join in in producing the
data, and supply us with as much labor market information as they
could produce from their organizational membership to help us under
stand the total labor market.
In addition, we have met with another group, with which you are
probably familiar, the college placement council.
R e p r e s e n t a t iv e C u r t i s . Y e s , in d ee d .
Mr. C a s s e l l . And we are meeting with
them later this month in
a joint meeting to discuss how this kind of information and data can
not only be used jointly, and to further the efficiency of the job
market, but to see what they can produce to help us, and what we
can produce to help them.
Representative C u r t i s . Very g o o d .
Now, a third big area is the relationship of our Military Establish
ment to the civilian sector. I have seen, over a period of years, articles,
and one just recently in the past 2 or 3 months in the Labor Depart
ment’s Bulletin, pointing out that about 80 percent of the skills needed
by the military have their counterparts in the civilian sector. But I
have been distressed to find that in the military itself, there seems to
have been very little development of nomenclature of these skills or a
liaison between the military and the Department of Labor and others
who are in the training field. We are spending over a billion dollars
a year, at least, in the military sector training people in skills existing
in the civilian sector. I have often thought that the civilian sector
could do a much better job, and that, therefore, considerable liaison
would be helpful. Would you comment, though, solely on how they
fit into your present studies?
Mr. C a s s e l l . I would like to leave the question of nomenclature
to Mr. Chavrid, whom I am sure is very familiar with this particular
problem.
Representative C u r t i s . I was only using the point on nomencla
ture as an illustration. What I am mainly interested in is what you
are doing in relating it to this big area of what the military is doing in
the field of training, and the need for skills, et cetera.
Mr. C a s s e l l . One of the most important things w~e are doing is,
we have employment service people at all the separation centers who
are working on counseling and guidance. Just last week I was down
in South Carolina, in Charleston, in that area where there are a lot
of people being separated from the military. And I talked with our
people down there about their experiences. And they are doing just
the thing you would like to see done, namely, trying to match up the
skills and the background that these people have accumulated during
their military service with appropriate civilian occupations.
This is at two levels. This is with the enlisted man who is coming
out. But then as you know, there are a whole group of people, retired
military officers with 20 years’ experience in the service, getting out at
the age of 45 or so, who have to learn new careers. And frequently
these are military executives who have to learn how to be other kinds
of executives. We are just coming to this last point. But the whole
matter of matching the military people who are coming out is being
done.
22
JOB VACANCY STATISTICS
Now, I would be the first to say that I don't think we are doing
enough of it. And I don’t think I personally know enough how this
ought to be done.
Representative C u r t i s . I am really interested in this aspect.
Say the military needs x number of welders, or x number of bull
dozer operators. Why take kids 18 or 19 years old and send them
by the numbers to welding school or bulldozer operating schools?
Why not tell the private sector what the military needs in the way of
these skills? Because they do exist. That was what was done with
the Seabee manpower procurement in World War II—matching the
skills needed with civilians having those skills.
I want to make one general observation, if I may, Mr. Chairman. I
think that the key to this whole business is, does automation create
more jobs than it destroys? I think it does. But these jobs are
frequently geographically apart from where the jobs destroyed existed.
And they are frequently in different skills.
And furthermore, a job destroyed is easier to identify that the new
jobs created. A job destroyed has nomenclature; it has a human being
attached to it. The newly created job frequently does not have no
menclature and does not have an individual human being attached to
it. But I think a great deal depends on accepting this and dealing
with it.
There are those who argue that automation destroys more jobs than
it creates. I think our disagreement lies in the fact that they have a
narrow definition of automation, and possibly a more correct dictionar}^ definition than the one I used. I am using it in the broad
sense. But if this is so, then certainly we have got to do a great deal
more in developing an early warning system, no matter what new jobs
are being created, and get the nomenclature, find out where the jobs
are, and get the men trained for them .
Thank you.
Chairman P r o x m x r e . Thank you, Congressman Curtis.
Our next witness is the Commissioner of Labor Statistics, Arthur M.
Ross.
We are delighted to have you, Mr. Ross. You may proceed in
your own way. We have a very impressive statement from you. If
you care to summarize, it would be quite satisfactory.
STATEMENT OF ARTHUR M. ROSS,
COMMISSIONER OF LABOR
STATISTICS, DEPARTMENT OF LABOR
Mr. Ross. Yes, I would like to have it introduced in the record.
And I will summarize it for 10 or 15 minutes if I may.
Chairman P r o x m i r e . Yes, sir.
Representative C u r t i s . And may I say, Mr. Cassell, that any
additional information that you would like to put in, and extend your
answers to questions, you may do so.
Mr. C a s s e l l . I would be happy to do so.
Chairman P r o x m i r e . That is a good addition.
Mr. Ross. Mr. Chairman, I am going to deal particularly with the
analytical uses of job vacancy data. And I think it might be helpful
to comment on a couple of points raised in the question before sum
marizing my statement.
The question was raised, How are the data collected? In the job
vacancy surveys the data are collected directly from employers gen
JOB VACANCY STATISTICS
23
erally through questionnaires. We are experimenting with two
methods. One is a separate questionnaire on job vacancies which is
sent to the employer by the local employment office. The other is
to combine the job vacancy questions with the labor turnover ques
tions; that is, another option. If that works out well, it might be
possible to get labor turnover information for all industries, manu
facturing, and nonmanufacturing, together with the vacancy data.
This is an open question of what technical option we use, depending
upon the outcome of the experiments.
Chairman P r o x m i r e . Will you give us a little more information
on what you mean by “labor turnover” ?
Mr. Ross. That is the information on accessions, terminations,
the quit rate, the layoff rate, and so forth. We now get those data
for manufacturing industries only. And we would like to extend
them also to the other industries.
Mr. Chairman, you also raised the question of whether there is a
concrete definition of the job vacancy, or whether it is a pretty
ambiguous thing.
I think we do have a pretty concrete operational definition. In
these surveys we define a vacancy as an opening for which they need
a new worker, somebody who is not presently in the company, a job
that is unoccupied, a job which is immediately available, and one for
which they are actively recruiting. And I would submit that that is
a pretty understandable definition. And we do find that employers
generally understand what we mean, and that they are able to identify
their vacancies. In the experimental surveys the Department has
made followup personal visits to many of these firms, and they have
corroborated that the employers’ concept of a vacancy generally
corresponds to our concept of a vacancy.
So I think that question of yours could be answered in the
affirmative.
Proceeding to my summary, Mr. Chairman, I think the general case
for vacancy data is that we ought to know about labor demands as
well as labor supply. Our information on unemployment and on the
growth of the labor force tells us about the oncoming supply and
existing supply of manpower. We do have more of a vacuum of
information on the demand side.
The interest and need for vacancy data is greater now because of
the manpower shortage problem. Secretary Wirtz called in Mr.
Ruttenberg, the Manpower Administrator, and myself back in
December. And he told us that he thought we were doing a pretty
good job in studying unemployment, but we had a long way to go
before we could master the problems of information about labor
shortages.
We have been working on that ever since. I filed a monthly report.
And the President has asked me to make public a monthly report on
both present and prospective shortages.
This calls for careful analysis of the whole economic picture. And
I think we all feel that if we had better information on labor demand
in the form of job vacancies we could do a better job at this.
Now, I don’t think this is limited to the present year. If we are
going to approach and maintain full employment, there are going to
be constant problems of imbalance between supply and demand of
manpower. There are going to be constant mismatches. It will not
24
JOB VACANCY STATISTICS
be an easy task to reconcile the two, and I think there is an increasing
and constant need for this.
In the statement I described some of the background, the experi
ments during World War II and the Korean war, the development of
foreign vacancy statistics. And 1 might say that we would not be
satisfied with the kind of program they have in foreign countries,
which is based upon employment service orders.
Representative C u r t i s . Would you identify the countries, if you
could?
Mr. Ross. Yes, sir.
Representative C u r t i s . Are they in there?
Mr. Ross. They were in the statement.
Representative C u r t i s . Which are the better? Could you jast
comment? i guass they are Western Europe mostly.
Mr. Ross. Yes. The ones mentioned in the statement, among
them are the United Kingdom, Sweden, the Netherlands, and France.
They are not the only ones, but they are the ones who have done
perhaps the most with it.
You are also familiar with the experimental surveys of the National
Industrial Conference Board in Rochester. And we will report at
greater length upon our own experimental surveys daring the past
couple of years.
You perhaps know the National Bureau of Economics Research
had a conferenca on job vacancies research last year. I was a member
of the planning committee. And I think that many of the conclu
sions about job vacancy research were brought together there.
Now, my statement does emphasize the analytical uses of the
information, and I think I would like to read four or five pages from
my statement, and that would be about what I would want to
emphasize.
The lack of job vacancy information constitutes the most significant
gap in our knowledge of labor market conditions. Statistics on job
vacancies would give us a measure of unsatisfied demand for labor
which, together with our data on employment, would provide a more
complete measure of the demand for labor—something we have
never had before.
Some of the major potential analytical uses for job vacancy in
formation are as follows:
1. Job vacancy information can be used to develop a picture of
the size and characteristics of unfilled demand for labor. Such in
formation can then be analyzed in its own right, just as many useful
analyses are made of the size and characteristics of unemplyoment.
2. Trends in job vacancies, especially if classified by occupation,
can be of considerable value in throwing light on the ability of our
economy to adjust to changes in the demand for labor, and a vacancy
information would be a good lead indicator.
3. When used in conjunction with data on employment, unemploy
ment turnover and hours of work, this can help us analyze the current
economic situation to bring to light the major policy decisions that
have to be made in dealing with unemployment and labor shortages,
inflation, and so on.
4. We could—well, I mentioned that the President has asked me to
prepare regular reports on shortages, and I feel we could do a better
job in addition to the indirect and circumstantial data we now have
JOB VACANCY STATISTICS
25
to use if we had direct evidence through measures of vacancies classi
fied by occupation, industry, and area.
5. The job vacancy information will throw additional light on de
mand and supply conditions in the labor market, in relation to chang
ing wage levels. And I believe myself that the wages for the vacancies
should always be taken into account in analyzing them.
6. The vacancy data can help us sharpen our projections of man
power requirements by occupation, which are a very essential tool in
estimating training nee^ds, zjid in counseling both, at the employment
service level and in the educational institutions.
7. The information, as Mr. Cassell has emphasized, can be used by
business firms to get a picture of the area within which they are
recruiting, and to develop their recruiting and in-plant training
policies.
And finally, I feel that this information can be of equal value to
labor organizations in evaluating the demand for the services of their
members and in developing policies for training, apprenticeship, and
collective bargaining.
In meeting these analytical needs more information is required
than merely the number of vacancies. We need to know how many
of the jobs employers are trying to fill have been vacant only briefly,
and how many of them represent hard-to-fill jobs. The latter may
indicate imbalances between supply and demand, resulting from a
disparity between the skills needed by industry and the skills available
among unemployed workers in the community. They may also
reflect unrealistic hiring standards, or low wage rates and unfavorable
conditions of employment. To get insight into these questions we
need job-vacancy data separately for each local area, and by specific
occupation. And we also need information on the wages to see what
proportion of the vacancies are being offered at substandard levels.
Job-vacancy information can add greatly to our analytical insights
into the labor market, particularly if studied together with other data.
Analysis of vacancies together with data on labor turnover will
suggest what proportion of the vacancies at any time represent a
normal condition reflecting the typical turnover experience in the
industry. For example, we would expect the construction industry
to have a higher rate of vacancies than an industry characterized by
more stable employment, such as banking.
Similarly, we need to analyze vacancies in relation to employment
and its seasonal and cyclical fluctuations. If you had a sudden in
crease in vacancies in an apparel manufacturing, it might only mean
that it is entering the busy season, but a similar increase reported by
a nonseasonal industry would be another matter all together.
It would also be essential to study vacancies in relation to the
apparent supply of labor as reflected in unemployment data for the
locality. In the pilot studies we made last year, we found a large
number of vacancies in relation to the apparent supply of workers in
professional, managerial, and sales occupations, and a smaller number
of vacancies in relation to supply in the blue-collar and less-skilled
occupations. In order to make these analyses possible, I might say
that we need better local data on unemployment, because as you
know, the unemployment estimates locally leave much to be desired.
The Department of Labor is working to improve these measures, and
hopes to secure resources to do so.
26
JOB VACANCY STATISTICS
In interpreting the significance of job vacancies in an industry or
occupation, we need to relate the vacancies to the total employment.
For example, if you had 5,000 vacancies for engineers, that would
be one-tenth of 1 percent of the total number of engineers in the
country. Five thousand vacancies for physicists would equal 12%
percent of all physicists.
So in order to put the vacancy data into perspective, we need
estimates of employment by industry, both on a nationwide and
local basis. We have those. We also need better information for
occupations which we are onlj now setting out to develop on a sys
tematic basis.
And finally, to make the data on job vacancies most useful for ana
lytical purposes, we need to study the trend and vacancies in relation
to those in the other measures, including employment, unemployment,
labor turnover, hours, and wages. ^
I am confident that the data on job vacancies will provide a valuable
additional dimension to our system of economic changes, and will
provide critical insights into the economy. In some cases we will be
able to identify what measures should be taken to improve the speed
and efficiency of local placement and recruiting mechanisms. In
another case we will have a clearer indication of the need to make a
better match between workers and jobs by retraining, counseling and
relocation grants, and programs to reduce discrimination.
Analysis of labor supply and demand information may point to the
need for higher levels of effective demands in periods other than the
present.
The analysis of job-vacancy data, together with other information
on the economy, will give better clues than previously have been avail
able to the selection and timing of private and public policies.
Now, we have given a lot of thought to the criteria and technical
definitions. I won't go into that, although Mr. Chavrid may cover
some of that. And we can do it in the questioning, if you like, Mr.
Chairman.
We do discuss in my statement the pilot surveys from the standpoint
of how you define vacancies—and there are problems there. And we
discuss the pilot surveys from the standpoint of occupational classifi
cation. And there are problems there. And we discuss the handling
of wage data in the analysis of vacancy information. We discuss the
problems of establishment sample and industry coverage. And we
discuss the response accuracy problem—I might say that in the re
sponse accuracy checks w^hich are summarized in my statement we
have had pretty encouraging results in terms of not only employer
cooperation but also the accuracy of their reports when checked against
the evaluation by skilled BLS people who actually went into the plants
and studied the vacancies themselves.
There is an understatement of vacancies of about 12 percent, as a
result of various factors which we hope can be corrected.
Now, there has been considerable concern, both in labor and man
agement circles, about the danger that job-vacancy data once collected
would be misused. One possible misuse that has caused some appre
hension is that the number of job vacancies might be subtracted
from the number of unemployed workers, and the difference between
these two figures might then be represented as the “ true” unemploy
ment figure. I do not know any serious student of labor problems
who would propose such a use; for the problem of unemployment is
JOB VACANCY STATISTICS
27
highly complex and must be considered in relation to the age, sex,
color, skills, local residence, and other characteristics of the unem
ployment. We have to recognize that no economic measure or
statistical measure is completely free from the danger of misuse.
This goes for our price statistics, our employment and unemployment
statistics, as well as the vacancy data we would like to develop. If
we were to refrain from collecting any data that could possibly be
twisted or used out of context, then we would have to close all of our
statistical agencies, as well as the scientific laboratories for the po
tential misuse is much more deadly in the field of science than in the
field of statistics.
I believe that the danger of possible misuse should not be a bar to
the development of data under responsible auspices that will be
valuable if properly used. We must recognize, of course, the obligation
of Government agencies, responsible for statistics to use the data
properly, and to caution the public against improper interpretations,
and we will be assiduous in this regard. Indeed, we intend to analyze
the data on job vacancies carefully, and acquire deep experience in
the behavior of the data before interpreting its significance.
I would be the last to claim that we have solved in advance all the
problems of concepts, all the problems of checks, all the problems of
analysis and interpretation. We will not rush to publication before
we have this experience.
And I might observe, Mr. Chairman, that this is something that is
true of all of our economic intelligence. We have been collecting and
publishing unemployment data for many years. But even now we
are experimenting with different concepts of unemployment through
a supplemental family household survey, and we will make a decision
later this year as to changes in the definition of unemployment, and
changes in the way that the families will be interrogated. If this is
true for unemployment after so many years of publishing these data,
it stands to reason that we have nothing to be ashamed of in the fact
that we have not solved all the conceptual or analytical problems of
vacancies.
My concluding paragraph is as follows: Let me say that although
the conceptual and technical problems described above are important,
they are not of such magnitude as to prevent the statistics collected
in a national program from being useful for analytical and operating
purposes. We have come a long way in the past few years in identify
ing and resolving problem areas. We have examined in detail the
arguments for and against the problem of job vacancies statistics. We
feel that the need for job vacancy data heavily outweighs the prob
lems involved. An ongoing and comprehensive vacancy program can
greatly assist in the implementation of an active manpower policy and
the development of economic analysis needed for major policy decisions.
Thank you, Mr. Chairman.
(The complete statement of Mr. Ross is as follows:)
P r e p a r e d S t a t e m e n t op A r t h u r M . R o ss, C o m m is s io n e r op L a b o r S t a t is t ic s ,
B u r e a u op L a b o r S t a t is t ic s , D e p a r t m e n t o p L a b o r
i.
in t r o d u c t io n
Over the past two decades, there has been continued discussion o f the potential
uses o f quantitiative information on job vacancy data. Interest in such a pro
gram intensified in the early 1960’ s as the governm ent instituted an active man
28
JOB VACANCY STATISTICS
power policy to com bat high unemployment. In part, interest centered on
vacancy statistics as a guide to labor market programs, including the placement,
training, retraining, and relocation o f unemployed workers. In part, interest
reflected the debate on what general economic policies should be adopted to
reduce unem ployment. As the econom y has m oved from high to low unem ploy
ment in the mid-1960’ s, shortages o f workers in some occupations and industries
have provided added impetus to obtain information on job vacancies.
The President’ s recent Manpower Message emphasizes the need for accurate
inform ation concerning the extent and nature o f labor shortages. “ Specific
shortages of labor slow up the expansion of the econom y. They can put pressure
on costs and prices. W e are determined to do whatever is necessary to keep the
econom y expanding and avoid inflationary bottlenecks. The time to deal with
manpower shortages is before they develop. Effective manpower policies can
reduce unem ployment and at the same time head off manpower shortages.”
The President also directed the Commissioner o f Labor Statistics to provide
the “ fullest possible inform ation on existing or threatening labor shortage situa
tions.” For each o f the past several months, in response to the President’ s
directive, I have prepared a report on manpower shortages and reserves. The
evidence on which these reports are based has been primarily circumstantial—
em ploym ent gains, unem ploym ent declines, changes in hours o f work, labor
turnover rates, job openings filed with the Em ploym ent Service, and other in
direct indicators of manpower problems. Our experience in preparing these
first reports has made it clear that direct evidence concerning unfilled vacancies
is essential for determining the true extent and nature of manpower shortages.
W ith such information, and with analyses of related data, we could more effec
tively pinpoint and deal with problems o f shortages as well as surpluses in relation
to current demand.
Attem pts to obtain inform ation on job vacancies and the unfilled demand for
labor go back many years and have encompassed many different approaches.
Indexes of help-wanted advertising date back to the mid-1920’s. Statistics on
actual job vacancies were extensively collected by State employm ent security
agencies during W orld W ar II and the Korean War, and are still collected on a
bi-m onthly basis in several labor areas. The United States Em ploym ent Service
record of jo b orders, which shows the number of jobs that employers place with
the Em ploym ent Service to recruit workers, has provided insight into the demand
for labor. Since only a portion of all job openings are placed with the E m ploy
ment Service, however, the usefulness of this inform ation is somewhat limited.
In the mid-1950’s, the Departm ent of Labor began to examine ways of de
veloping jo b vacancy inform ation on a systematic and statistically sound basis.
The Bureau of Labor Statistics undertook a pilot study in 1956 to determine if it
were feasible to collect job vacancy data by mail. The Bureau’ s initial reaction
was that employers did not have sufficiently complete records and could not make
sufficiently accurate estimates of vacancies.
Nevertheless the President’s Com m ittee to Appraise Em ploym ent and Unem
ploym ent Statistics (the Gordon Committee) asked the Bureau of Labor Statistics
in 1962 to re-examine and report on the possibility of setting up a m ajor job
vacancy statistical program, pointing out the widespread interest in such a
program and its potential uses. The Bureau’s report, A National Statistical
Program on Job Vacancies,1 as well as the many comments the Com m ittee received
from those w hom it consulted, favorably impressed the Com m ittee which recom
mended that the Department of Labor initiate the program of research suggested
in the report. This research, undertaken on a cooperative basis by the D epart
ment’s Bureau of Em ploym ent Security and the Bureau of Labor Statistics in
fiscal year 1964, concentrated on conceptual and definitional problem s. It
included pilot feasibility studies in the Chicago, and Buffalo areas and a survey of
the nature and uses of job vacancy statistics being collected in foreign countries.
An exploratory study was also conducted by R obert Ferber and Neil Ford of the
University of Illinois in 1963 and 1964.2
The research undertaken in 1964 identified many of the conceptual and technical
problems that had to be solved. In fiscal year 1965, a more extensive pilot
program was set up to find out whether employers would be willing and able to
report job vacancy information by occupation on a voluntary basis. This pro
gram which was continued in fiscal 1966, demonstrated clearly that employers can
1 Printed as Appendix B of the Committee’s Report.
2 Robert Ferber and Neil Ford, “ The Collection of Job Vacancy Data Within a labor Turnover Frame
work,” in Employment and the Labor Market, edited by A. M . Ross, University of California Press, 1965.
JOB VACANCY STATISTICS
29
and will provide such information, and that reasonably accurate estimates can be
made.
The results of a conference held by the National Bureau of Econom ic Research
in February 1965 also fortified the Departm ent’s belief that a jo b vacancy re
porting program was feasible and desirable. A t this conference, a large number of
papers were presented covering all aspects of the measurement and interpretation
of jo b vacancies, including the potential uses, problems, advantages and dis
advantages of such a program. The insights gained from the discussions helped
to resolve many of the D epartm ent’s doubts, and pointed the way toward further
im provem ent in the collection techniques and refinement of conceptual issues.
The Departm ent’s conclusion as to the feasibility of collecting meaningful job
vacancy data wTas further corroborated by an exploratory study in Rochester,
N .Y ., done in 1965 by the National Industrial Conference Board, with support
from the Ford Foundation. The report of the Conference Board stated that
“ statistics on the number of jo b vacancies b y occupation can be collected with a
level of accuracy that renders the figures meaningful.” It further stated, after
examining the uses and the money costs of such a program, that
. . in our view
the documented uses and the additional expected uses am ply justify the estimated
costs.”
Interest in and use of job vacancy data is not limited to the United States.
M any foreign countries have instituted such programs, among them the United
Kingdom, Sweden, the Netherlands, and France. Foreign experience has pro
vided us with valuable insights into the advantages, disadvantages, and problems
connected with jo b vacancy programs. Let me describe briefly some of the char
acteristics of the jo b vacancy programs of foreign countries.
Job vacancy statistics in foreign countries are primarily the product of the
administrative statistics of local employment service offices. M ost countries
tabulate and publish statistics on the number of vacancies registered each month,
the number filled through their employm ent services (placements), and the
number remaining unfilled as of a particular day. Published figures are frequently
broken down by regional, occupational, and industrial classification, generally
corresponding to the classifications used for the em ploym ent service statistics on
registered unem ployed jo b applicants. A few countries obtain jo b vacancy
statistics from establishment surveys, usually in addition to their em ploym ent
office statistics.
M ost foreign programs have a relatively loose definition of a jo b vacancy,
corresponding to that used in registering employer jo b orders. Similarly, there
are no standards as to how actively an employer must be seeking a worker. M ost
countries count not only current vacancies, but also registered vacancies which
the employer does not wish to fill until some future date. Generally, they count
registered vacancies involving part-time work, tem porary or even casual labor,
as well as regular full-time work.
Statistics collected in most foreign countries represent an incomplete count of
existing vacancies. The proportion of total vacancies covered varies among
countries, depending on the degree of employer utilization of the em ploym ent
offices, and varying with occupation and industry. The actual proportion is
unknown in most foreign countries.
Despite the limited coverage of their vacancy statistics, foreign countries make
extensive use of the inform ation resulting from their statistical programs. The
primary use is for recruitment and placement.
Some countries find m onthly employm ent office job vacancy statistics to be
useful indicators of trends in the demand for labor, both total and by area, in
dustry, and occupation.
A number of countries use job vacancy figures by area and occupation in
conjunction with unem ployment figures to analyze the causes of em ploym ent and
to determine needed corrective measures. An examination of the tw o sets of data
provides a useful indication of labor shortages and surpluses in particular occupa
tions, nationally, regionally, and locally. Coupled with other statistics, job
vacancy inform ation is used in the formulation of policies on labor mobility, loca
tion of industry, geographic maladjustments, and structural unemployment.
A number of countries use job vacancy statistics as a lead indicator of changing
econom ic conditions, particularly as an indicator of approaching downtrends in
the econom y. Others indicate that they are an excellent indicator of emerging
labor shortages, and of inflationary pressures resulting from labor market condi
tions.
A few countries use job vacancy statistics as a guide to training needs, mostly
for short-term worker training programs. Current vacancy statistics, how'ever,
63 - 947— 66 ------------ 3
30
JOB VACANCY STATISTICS
are not usually viewed as a good indicator of future trends, and are therefore less
useful for determining training programs of a more extended nature. They may
also be useful for vocational guidance, but only as a very broad indicator of trends.
M any insights have been gained from foreign experience with job vacancy
statistics. One is that these figures should be used in conjunction with other
data that are relevant in analyzing conditions in the job market.
W e would not be satisfied with a system of statistics on job vacancies for job
market areas in the United States as inadequate as those described above. Our
econom y is larger and more complex than those of the countries cited and our other
manpower statistics are correspondingly more detailed and comprehensive. The
system we are proposing for job vacancy information is, therefore, more elaborate
and will suit our needs much better than the kind of simple tabulation described
as a result of compiling applications to employm ent service offices.
II. ANALYTICAL USES OF JOB VACANCY INFORMATION
The lack of job vacancy information constitutes the most significant gap in our
knowledge of labor market conditions. Statistics on job vacancies would give us
a measure of unsatisfied demand for labor which, together with our data on em
ploym ent, would provide a more complete measure of the demand for labor— som e
thing we have never had before.
I would like to summarize briefly some of the major potential analytical uses for
jo b vacancy information and then discuss some of them at greater length.
1. Job vacancy information can be used to develop a picture of the size and
characteristics of unfilled demand for labor. Such inform ation can then be ana
lyzed in its own right, just as many useful analyses are made of the size and charac
teristics o f unemployment.
2. Trends in jo b vacancies, especially if classified by occupation, can be o f con
siderable value in throwing light on the ability of our econom y to adjust to changes
in the demand for labor. They may serve as a lead indicator of changing eco
nom ic conditions.
3. Job vacancy information, when used in conjunction with inform ation on
em ploym ent, unemployment, labor turnover, and hours of work, can enhance
our ability to analyze the current econom ic situation for light on m ajor policy
decisions that have to be made in dealing with unem ployment, labor shortages,
and inflation. I shall discuss this in more detail below.
4. In the present econom ic situation, the question of labor shortages has becom e
sufficiently critical, especially in relation to skilled manpower, that the Presi
dent, as previously indicated, has asked the Departm ent of Labor to w atch the
situation closely and to prepare regular reports. I have already pointed out
that much of our evidence on labor shortages is indirect and circumstantial. W e
could do a much better jo b if we had direct evidence on labor shortages through
measures of jo b vacancies classified by occupation, industry and area.
5. Job vacancy information will throw additional light on demand-supply
conditions in the jo b market in relation to changing wage levels. Analyses of
the effect of em pk^m ent changes upon wage rates, although potentially very
useful in appraising wage developments and policy, has not exhibited highly pre
cise results when applied to data available for the United States. The additional
dimension of jo b vacancies in the measurement of labor demand would contribute
another powerful tool of analysis.
6. Job vacancy data can help us to sharpen the Bureau of Labor Statistics’
projections of manpower requirements by occupation which are so essential in
developing estimates of training needs to guide in the planning of the many
education and training programs supported by the Federal Government. Up to
now, these projections have been based on analysis of past trends in manpower
requirements as measured by employm ent. We have recognized that in so far
as there is unsatisfied demand for labor, the figures on employm ent are an imperfect
measure of demand for labor.
7. Job vacancy information can be used by business firms to get a picture
o f the area within which they are recruiting workers, and to help in developing
more effective recruiting policies. This would be especially valuable to firms
considering new plant locations.
8. Such inform ation could be of equal value to labor organizations in evaluating
the demand for the services of their members and in developing policies for
training, apprenticeship, and collective bargaining.
In meeting these analytical needs more information is required than merely the
number of vacancies. We need to know how many of the jobs employers are
JOB VACANCY STATISTICS
31
trying to fill have been vacant only briefly, and how many of them represent
hard-to-fill jobs. The latter m ay indicate imbalances between supply and
demand, resulting from a disparity between the skills needed by industry and the
skills available among unemployed workers in the community. They may also
reflect unrealistic hiring standards, or low wage rates and unfavorable conditions
of employm ent. T o get insight into these questions we need job vacancy data
separately for each local area, and by specific occupation. W e also need informa
tion on wages to see what proportion of the vacancies are offered at wage levels
below prevailing entry rates for the occupation in the community. From m any
points of view additional information on the personal and skill requirements;
associated with each vacancy would be desirable, but it would be difficult to collect
all these specifics as part of a large-scale survey program. Job vacancy informa
tion can add greatly to our analytical insights into the labor market, particularly
if studied together with other data. Analysis of vacancies together with data on
labor turnover will suggest what proportion of the vacancies at any time represent
a normal condition reflecting the typical turnover experience in the industry.
For example, we would expect the construction industry to have a higher rate of
vacancies than an industry characterized by more stable employment, such as
banking.
Similarly we need to analyze vacancies in relation to employm ent and its sea
sonal and cyclical fluctuations. W e would expect a sudden increase in vacancies
to be reported by a seasonal industry such as apparel manufacturing at the start
o f its busy season, but a similar increase reported by a non-seasonal industry
would be another matter altogether.
It would also be essential to study vacancies in relation to the apparent supply
o f labor as reflected in unem ployment data for the locality. In the pilot studies
we made last year, we found a large number of vacancies in relation to the ap
parent supply of workers in professional, managerial, and sales occupations, and
a smaller number of vacancies in relation to supply in the blue collar and lessskilled occupations. In order to make these analyses possible we need better
local data on unemployment, and the Departm ent of Labor is working to im prove
these measures.
In interpreting the significance of job vacancies in an industry or occupation
we need to relate the vacancies to total employment. Five thousand vacancies
for engineers represent only y¿ of 1 percent of the total; 5,000 for physicists would
equal 12y2 percent of all physicists in the United States. In order to put the
vacancy data into perspective we need estimates of em ploym ent by industry,
nationally and for States and local areas, which we already have, and similar
inform ation for occupations, which we are only now setting ou t to develop on a
systematic basis.
Finally, to make data on job vacancies more useful in analyzing job market
developments we need to study trends and changes in vacancies in relation to
those in other measures— employment, unemployment, labor turnover, hours
and wages. In order to make this possible we need systematic statistics on job
vacancies, comparable from locality to locality, and collected at regular intervals.
I am confident that data on jo b vacancies will provide a valuable additional
dimension to our system of econom ic measures and will provide critical insights
into the econom y. In some cases we will be able to identify what measures should
be taken to im prove the speed and efficiency of local placement and recruitment
mechanisms. In other cases we will have much clearer indication of the need to
achieve a better match between workers and jobs, by such measures as training,
retraining, counseling, relocation grants, and programs to reduce discrimination.
Analysis of labor supply and demand information may point to the need for higher
levels of demand.
The analysis of jo b vacancy data together with other information on the
econom y will give better clues than have previously been available to the selection
and timing of public and private policies.
T o make these analytical uses possible, certain requirements are imposed on the
jo b vacancy data we develop. In particular, the definitions and concepts used in
the measurement of vacancies should be appropriate for their analysis as part of
the whole structure of job market information. M oreover, the occupational and
industrial classification of the data should be comparable with the classification
systems used for other job market data.
There has been considerable concern about the danger that jo b vacancy data,
once collected, would be misused. One possible misuse that has caused some
apprehension is that the number o f jo b vacancies would be subtracted from the
number of unem ployed workers and the difference between the tw o figures would
32
JOB VACANCY STATISTICS
be said to be the “ true” unem ployment figure. N o serious student of labor
problem proposes such use, for the problem of unem ploym ent is highly complex
and must be considered in relation to the age, sex, color, skills, and other char
acteristics o f the unemployed.
We must recognize that no econom ic statistical measure is com pletely free
from the danger of misuse. If we w'ere to refrain from collecting any data that
could possibly be twisted, or used out of context, we would have to close all the
statistical agencies of government (as well as all scientific laboratories— where
potential for misuse is even more deadly than in the field of statistics).
I
believe strongly that the danger of possible misuse should not be a bar to the
developm ent of data that are valuable if properly used. We must recognize, of
course, the obligation of government agencies responsible for statistics to use the
data properly and to caution the public against im proper interpretations, and
we will be assiduous in this regard in this new statistical program. Indeed, we
intend to analyze data on jo b vacancies carefully and to acquire deep experience
in the behavior of the data in reflecting the jo b market before interpreting its
significance. I would be the last to claim that we have solved, in advance, all the
problems of analysis and interpretations. We will not rush to publication before
we have this experience.
III. THE PILOT SURVEYS OF JOB VACANCIES BY THE DEPARTMENT OF LABOR
The pilot studies made by the Departm ent last year, and those still going o n 1
will be described in some detail by M r. Chavrid of the Bureau of Em ploym ent
Security. I will com m ent only on the technical aspects of the surveys— problems
of definition, sampling and estimating methods, the results of our studies of the
accuracy of response, and the problems of assuring that the vacancies reported
were being offered at or above prevailing entry wage rates in the com m unity.
Despite a strong belief that a job vacancy statistics program is feasible and
can add much to our knowledge of the functioning of the job market, we are
fully aware that some conceptual and technical problem s remain to be solved.
We have already solved many of them, but to im ply that the solutions have met
with unanimous acceptance would be misleading and might endanger the future
success of the program.
Definitional Problem s: Defining a job vacancy has been the subject of intensive
research by the Department. The merits and demerits of various definitions,
including those used by the National Industrial Conference Board and by foreign
countries, have been explored in depth. In the course of developing a definition,
many difficult questions had to be resolved.
Should vacancies to be filled by recall of laid-off workers be included?
Should vacancies expected to arise at some future date be included?
Should tem porary or seasonal openings be included as well as permanent ones?
Should vacancies to be filled by transfer, reassignment, prom otion, or dem otion
of present employees be included?
Should a vacancy be counted only if it is open for a period of time, or should
it be counted the moment it opens up?
Should a vacancy be counted if it results from a work stoppage?
Should part-time as well as full-time jobs be counted?
After considering all the alternatives and their ramifications, the definition we
have provided to employers in the pilot surveys is as follows:
A current jo b vacancy is an existing em ploym ent opportunity in your estab
lishment for some worker from outside your firm (i.e., a “ new” worker— not a
com pany employee) for a job that is unoccupied and immediately available for
occupancy b y a “ new” worker for whom your firm is actively searching or re
cruiting. Include such vacancies for all kinds of positions, classifications and
em ploym ent (full-time, part-time, permanent, temporary, seasonal), including
those outstanding on orders with em ploym ent agencies and notifications to unions.
Exclude jobs to be filled by recall, transfer, prom otion, dem otion or return from
paid or unpaid leave; jobs unoccupied because of labor-management disputes;
and jo b vacancies for which “ new” workers were already hired and scheduled to
start work later. 11Actively searching or recruiting” means current efforts to fill
the jo b vacancy through orders listed with public or private em ploym ent agencies
and school placem ent offices; notifications to labor unions and professional orga
nizations; “ help-wanted” advertising (newspaper, post office, e tc .); recruitment
programs; interview and selection of applicants.
The m ost im portant feature of this definition is that it spells out three condi
tions which must be met before a job vacancy can be counted. (1) The job must
JOB VACANCY STATISTICS
33
be unoccupied; (2) the job must be available for immediate occupancy by a new
worker outside the com pany; and (3) the job must be the object of an active search
for a new worker outside the company.
Through use of these criteria, the job vacancy definition used is reasonably
comparable with the unem ployment definition used in the BLS m onthly house
hold survey of the labor force. Just as an unem ployed individual must be
actively seeking work to be counted as unemployed, a true job vacancy must
entail a positive com pany effort to fill it. Similarly, the vacancy must be un
occupied and available for immediate filling, just as an unem ployed worker must
be available for immediate employm ent. (In our present unem ployment meas
ure a few persons seeking a jo b for which they would only be available at a future
date— such as a student out of school in Easter week who is looking for a Summer
jo b — are counted as unem ployed; we are seeking a way to change this definition
to include only unem ployed persons available for work in the survey week.)
This approach thus eliminates the vacancy which will not be filled from outside
he firm, i.e., by the unem ployed or by new workers, such as the vacancy for the
com pany president to be filled from the roster of existing officers.
One question about the current job vacancy definition still under study relates
to how long a vacancy must exist before it should be included in the count of job
vacancies. One point of view is that a job vacancy should exist for an entire
week before being counted for this purpose. Since an unem ployed person is
generally one who was not at work at any time in an entire week and was looking
for a job, it is argued that a vacancy should be “ looking for a new worker” for an
entire week. A contrary viewpoint holds that all vacancies existing at a given
point in time, such as the last day of the week, should be included in the vacancy
count. It is argued that any vacancy which is immediately available for occu
pancy and for which new workers are actively being sought at any given time is
immediately available to an unem ployed person and would change his em ploy
ment status if he were to begin working immediately in the available vacancy.
In the pilot studies in fiscal year 1965, data were collected using both defini
tions— i.e., vacancies existing on the last day of a certain week and those which
had been vacant the entire week. Over four-fifths of the estimated total vacancies
as of the last day of the week had existed one week or longer. In two-thirds of
the areas the proportion of vacancies existing one week or longer exceeded 85
percent and in one-half of the areas it was 90 percent or more. In my opinion,
the issue is a close one and should be given further study.
Our research indicates that employers can respond reasonably accurately to a
request for data under the definition used in the pilot surveys, despite the absence
in some companies of formal records on vacancies. Nevertheless, it would be
misleading to im ply that this, or any definition for that matter, can resolve all
the problems entailed in collecting the information or put to rest all the criticisms
of the program. We have examined and will continue to examine the problem,
hoping thereby to im prove the definition and the statistics collected.
We are, for example, collecting an additional statistic in the tests now under
way. This figure, represents the number of jobs for which active recruitment is
underway but for which no actual vacancy exists at the moment.
Occupational classification
Identification o f occupations is another major problem o f the jo b vacancy
program as it is with many other statistical collection programs. Vacancy in
form ation must be provided by specific occupation in order to meet the require
ments for either an operating program or an analytical program. The amount of
occupational detail to be collected is a particularly im portant problem .
In the D epartm ent’s pilot surveys, vacancies were requested only by jo b title;
those reported were then classified and coded by occupation according to the
Dictionary of Occupational Titles, by local em ploym ent service personnel.
(Employers themselves were not asked to code the occupations, since m any
employers are not familiar with or do not use the D O T in their classification
system.) Coding of jo b titles is a very difficult task even for experienced occupa
tional specialists, since no standard term inology is universally used by industry.
M any employers do not have formal jo b classification systems, often providing
jo b titles to fit a particular worker or a particular work situation. Other establish
ments use general jo b titles to cover a wide range of occupational duties and respon
sibilities. Thus, the precision of the occupational classification system now used
in the program still needs to be im proved.
Several alternatives for the classification of occupations were examined during
the course of the Departm ent’s experimental research. These included the
34
JOB VACANCY STATISTICS
method now used of asking for job titles and having local personnel of the State
em ploym ent office classify them by D O T number. Another possibility was to
provide each respondent with a precoded list of the occupations on which data
were required. Still another was the developm ent of a carefully worked out,
precoded list of occupations with accompanying brief descriptions that were
adapted to each industry but would be comparable among industries. None of
these possible methods, including the method finally decided upon, was w ithout
its conceptual and technical problems. Such problems, however, are not peculiar
to the job vacancy program alone, but extend through all statistical programs deal
ing with occupations. M ore work still remains to be done on occupational
classification before a fully satisfactory solution can be found.
Wage rates offered for vacancies
It is obviously im portant in evaluating job vacancies to distinguish between
bona fide vacancies and those offered with wages or conditions of work below pre
vailing norms. For analytical purposes what is needed is a measure of demand at a
given level of wages, and when the intention is to compare demand with supply the
appropriate w^age level is that currently prevailing for the occupation, at the poin t
of hiring, in the community. Another reason for our interest in knowing the wage
levels being offered for the vacancies was the apprehension that some employers
might report an unrealistically high number of vacancies.
For these reasons, in the tw o survey rounds of F Y 1965 an attem pt was made on
a limited scale (in the five Standard M etropolitan Statistical Areas of Baltimore,
Chicago, Los Angeles, Miami, and Minneapolis-St. Paul) to check the wage rates
offered against prevailing entry rates. The questionnaire used in these areas
asked for the rate of pay offered for the job to which the vacancy related. The
respondent was encouraged to furnish a single rate of pay but was also told that
where a range of rates was offered, depending on the varying educational, training,
and experience qualifications of prospective applicants, a statement of the offered
pay rate range would be acceptable. Encouragingly, most employers with
vacancies showed no reluctance to report wage data.
T o obtain information needed to check the wage rates, local em ploym ent
service offices were instructed to furnish prevailing area entry wage rates for each
occupation for which one or more employers reported both vacancies and pay
rate information. This information might be based on employer orders, “ suitable
wage” determinations under the Unemploym ent Insurance program, union wage
scales, BLS Com munity Wage Surveys, management-labor contracts, and the
best judgm ent of local office personnel most knowledgeable about wage rates for
particular occupations. In reporting this information, the employm ent services
offices rejected the entry wage rates at the extreme low end of the distribution of
actual rates of which they had knowledge, although in many cases workers have
actually been hired at these rates.
The wages offered for job vacancics reported by employers were compared with
the prevailing entry wage rates reported in the employm ent service offices. In
those cases in which employers indicated a range of rates for a vacancy, the
lowest rate offered was used for the comparison. For this reason, and because the
extreme low end of the distribution of actual entry wage rates was omitted, we
believe the analysis shows the maximum probable occurrence of vacancies offered
at below prevailing rates.
In the five cities the pattern was remarkably similar. The vacancies found
to be below prevailing rates ranged from about 15 to 20 percent of the total
vacancies reported in April 1965 surveys. Thus the results of this comparison,
which are indicative but not definitive, show that the great bulk of the vacancies
reported were at or above the prevailing entrance rates of pay.
Further study of wages below prevailing rates is necessary and will be made
in present and future surveys. For instance, in all of the current vacancy surveys
being made in fiscal year 1966, the rate of pay offered for a particular vacancy
is requested. These data will be analyzed in the light of whatever inform ation
may be available on the prevailing entrance wages in each occupation in each
industry. It may be possible, after considerable research on the subject, to work
out a basis for screening vacancies and to separate those offering less than going
rates from the others. Conceivably, with such an analytical tool an estimate
could be made of the approximate proportions of vacancies with less than prevailing
wage rates— a useful guide to interpretation of vacancy data.
Further study is necessary of other aspects of this issue. W hat is the relation
ship between hard-to-fill vacancies (i.e., those unfilled one month or longer)
and wage rates offered? Are such wage rates below prevailing levels? Or, as
JOB VACANCY STATISTICS
35
Charles H olt and M artin D avid 3 suggest, does ,f* * * a com pany that has been
seeking a particular type of worker for two months * * * behave quite dif
ferently, in terms of its * * * offering wage, than it did when it had been
recruiting for only a week?”
W hat about other conditions of employm ent such as unrealistically high educa
tional or experience requirements, or unfavorable hours of w~ork? N o way
has yet been devised for collecting an anaylzing information on all relevant con
ditions of em ploym ent within the framework of a viable statistical program.
Thus there is no lack of problems for further study to clarify the relationship
between wage rates and job vacancies. In an era of potential labor shortages
few matters are more w orthy of study.
Establishment sample and industry coverage
In the tw o rounds of pilot jo b vacancy surveys in fiscal year 1965 a prescribed
m ethod for the selection of a probability sample of employers in each area was
provided to the State agencies in order to make it possible to use the area sample
results to estimate total vacancies in each area. The universe from which the
sample was selected included all establishments with four or more employees which
were covered by the State unem ployment insurance law in the first quarter of
1964, supplemented by a list of known nonagricultural establishments employing
100 or more workers which were not covered by the State unem ployment insurance
law. Establishments with fewer than four employees, as well as farms and pri
vate households were not included in this experimental program because they
require the development of special techniques with regard to sampling, collection,
and estimation. Noncovered establishments with fewer than 100 employees
were included where this was considered necessary to assure adequate representa
tion in certain nonagricultural industries. The universe from which the sample
was drawn, therefore, generally extended across all nonagricultural wage and
salary payroll employm ent except for the very small establishments.
The sample size in each o f the 16 area experimental surveys was rather large,
since it included all of the larger establishments which, when arrayed by size
in descending order, had a cumulative employm ent total equal to 50 percent of the
area employm ent universe. In addition to this rather large certainty stratum for
larger establishments, the sample included a fixed number of smaller establish
ments, amonting to 1,000 in each of the four largest areas and 500 in each of the
remaining areas. In order to assure good industry and size mix among the
smaller establishments, sample members were selected at random from a listing
o f all of the smaller establishments in the universe (comprising the remaining 50
percent of the area em ploym ent universe) arrayed by em ploym ent size within
each tw o-digit standard industrial classification. Nine broad industry cate
gories and three establishment size groups were used to form sampling strata.
Estimated area totals of job vacancies were prepared for each three-digit
occupational classification for which job vancanies were reported. This was done
by inflating the area sample results to estimated area universe totals. The
inflation factor for each stratum was derived from the base period total em ploy
ment of the universe of establishments and o f the responding establishments
falling into that stratum.
Since the estimates in this survey are based upon samples, they may differ from
the figures that would have been obtained if it were possible to make a com plete
enumeration using the same schedules and procedures. The standard error is a
measure of sampling variability, that is, the variations of sample estimates from
a complete count that might occur by chance because only part of the universe
is surveyed. The chances are about tw o out of three that an estimate from the
sample would differ from a complete enumeration b y less than the standard error.
The chances are about 19 out of 20 that the difference would be less than twice the
standard error. In the April 1965 round of surveys in all 15 Standard M etropolitan
Statistical Areas the relative standard error of estimated total current jo b vacan
cies was 1.7 percent for all establishments; it was 0.5 percent for larger establish
ments and 3.4 percent for smaller establishments. The standard error ranged b y
area for all establishments in the April 1965 surveys from 3.3 percent for Rich
mond to 7.1 percent for Kansas City.
There is less precision in the estimates of job vacancies in occupations peculiar
to smaller establishments (e.g., barbers) than in occupations predominantly in
larger establishments, reflecting the higher degree of sampling error associated
3 Charles C. Holt and Martin H. David, “ The Concept of Job Vacancies in a Dynamic Theory of the
Labor Market/* in The Measurement and Interpretation of Job Vacancies, National Bureau of Economic
Research, New York, 1966.
36
JOB VACANCY STATISTICS
with estimation from a few small units wTith large sampling weight. In the current
round of surveys an effort has been made to lower the sampling error b y including
greater representation of small establishments.
Checking the accuracy of response to the pilot surveys
In the pilot projects in 16 areas in fiscal year 1965, the Bureau o f Labor
Statistics conducted response quality check surveys for both rounds o f jo b vacancy
surveys in several selected areas— 6 for the first round and 5 for the second.
In each o f these areas, trained and experienced BLS staff personally interviewed
a random ly selected subsample o f 611 establishments, or about seven percent o f
the respondents to the 2 rounds of job vacancy surveys immediately after their
completion.
The interviewers completed questionnaires and probed to get
meaningful information about the quality of reporting, response problems, com
pleteness and accuracy o f data reported, the kind of records used as sources for
the information, problems in providing the information, and problem s associated
with the feasibility of reporting and collecting meaningful job vacancy data on a
continuing basis by mail.
Highlights o f the response quality check surveys were as follow s:
1. The concepts, definitions, instructions, and schedules used proved satis
factory and effective and were generally understood b y the respondents; 95
percent o f the persons interviewed in the follow-up quality check surveys reported
no difficulty in understanding and no particular questions about the jo b vacancy
definition, or what to include and exclude.
2. As a corollary to the above, there was a high degree of accuracy in the
reporting. Nearly nine ou t of 10 establishments correctly reported their total
number of job vacancies.
3. The degree of cooperativeness and the positive attitude of respondents in the
survey were highly encouraging. A bout 85 percent of the reporters in the follow up quality check survey signified their willingness to report vacancies on a con
tinuing basis in the future; and only five percent indicated an unwillingness to do
so. A number of employers went further and expressed interest, during the re
sponse analysis survey, in the potential use of the jo b vacancy data for training
and guidance purposes with the expectation that such use would improve the
quality of the supply of labor available to meet their needs.
4. The total number of jo b vacancies was understated. A three percent under
statement (varying from an understatement of about seven percent for the larger
establishments to an overstatement of one percent for the smaller establishments)
of the number of vacancies was found as a net result of the detected error in report
ing by respondents. There may be some further undetected downward bias in re
porting owing to the lack of employer records from which jo b vacancies can be re
ported; it was found that establishments which had com pleted or partial records on
jo b vacancies were more prone to report vacancies. Smaller establishments are
less likely to keep records and hence more likely to have undetected downward
bias than were the larger establishments.
5. Questions in the response quality check surveys elicited information on the
positive actions that employers had taken to fill the specific job vacancies they had
reported. On the average, for each occupation for which vacancies were reported,
two positive actions were undertaken by employers to fill the vacancies. About
one-third of these actions consisted of newspaper and other advertising, while
another third were requests to public and private em ploym ent services.
Small random samples of nonrespondent establishments were also interviewed
as part of the follow -up response analysis surveys. The primary purpose was
to determine the extent and direction of any bias that might be introduced in
the estimated total of jo b vacancies by assuming the same vacancy rate for re
spondent and nonrespondent establishments.
Interviews conducted with six percent of the nonrespondents in the 11 area
surveys disclosed that nonrespondent establishments, both large and small, had
substantially higher levels of current job vacancies than did the respondents.
In consequence, it was found that the level of vacancies estimated for nonrespond
ent establishments would have been understated by about 35 percent if it had
been assumed that there was no difference in the rate of job vacancies between
respondents and nonrespondents.
The seriousness of this understatement b y nonrespondents is materially de
creased in view of the fact such establishments represented less than one-fourth
of all surveyed establishments in the 11 area response analysis surveys. Hence,
combining the detected error for both respondents and nonrespondents reveals
an over-all understatement of 12 percent.
JOB VACANCY STATISTICS
37
M ethod of collection: There still remains some question as to how inform ation
on jo b vacancies should actually be collected. This year, the D epartm ent’ s
experimental survey program includes two additional rounds of jo b vacancy
surveys— one at the end of M arch 1966 and the other at the end of M ay 1966.
In 12 areas, the experimental surveys parallel those conducted in fiscal year
1965, which collected jo b vacancy inform ation independently from any existing
data program. In tw o States, however, a new approach is being tried. Pilot
surveys are being conducted in these States to test the feasibility of collecting
jo b vacancy inform ation through the cooperative Labor Turnover Statistics
system currently operated by the Bureau of Labor Statistics, the Bureau of
Em ploym ent Security and the affiliated State em ploym ent security agencies.
This test collection of jo b vacancy data in conjunction with labor turnover in
form ation is also being extended to industries other than manufacturing and
mining, which are not now covered by the cooperative Labor Turnover Statistics
program. This m ethod is being tested in the belief that it may offer analytical
advantages, as well as operating economies since turnover and vacancies can be
usefully studied togetcr.
The Departm ent now has a fiscal year 1967 budget request before Congress for
$2.5 million to launch a large scale job vacancy program in January 1967. The
program would cover 75 or 80 Standard M etropolitan Statistical Areas, and
possibly all 50 States, at quarterly intervals. The m ethod used to implement the
program will depend on the outcom e of the fiscal year 1966 experiments.
One alternative would be to institute an independent quarterly collection of
job vacancy inform ation in the 75 or 80 areas. Another would be to use the exist
ing m onthly collection program of labor turnover data (extended to cover all
nonfarm industries) as a vehicle to collect jo b vacancy inform ation on a quarterly
basis in all 50 States. This would thus make both jo b vacancy and labor turnover
data available for all 50 States and the Nation, as well as for 75 or 80 areas.
The Departm ent will study the results of the pilot studies and make a decision
later this year as to the technical option to be followed.
IV . CONCLUSION
In conclusion, let me say that although the conceptual and technical problems
described above are important, they are not of such magnitude as to prevent the
statistics collected in a national program from being useful for analytical and
operational purposes. W e have come a long way in the past few years in indentifying and resolving problem areas. We have examined in detail arguments for and
against a program of job vacancy statistics. We feel that the need for job vacancy
data heavily outweighs the problems involved. An ongoing and comprehensive
jo b vacancy program can greatly assist in the implementation of an active man
power policy and the development of econom ic analysis needed for major policy
decisions.
Selected
B ib l io g r a p h y
on
J ob V a c a n c ie s
1. Bevridge, William A., “ The Cyclical Behavior of H elp-W anted Advertising
and Selected Labor Turnover Variables,” Annual Proceedings, American Statis
tical Association, 1961 p. 31 ff.
2. Chavrid, Vladimir C., “ An Approach to Job Vacancy Inform ation,” Employ
ment Service Review, 1964.
3. Dow, J. C. R., and Dicks-Mireaux, L. A., “ The Excess Demand for Labour,
A Study of Conditions in Great Britain, 1946-56,” Oxford Economic Papers,
February 1958, pp. 1-33.
4. Edelman, M urray, Channels of Employment, Urbana, Illinois, 1952.
5. Ferber, R obert and Ford, Neil, “ The Collection of Job Vacancy Data
Within A Labor Turnover Framework,” in Employment Policy and the Labor
Market, Arthur M. Ross, ed., Berkeley, 1965, pp. 162-190.
6. Gainsbrugh, M artin R., “ The Need for Job Vacancy Measures,” Voluntary
and Involuntary Unemployment, National Industrial Conference Board, Public
Affairs Conference Report Number 1, New York, 1964.
7. Hawk, Em ily and Toliver, Evelyn, “ Indiana Job Vacancy Surveys,” Employ
ment Service Review, April 1965.
8. Illinois Departm ent of Labor and Governor’s Com mittee on Unemployment,
Illinois Job Seekers Survey (2 Vols.), 1962.
9. Joseph, M yron L., “ Current Surveys on Measuring Job Vacancies,” Pro
ceedings of the Business and Economic Statistics Section, American Statistical
Association, 1955, pp. 306-316.
38
JOB VACANCY STATISTICS
10. Malm, F. T., “ Recruiting Practices and the Functioning of Labor M arkets,”
Industrial and Labor Relations Review, Vol. V II, July 1964, pp. 507-525.
11. Myers, John G., “ Can Y ou Measure Job Vacancies?,” The Conference
Board Record, M ay 1965, pp. 50-57.
12.
, “ M ore on Measuring Job Vacancies,” The Conference Board Record,
September 1965, pp. 23-31.
13.
, “ Measuring Job Vacancies: The Third Survey,” The Conference
Board Record, N ovem ber 1965, pp. 27-36.
14. National Bureau of Econom ic Research, The Measurement and Interpreta
tion of Job Vacancies, New York, 1966.
15. President’s Committee to Appraise Em ploym ent and Unemploym ent
Statistics, Measuring Employment and Unemployment, Washington, D .C ., 1962.
16. Shelton, William C., and Neef, Arthur F., Foreign Job Vacancy Statistics
Programs, U.S. Departm ent of Labor, Washington, D .C ., D ecem ber 1964.
17. Slotkin, Elizabeth J., Feasibility Study of Problems in the Collection of Data
on Job Vacancies, Illinois Bureau of Em ploym ent Security, M arch 1964.
18. Zellner, Wellsey E., “ Portland Studies Job Vacancies,” Employment Service
Review, April 1965.
Chairman P r o x m i r e . Thank you, Commissioner Ross. This is
very encouraging testimony.
And my questions both to you and Mr. Cassell and Mr. Chavrid
certainly do not indicate that I am opposed to this program. I
am all for it. But I do want to do my best to bring out some of the
problems that I think we are likely to run into.
Mr. Ross. We do appreciate the continued interest of the commit
tee in this program, and we feel it has been very helpful.
Chairman P r o x m i r e . Thank y o u v e r y much.
Now, you have said something that I think ought to be underlined
and repeated, especially coming from you. You say—
The lack of job vacancy information constitutes the most significant gap in our
knowledge of labor market conditions. Statistics on job vacancies would give
us a measure of unsatisfied demand for labor which, together with our data on
employm ent, would provide a more complete measure of demand for labor—
something we have never had before.
Mr. Ross. Yes, sir.
Chairman P r o x m i r e . And I think that coming from you, and with
your reputation, and with your authority, that should be quite per
suasive with Congress.
You indicate also that the pilot program shows that the employer
can and will reach reasonably accurate estimates. Can you elaborate
a little on that, how they show that?
Mr. Ross. Yes, sir.
Chairman P r o x m i r e . I think you show earlier that there has been
a considerable question. The Bureau's initial reaction was that the
employer did not have sufficiently accurate records from which to
make estimates on vacancies.
Mr. Ross. The 16 pilot survej-s were made cooperatively with the
Bureau of Employment Security, the Bureau of Labor Statistics, and
the States and localities. The BLS conducted establishment response
surveys for the 16 areas that we tested. An experienced BLS staff
personally intereviewed 611 establishments to see if the responses were
good. And this was about 7 percent of the total respondents in the
two rounds of vacancy surveys.
We found that the concepts definition and instructions and sched
ules were satisfactory, and were generally understood by the respond
ents. Ninety-five percent of the employers said they had not had
any difficulty in understanding what we meant by a vacancy. And
JOB VACANCY STATISTICS
39
what we should include or exclude. There are very careful instruc
tions on what to include and what not to include in reporting the
vacancy.
We found that 9 out of 10 establishments correctly reported the
total number of vacancies. In other words, our people agreed with
them as to the count of vacancies.
We found a high degree of cooperativeness from about 85 percent
of the reporters and the foil ownp check signified their willingness to
report vacancy on a continuing basis in the future. Only 5 percent
said they were unwilling.
We have in mind quarterly surveys of vacancies.
Chairman P r o x m i r e . But you also found, as I understood you to
say, an understatement of vacancy of something like 12 percent?
Mr. Ross. Yes.
Chairman P r o x m i r e . H o w did you arrive at such a precise estimate
of the understatement?
Mr. Ross. A 3-percent understatement was found as a net result
of detected errors in reporting by respondents.
Chairman P r o x m i r e . And you always made this kind of correction?
Mr. Ross. We have not corrected the figures. Those are so far
experimental.
Chairman P r o x m i r e . In general, would you have a correcting
percentage?
Mr. Ross. No. I think the attempt would be to iron out these
response errors. I would say that for the first couple of surveys errors
of this magnitude are not discouraging. And we would attempt to
iron them out. Of course, some employers do not yet have good
records. And part of the task would be to get the employers to
establish records to pinpoint the number of vacancies. Some em
ployers have rather decentralized hiring procedures. And for them to
cooperate it would be necessary for them to put together from their
different hiring points the information on vacancies.
Then we say there may be some further undetected downward
bias in reporting due to the lack of employment records. It was
found that the employers which had complete records gave us better
and more accurate counts than those that didn’t have complete
records. We found the smaller establishments are less likely to keep
complete records. And we feel that there was more underreporting
on the part of the smaller establishments than the larger establish
ments.
Chairman P r o x m i r e . H o w large a sample would you feel would
be necessary to get an accurate picture?
Mr. Ross. We think that all of the large firms-----Chairman P r o x m i r e . The few largest firms, or something of that
kind?
Mr. Ross. All of the large firms and a good sample of the small
firms in each industry.
Chairman P r o x m i r e . H ow do you define large firms, how big?
Mr. Ross. I think Mr. Chavrid could answer that better, Mr.
Chairman.
Chairman P r o x m i r e . I will wait until he testifies.
Mr. Ross. Now, combining the detected error for both the re
spondents and the nonrespondents—we had to account for whether
respondents would have less or more reporting error than the non
respondents—revealed that overall understatement of 12 percent.
JOB VACANCY STATISTICS
40
This, as I say, is not discouraging for these pilot surveys. And I
believe that through getting more employers to cooperate— and 85
percent is again a very encouraging tally the first couple of times
around—by keeping systematic records, this understatement could be
reduced.
Chairman P r o x m i r e . Is there any area that you feel has to be ex
cluded from this? Would it include government? Would it include
Federal, State, and local government—all three? Are there any pe
culiar problems involved in the governmental sector which would be
quite different?
Mr. Ross. I think that Mr. Chavrid could give you a more intelli
gent response.
Chairman P r o x m i r e . How about the Armed Forces? Is that also
in this area?
Mr. Ross. No, this would be civilian employment.
Chairman P r o x m i r e . It would not include Armed Forces?
Mr. Ross. No, because I think their vacancies take the form of
draft calls, and we get pretty good information on that.
Chairman P r o x m i r e . H o w about the problem—you have men
tioned this, but I am not sure how you correct for it—of a very large
agricultural area where people would like to have hired hands, but
figure that they obviously cannot hire them, because nobody will
work for the wage that a farmer can afford to pay?
Mr. Ross. These surveys are taken in metropolitan areas. And
the program which is now before Congress would make it possible to
extend the surveys from 16 to about 80 metropolitan areas. They
would, therefore, not cover agriculture.
Chairman P r o x m i r e . N o w , you are asked to make estimates of
labor shortages for the President.
Mr. Ross. To make an appraisal, I do not make theoretical esti
mates.
Chairman P r o x m i r e . H o w did you do that without these statis
tics?
Mr. Ross. Well, we do the best we can with what we have. I
think the reports are respectable, but we could do a lot more if we had
vacancy statistics.
Chairman P r o x m i r e . You say they are respectable now?
Mr. Ross. If you like, I would be glad to supply for the record
copies of the last three manpower shortage reports which I have made
public at the request of the President.
Chairman P r o x m i r e . Please do so.
(Reports referred to follow:)
T h e M a n p o w e r S it u a t io n : E m e rg in g S h o r ta g e s an d R e s id u a l S u r p l u s e s 1
(B y Arthur
M.
Ross, Commissioner, Bureau of Labor Statistics)
The N ation’ s manpower situation is in transition. Civilian em ploym ent has
risen by 2.4 million during the year, while unem ployment has fallen by 600,000.
The seasonally adjusted unemployment rate for December, at 4.1 percent, was
at the lowest point since M ay 1957. The rate was 2.6 percent for adult men and
only 1.8 percent for married men.
Clearly the picture is changing, but it is im portant to see it in perspective.
Opinions are widely expressed that a general labor shortage has arrived or is
imminent. W hat are the facts?
1 Statement issued with release of Summary Employment and Unemployment Estimates: December 1965.
JOB VACANCY STATISTICS
41
1. The current situation shows stringency in some areas and occupations and
in a fewT industries. There is still no evidence of a general shortage of labor.
2. Geographically, the tightest job markets are found in the Great Lakes area
(e.g., Chicago, Detroit, Milwaukee, Cincinnati, and Cleveland), where the heavy
manufacturing industries are centered. Estimated unem ployment rates are in
the neighborhood of 2 or 2J/£ percent in such areas. Job markets are considerably
looser in many New England and West Coast areas (e.g., Fall River, Lawrence,
Los Angeles, San Diego and San Francisco-Oakland). There is every indication
that unemployment in some of the Negro ghettos (such as those of Watts, Harlem
and W est Oakland) remains unacceptably and dangerously high.
3. The labor situation appears to be tightest in the metal-working machinery
and construction industries. Elsewhere there is little clear evidence that pro
duction schedules are being delayed by manpower shortages. D efense-production
industries such as aircraft, ordnance and electronics have been able to find needed
workers except that difficulties are experienced in recruiting skilled personnel such
as engineers, scientists, mathematicians, tool and die makers, and electronic
maintenance workers.
4. M anpower shortages are heavily concentrated in the professional and skilled
occupations such as those just mentioned. Other widely-sought skills are those
of machinists, turret-lathe and milling-machine operators, sheet-metal workers,
some construction workers, shipfitters, boilermakers and arc welders. There are
shortages of physicians, nurses, hospital attendants and other health-related
personnel. Low-paying jobs in some industries are difficult to fill.
5. As noted above, the unem ploym ent rate for adult men is low. Yet there
were 1,249,000 jobless men in Decem ber, mostly seeking fulltim e employm ent.
There were about 838,000 women out of work, with 80 percent desiring fulltim e
jobs. There were 800,000 unem ployed teenagers, with about 45 percent wanting
fulltime jobs. Am ong the unem ployed there w~ere 600,000 who had been out of
work 15 weeks or more. The unem ployment rate for Negroes remains twice as
great as that for whites.
6. It should also be kept in mind that the unem ployment would be considerably
higher if hundreds of thousands were not enrolled in manpower and social programs,
such as Neighborhood Y outh Corps, Job Corps, the w ork-studv program, etc.
At the same time, people in these and related programs such as M D T A and
vocational education are acquiring the qualifications which will enable them to
take regular jobs and help avert manpower shortages in the near future.
7. Thus we do not have a general labor shortage but rather a mixture of emerg
ing shortages in a few areas, occupations and industries; residual surpluses in
others, and an approximate balance in the remainder. A t the same time, it seems
clear that manpower demands will intensify during 1966 as a result of increased
defense production, continuing capital expansion and greater numbers of men in
military service.
8. To meet this new situation, the first step is to know the problems in detail
and in depth. A t the initiative of Secretary Wirtz, the Departm ent of Labor is
working with the Departm ent of Defense, the Departm ent of Com merce and other
agencies in order to ascertain specific manpower requirements, shortages and
surpluses in specific areas, occupations and industries. Conferences will be held
with representatives of several industries and unions in the near future.
9. These efforts, it is hoped, will facilitate more effective use of vocational
training, on-the-job training, apprenticeship training, em ploy ment-ser vice
operations, inter-area recruitment and worker relocation in order to remedy skill
shortages. But other measures are also needed. In view* of the severely limited
supply of adult men, some jobs will have to be redesigned so that women, older
workers, young people and part-time jobseekers can perform them. Unneces
sarily stringent hiring specifications with respect to education, age and experience
will have to be relaxed. Transportation facilities should be im proved in some
metropolitan areas where low -incom e neighborhoods are quite distant from
available jobs. Wages should be more com petitive in certain poorly-paid yet
essential occupations. Needless to say, equal em ploym ent opportunity is more
essential than ever.
10. In conclusion, the period just ahead will test our ability to use our human
resources to an even fuller extent in order to meet production schedules and main
tain satisfactory econom ic growth. Likewise it will test our capacity to eliminate
potentially inflationary manpower bottlenecks as they develop.
[News from U.S. Department of Labor, Friday, April 8, 1966]
N o overall labor shortage exists in the U.S. today but there are some “ im
balances” between labor supply and demand.
42
JOB VACANCY STATISTICS
A t the same time, there also are “ considerable” untapped reserves of manpower
and the governm ent itself can accomplish only part of the task of matching up
manpower requirements and human resources. Private industry must do a great
deal more to provide training and upgrading for its own workers in order to pre
pare them for greater responsibilities.
These are m ajor points made in a “ Report on M anpower Shortages and Re
serves’ ’ b y Commissioner Arthur M . Ross of the Labor D epartm ent’s Bureau of
L a b or Statistics.
The report comes in response to President Johnson’s directive that the Com
missioner provide m onthly “ the fullest possible” inform ation on possible man
pow er shortages.
The full-scale manpower review— the first of its kind— reveals that—
In some areas, such as Boston, Cincinnati, Los Angeles, Seattle and St.
Louis, skilled workers, engineers and scientists are hard to find.
Defense industries have accounted for about one-fourth of the em ploym ent
gain in manufacturing over the past year.
W ith employm ent growing by 900,000 since last December, “ sp ot”
manpower problems “ are being met up to now.”
R
eport
on
M
anpower
Sh o r tag es
and
R
eserves
(By Arthur M . Ross, Commissioner, Bureau of Labor Statistics, U.S. D epartm ent
of Labor)
I. INTRODUCTION
In his recent M anpower R eport to the Congress, President Johnson directed
that “ the Commissioner of Labor Statistics should include in the em ploym ent and
unem ploym ent reports, beginning in March, the fullest possible inform ation on
m anpower shortages, both actual and threatened.”
Since this report, covering the situation in the month of M arch, is the first to
be issued, a brief statement of its purposes may be helpful. The principal pur
pose is to provide accurate Up-to-date information on manpower shortages, as
well as surpluses, to employers, union officials, education and training authorities,
and the general public. Understanding of the manpower situation in general is
im portant in order to gauge whether expansion of production and em ploym ent
m ay continue as in the past. Identification of manpower shortages in specific
areas, industries, and occupations should be helpful in gearing up recruitment,
training, upgrading and related policies to meet the requirements m ost effec
tively. Likewise, knowledge of the characteristics of available workers will
assist employers in making necessary adjustments in their personnel policies.
M atching up manpower requirements and human resources in an indispensable
step in maintaining econom ic progress, avoiding inflationary bottlenecks, elimi
nating hardship and achieving full employm ent. In the last analysis, the gov
ernment itself can accomplish only a small part of the task. It is therefore essen
tial that employers, workers, union officials, educational and training authorities
and the general public be well inform ed concerning the manpower situation.
M uch of the data concerning areas and occupations was provided by the
United States Em ploym ent Service of the Bureau of Em ploym ent Security and
the affiliated State Em ploym ent Services. Industry data were supplied in large
part by the Business and Defense Services Administration in the Departm ent of
Commerce. Other cooperating agencies include the Bureau of Apprenticeship
and Training, the Office of Emergency Planning, the Departm ent of Defense and
the Council of E conom ic Advisers. Needless to say, statistics and analysis by
the Bureau of Labor Statistics have been drawn upon.
General Indicators of the Manpower Situation
Employers, it is evident, are finding it increasingly difficult to locate some types
of workers. W ant-ad columns are growing longer, unfilled vacancies are more
numerous, and widespread complaints are heard that applicants do not have the
proper qualifications or experience.
I t does not follow, however, that we are suffering from a general manpower
shortage or an exhaustion of manpower reserves. The explanation is that there
are imbalances between manpower supply and demand which are capable of
JOB VACANCY STATISTICS
43
being remedied b y private and public efforts. The untapped reserves of labor
supply are also considerable.
The general indicators of the manpower situation, as of M arch, are consistent
in showing that manpower problem s exist in some occupations, industries and
areas, but that we do not have an overall shortage of labor.
A. The Employment Trend
During each of the past several months, payroll em ploym ent has increased
about 300,000 more than the seasonal expectation. The M arch 1965-M arch
1966 increase, at 2.9 million, was nearly a third greater than the increase over
the same months from 1964 to 1965. This em ploym ent advance could not have
been possible if there were an overall manpower shortage or if skill bottlenecks
had become unbreakable.
B. The Unemployment Rate
For the better part of a year now, the national unem ploym ent rate has been
moving downward at the rate of about 0.1 point per month, from 4.8 percent in
April 1965 to 3.8 percent in M arch 1966. While the im provem ent has been very
great, the rate is still higher than the 3.5 percent of N ovem ber 1953 or the 2.6
percent of M ay, June and July, 1953.
In M arch, the total number of unem ployed was 3.0 million. Of these, about
I.4 million were adult men, and all but 75,000 of them were seeking full-tim e
work. There were 900,000 unem ployed women, more than 80 percent of whom
were seeking full-time jobs, and 750,000 teenagers, about half looking for full
time work. Even though unem ployment among men is still close to a seasonal
peak and can be expected to drop sharply in the next 2 months because of the
pickup in seasonal work, there still remains a considerable reservoir of workers
available for em ploym ent expansion.
C. Average earnings and hours of work
In a period of severe manpower shortage, average earnings escalate as a result
of heavy overtime assignments, premium wage rates and loose classification o f
workers. Data on average hourly earnings of factory workers ($2.68 in M arch
1966 as compared with $2.59 in M arch 1965) indicate that these practices are
not widespread. The average workweek in manufacturing was 41.5 hours in
M arch 1966 and 41.2 hours a year previously.
Hours of work have m oved up more rapidly, however, in some industries where
the manpower situation is tight. Examples include the instrument industry,
where hours of work rose from 41.2 in M arch 1965 to 42.3 in M arch 1966; the
aircraft and parts industry in which hours rose from 41.8 to 44.0; and ship and
boat building and repairing, in which hours rose from 40.1 to 42.1. In other
industries such as metalworking machinery and screw machine products, hours
of work have showed smaller increases over the past year, but are at extremely
high levels.
In the bulk of industry, hours of work have only risen slightly and are still
not at critically high levels. In industries such as primary metals, fabricated
metal products, apparel, and rubber and plastic products, hours of work in
M arch 1966 were the same or lower than in M arch 1965, indicating that workers
were still available in sufficient quantity to make unnecessary any increase in the
workweek.
D. Labor turnover
Rising quit rates in manufacturing indicate that workers are able to find abun
dant alternate opportunities for employm ent. Over the February 1965-February
1966 period, quit rates in manufacturing rose from 1.7 percent to 2.4 percent
(seasonally adjusted). However, the rate was unchanged from January to Feb
ruary, and stiil was w^ell below the rates of the Korean conflict and W orld W ar
II. The rate of accessions (the rate at which employees are hired) also increased— ■
from 4.0 percent (seasonally adjusted) in February 1965, to 4.7 percent in Febru
ary 1966, reflecting the fact that employers in general are able to find additional
workers to expand their payrolls.
E. Manufacturer’s orders and shipments
While manufacturer’s orders and shipments are rising, the ratio of shipments
to unfilled orders has not changed particularly in most industries. This means
that producers are keeping up with the flow’ of new business. There are only a
few authenticated instances in which production schedules are being missed
because of manpower shortages.
JOB VACANCY STATISTICS
44
F. Price trends
Recent consumer price increases— such as those of meat and poultry, apparel
and footwear, oil and gasoline have not been connected with manpower shortages,
for the most part. City consumer price indexes do not show a larger than average
increase in communities, such as those in the Great Lakes region, where qualified
labor is hardest to find. In some of the lower paid consumer services, however,
labor costs are rising as employers find it necessary to make their jobs more
attractive in order to compete for manpower in a period of high em ploym ent.
II.
TH E
M ANPOW ER
S IT U A T IO N
IN
LABOR AREAS
The m ost recent reports indicate that about one-third of the 150 m ajor metro
politan areas have relatively tight manpower situations. Those areas reporting
the lowest unem ployment rates and the most severe hiring difficulties are mainly
in the industrial, durable goods producing sections of the M idwest and East.
The lowest rates were reported for Flint and Lansing, M ichigan; Fort Wayne,
Indiana; Cedar Rapids, Iow a; and Reading, Pennsylvania. Larger areas with
tight situations include Detroit, Chicago, Minneapolis, Cincinnati, Cleveland and
Milwaukee. States with the greatest concentration of shortage areas include
Indiana, Iowa, Michigan, Pennsylvania, and Georgia.
On the other hand, there are numerous labor market areas with unem ployment
rates of 5 or 6 percent or more. The bulk of these are in California, Massachu
setts, Pennsylvania, and West Virginia. They include Stockton, Fresno, San
Bernardino, San Diego, Lowell, Fall River, Atlantic City, Duluth-Superior,
Altoona, Scranton, Wilkes-Barre-Hazelton, and Wheeling.
Defense Plants
Special studies have been made of a dozen areas in which defense-related plants
with “ mass hiring needs” (up to 15,000 workers in individual cases) are located.
These plants are engaged in the production of aircraft, ships, ammunition and
explosives. On the basis of studies by the United States Em ploym ent Service,
it can be expected that eight of the plants will not have serious difficulty in meeting
their 1966 manpower requirements. The reasons for optimism are that the firms
are among the higher-paying employers in their communities; the working con
ditions are above average; the bulk of the needs are for semi-skilled workers, who
can be trained and upgraded on the jo b ; and the plants are not located in tight
labor areas.
Of the remaining four plants, tw o are likely to have moderate problems and
two (which are located in Seattle and Cincinnati) will encounter more serious
difficulties.
The Situation in Five Specific Areas
It would not be practical to summarize the manpower situation in all 150
m ajor areas. As an alternative, job market conditions will be described for a
few of the more im portant areas. The purpose will be to give a sample of looser
as well as tighter areas in various parts of the country.
A. Boston.— In the early months of 1966, unem ployment in Boston was the
lowest for any year since 1957. M any industries are finding it more difficult
to recruit qualified workers. Three out of every five vacancies listed at public
em ploym ent offices have been unfilled for 30 days or more, with the largest
number being in medical and health service occupations._
Governm ent-supported training activities include the following: There are
2,437 trainees enrolled in M D T A courses, and additional courses for 732 work
ers have been approved. Over 1,000 apprenticeship opportunities have been
developed in recent months. New OJT contracts will absorb 318 trainees, and
two Y outh O pportunity Centers will accom m odate 700.
B. Cincinnati.— Labor supply has been tightening in this manufacturing cen
ter, with the unem ployment rate the lowest in recorded history. The most press
ing needs are for clerical, professional and skilled manual workers. The shortage
of skilled metal workers is so great as to cause possible delays in production
schedules. Yet, there are surpluses of material handlers, custodial workers,
female assemblers, and cafeteria workers. There were also unemployed construc
tion laborers during the winter.
C. St. Louis.— The job market in St. Louis has also been tightening, although
n ot yet to the same extent as in Cincinnati. There have been shortages of highly
skilled workers during the past year, which have been met to some extent by
recruitment from other areas. A t the same time there are untapped reservoirs
of unskilled and entry-level workers, particularly in East St. Louis.
JOB VACANCY STATISTICS
45
D. Seattle.— The unem ployment rate in Seattle is higher than the national
average. Despite recent gains in aircraft employment, the number of workers
in Seattle’ s leading industry is less than it was in 1962.
By the end of 1965, about 90 percent of the unfilled job s listed at the*local
unem ployment office had been vacant for 30 days or more. A bout half of these
were for skilled workers in aircraft. In addition, there were critical openings
for engineers, draftsmen, and engineering technicians which could not be filled
from within the community.
Looking ahead further into 1966, the supply of unskilled workers appears to
be more than adequate. Nevertheless, real difficulties will be encountered be
cause so many of the new jobs will be for well trained and experienced workers
already in short supply.
E. Los Angeles.— The highest unem ployment rate in these five areas is found
in Los Angeles-Long Beach, partly as a result of continuing in-migration of
workers. Recent surveys have confirmed the alarming extent of joblessness in
Watts. The home-building industry has been retarded b y the decline in housing
starts, despite the fact that winter ordinarily has little effect on construction
activity in Southern California.
A t the same time, shortages of qualified workers in nationally scarce occupations
have been intensifying. Unmet manpower needs are concentrated in aerospace,
resulting from increased orders for both military and civilian aircraft. Aside
from engineers, technicians and skilled metal tradesmen, health service workers
are also in short supply.
The public employm ent service has placed about 2,200 jobs in interarea recruit
ment. Greater emphasis is being placed on M D T A and OJT programs, drawing
in some measure from poverty groups. Em ployers are being urged to reduce
hiring specifications, redesign their jobs to use more women and lesser skilled
individuals, and recruit outside the area.
The manpower situation is quite different in these five areas. They range from
Cincinnati, classified as rather tight, to Los Angeles, regarded as rather loose.
There are almost 10 times as many unemployed persons in Los Angeles as in
Cincinnati, and the unem ployment rate is at least tw o points higher. Yet there
is a basic similarity among these areas in one respect. Skilled workers, engineers
and scientists are hard to find, while there are unused reserves of less skilled, less
experienced and otherwise less attractive job candidates. The extent of the
mismatch varies from one com m unity to another; but it is plain that the com m on
problem is one of imbalance, rather than overall labor shortage. It is the task
of private and public manpower policies to remedy these imbalances in order
that employm ent growth may continue in orderly fashion.
III.
THE
M AN POW ER
S IT U A T IO N
IN
S P E C IF IC
IN D U S T R IE S
This section of the report will analyze the situation in several industries which
are of concern because they are heavily involved either in defense production or
in the manufacture of capital goods. Included are some of the industries which
have voiced frequent complaints concerning manpower shortages.
A.
Metalworking Machinery.— Recent data indicate no real let-up in the tight
ening manpower situation. Although employm ent fell by 600 to 315,500 in M arch,
average weekly hours and overtime hours continued to rise, and were significantly
above the levels of one year ago. Average weekly hours in M arch 1966 were
46.9 up 0.8 hour from M arch 1965. Overtime hours reached 8.1 in February
1966 (the latest month for which data are available), up 1.0 hour from February
1965 and the highest level for the month since the series began in 1958.
In the past three years the metalworking machinery group has shown the most
significant increases in unfilled orders backlog of the individual manufacturing
industires. In recent months, however, the backlog situation has eased some
what. The ratio of unfilled orders to shipments rose from an annual average of
4.2 months in 1963 to 7.1 months in 1965, with the greatest change occurring
between 1963 and 1964. The peak month was reached in June 1965, when a
7.6 months backlog was recorded by the group. Since that time, the backlog
has declined somewhat, and in January 1966 the ratio was 6.7 months, as compared
with 7.3 months for December 1965. Nevertheless, in view of the high skill
levels of workers in this industry, the labor supply situation may be one of the
N ation’s most critical.
Am ong the specific segments of the industry, machine tools (and especially
metal-cutting and metal-forming tools) have had the greatest increases in the
63- 947— 66 ---------- 4
46
JOB VACANCY STATISTICS
backlog of orders. M etal-cutting tools had a backlog of 7.5 months in D ecem ber
1965, while metal-forming tools had a backlog of 10.2 months.
Other machinery industries or industry groups show different movements in
the ratio of unfilled orders to shipments. The engine and turbine industry has
shown a rising rate since 1963; the annual average was 8.2 months in 1963, 8.5 in
1964, and 10.0 in 1965. In January 1966, however, the rate declined to 10.8
months, as compared with 11.1 months in January 1965. The ratio for General
Industrial Machinery also has been rising consistently since 1963 and is still in
creasing. The annual rate for 1965 was 2.9 months as compared with 2.4 months
in 1964. The upward trend became sharper in the latter part of 1965 and has
continued into this year. In January 1966, the backlog was 3.4 months as com
pared with 2.5 months for the same month last year.
B. Iron and Steel Foundries.— Em ploym ent in iron and steel foundries increased
by 500 workers in M arch to 234,000. This relatively small increase in em ploy
ment in the face of continued rises in demand for the industry's products may
indicate problem s on the horizon. However, in M arch 1966 the average work
week was down from year-ago levels (43.2 hours this year, as compared with 44.8
hours in M arch 1965). Average overtime hours, at 5.6 hours in February 1966
were up slightly from January, but unchanged from a year earlier.
C. Construction Equipment and Machinery.— D ata for this industry indicate a
tight but not serious manpower situation. Em ploym ent in M arch 1966, at
261.000, was up 14,500 over M arch 1965. The workweek— at 43.3 hours in
M arch 1966— was up yi hour from M arch 1965. Overtime hours were also up
somewhat over 1965 levels, rising 1.1 hour from February 1965 to February 1966
(from 3.9 hours to 5.0 hours).
D. Defense Industries.— Em ploym ent in defense industries— aircraft, com m u
nications equipment, electronic components and accessories, ordnance and
ammunitions, and shipbuilding and repairing— rose by 27,700 in M arch to
1.965.000. A t that level, employm ent was up by 278,000 from a year earlier.
Defense industries accounted for over a fourth of the gain in manufacturing
employm ent.
W ithin the broad group, there were some divergence in trends. Em ploym ent
in Communications Equipment rose slightly in March, continuing the uptrend of
recent months. A t a level of 462,000, employm ent was up 43,000 over the year
and above its 1962 peak. However, the average workweek declined slightly in
M arch to 42.0 hours, but was still 0.6 hour higher than a year earlier.
The Ordnance Industry continues to show the marked expansion begun in mid1965. Between M arch 1965 and M arch 1966, em ploym ent increased by 27,000,
with three-fourths of the advance occurring among production workers. The
(seasonally adjusted) factory workweek, declined in M arch, from 42.4 hours to
42.3 hours, and was not at critically high levels.
In the Aircraft Industry the situation remains tight but not critical. E m ploy
ment rose by 14,700 between February and March, but still remained substan
tially below 1955-59 levels. A t 44.0 hours in March, the average workweek was
2.2 hours higher than a year ago. Average overtime (at 5.3 hours in February
1966) was down from January, but was significantly higher than the level of
February 1965. A recent survey indicated that em ploym ent in the aircraft
industry is expected to increase by nearly 50,000 between D ecem ber 1965_and
June 1966, intensifying pressures on the industry.
D ata on the backlog of unfilled orders for “ defense products industries,” show
some easing of the situation in late 1965 and early 1966. The group as a whole
showed considerable m onth-to-m onth variation w ithout any particular trend until
the beginning of 1965, when the backlog rate rose from 9.1 months to a peak of
10.5 months in September. The backlog has since declined and stood at 9.6
months in January 1966.
There were some disparate movements in the backlog among the industries
in this group over the past year, but in late 1965 and early 1966 some reduction
in the backlog occurred for all the industries. The communication equipm ent
industry’s backlog reached a peak of 6.7 months in July and has been declining
since then. By January 1966 the rate was 5.8 months, as compared with 6.2
months a year ago. The aircraft and parts industry showed an upward trend
since the beginning of 1965, rising from 11.0 months at the beginning of the year
to a peak of 13.2 months in N ovem ber 1965. However, in Decem ber 1965 and
January 1966 the rate dropped, to 12.2 months. This would indicate that they
are able to find the equipment, materials and labor to hold their own.
E. Screw Machine Products.— Some 2,000 companies in this industry manu
facture custom-m ade com ponent parts for nearly all hard-goods manufacturing.
JOB VACANCY STATISTICS
47
Serious shortages of skilled labor have been reported. Orders booked in the first
11 months of 1965 were 27 percent above 1964. T o meet the increased demand
for their products, companies have extended the workweek and have accelerated
recruitment and training. They also indicate that the shortage o f skilled workers
in the industry may restrict the flow of com ponent parts in the near future.
Em ploym ent reached an all-time high of 99,000 in March. The gain from a
year earlier totaled 7,200, with the bulk o f the new jobs falling in the production
worker category. The steady upward trend in employm ent has been in evidence
since late 1964 and has been accompanied by a sharp rise in the workweek. In
M arch, the workweek averaged a relatively high 45.0 hours, down slightly from
February, but up 0.7 hour from a year earlier. Similarly, overtim e was up
sharply, showing an increase of 1.5 hours between February 1965 and February
1966 (to a record February level of 7.1 hours).
Although data on prices for the specific industry group are n ot available, the
Wholesale Price Index includes a related category of “ bolts, nuts, screws, and
rivets.” Since January 1965, the Wholesale Price Index for these products has
risen 5.4 percent, more than 3 times the rate for all industrial commodities (1.6
percent).
IV. OCCUPATIONAL SHORTAGES AND SURPLUSES
A . Employment service data
The intensity of occupational shortages has increased measurably over the past
year and a half. Unfilled Em ploym ent Service job openings as of M arch 1, 1966,
totaled 344,000— the highest figure for that date since the end of W orld War II.
The M arch 1 total was 40 percent above that reported in July 1964.
Professional and skilled jobs still account for most of the hard-to-fill openings,
but unmet needs at lower skill levels have increased sharply since m id-1964.
Em ploym ent offices report that 7 of every 10 openings available for 30 days
or more have gone unfilled for lack of qualified applicants. About 15 percent
were not filled because of unfavorable working conditions; all other reasons ac
counted for another 15 percent.
Shortages are still of the “ spot” variety, rather than general in character.
Even with unem ployment down to 3.8 percent in March, there are 3 job appli
cants at local employm ent offices for every unfilled opening.
Among professional and technical categories, draftsmen, social welfare workers,
trained nurses, mechanical and electrical engineers, laboratory technicians and
assistants, and accountants and auditors were in shortest supply. The number of
unfilled openings has increased most rapidly for draftsmen, electrical engineers
and technicians.
In the blue-collar category, the most prominent shortages are for mechanics
and repairmen, machinists, welders and flame cutters, pattern-makers, tool makers
and die setters. A t the lower end of the skill spectrum, local shortages of con
struction laborers, warehousemen, autom otive washers and greasers and machine
shop laborers are reported.
Among service workers, the occupations in shortest supply include waiters and
waitresses, housemaids, nursemaids, restaurant cooks and kitchen workers,
private housekeepers and porters. Some workers have m oved out of these oc
cupations into others offering greater status and compensation.
B. Occupational unemployment rates
Unemploym ent rates for the broad occupational groups of experienced workers
show considerable declines over the past year. The seasonally adjusted rate for
professional and technical workers fell from 1.7 percent in the 1st quarter 1965
to 1.2 in the first quarter 1966. Similar declines occurred in jobless rates for
clerical workers— 3.6 percent to 2.8, and for sales workers— 3.5 percent to 2.7.
At the same time, unem ployment rates are still fairly high in the blue-collar
field. In M arch 1966 the rate was 3.7 percent for skilled craftsmen, 5.0 percent
for semi-skilled operatives, 8.9 percent for nonfarm laborers, 4.7 percent for
service workers and 6.6 percent for agricultural wage workers. While som e of
this unem ployment represents seasonal layoffs, and normal turnover in the job
market, nevertheless it also includes an untapped labor supply, a considerable
margin for econom ic expansion.
In addition, the labor force continues to grow. On the basis of population
trends, the projected increase for 1966 was 1.3 million persons. W e expect that
an additional 300,000 will be drawn into the jo b market by the attraction of
plentiful opportunity.
But the extra margin of labor supply, in the full sense, is not limited to the un
em ployed and the new entrants. M any persons are not working to capacity.
JOB VACANCY STATISTICS
48
There are still about 1,800,000 who want full-time work but are only obtaining
part-tim e; and half of these are adult men. A much greater number can assume
greater responsibilities if employers will upgrade them along with necessary train
ing and break-in arrangements.
V. CONCLUSION
Although this is the first published report in this series, we have been studying
manpower shortages intently since last December. It is significant that the
shortage areas, occupations and industries at the present time are largely the same
as those we identified several months ago. Meanwhile, payroll em ploym ent has
grown by over 900,000, and there is every indication that the expansion is con
tinuing. Thus, despite difficult “ sp ot” shortages, the problem s are being met
up to now.
[News from U.S. Department of Labor, Thursday, May 26,1966]
The N ation’s manpower supply is still ample for employers to maintain a rapid
rate of econom ic growth.
This fact is emphasized in the second monthly Report on M anpower Require
ments and Supply, made by Commissioner Arthur M. Ross of the Departm ent of
L abor’s Bureau of Labor Statistics at the request of President Johnson. The first
report w^as published in April.
The M ay review points out that even with recent increases in employm ent—
with nonagricultural payrolls up by 650,000 persons between M arch and April—
new additions to the labor force remain great enough to permit further business
expansion.
The report also notes that—
Seasonal pickup in certain industries— particularly in construction— has
increased the demand for experienced workers.
Washington, D .C ., Atlanta, Minneapolis, St. Paul, Cleveland, H ouston
and Milwaukee show the lowest unem ployment of all large metropolitan
areas.
W orkers are in the shortest supply in machinery manufacturing.
New defense contracts on the W est Coast have created unusually heavy
demands for engineers and metal workers in the aircraft industry.
M D T A training activities are playing an im portant role in providing
m anpower for hard-to-fill openings.
R e p o r t on
M a n p o w e r R e q u ir e m e n ts and
S u p p ly ,
M ay
1966
(B y Arthur M. Ross, Commissioner, Bureau"of L ab or Statistics, U.S. Department
of Labor)
[This is the second monthly report on manpower shortages and reserves prepared
in response to President Johnson’ s request in his M anpower Report to the
Congress. M uch of the inform ation was provided by the United States E m
ploym ent Service, Bureau of Em ploym ent Security, and the affiliated State
Em ploym ent Services. Other cooperating agencies were the Business and
Defense Services Administration of the Departm ent of Commerce, the Council
of econom ic Advisers, and the Bureau of Labor Statistics]
I. GENERAL ECONOMIC SUMMARY
As we reported last month, manpower reserves are still sizable enough to permit
further expansion of employm ent. The number of wage and salary earners on
nonagricultural payrolls rose 650,000 between M arch and April. The increase
since April 1965 has been almost 3,000,000. Increases of this magnitude indicate
that employers are still able to staff new positions and maintain a rapid rate of
econom ic growth.
Even with the recent increases in employment, the N ation’ s manpower reserves
remain substantial. M uch of the additional manpower consists of new additions
to the labor force, which will become greater at the end of the school year in June.
The number of unem ployed workers declined slightly in April, to 2,800,000.
Am ong the unem ployed were 1.1 million adult men, 840,000 adult women, and
860,000 teenagers. The national unem ployment rate (seasonally adjusted) was
3.7 percent, the lowest April rate since 1953, but virtually unchanged from
February and M arch. Rates of unem ployment for adult men and for married
men, however, edged downward in April (to 2.4 and 1.8 percent, respectively).
JOB VACANCY STATISTICS
49
Thus the reserve of male workers is still being squeezed down to meet manpower
requirements.
It is apparent, however, that manpower problems in particular industries, areas
and occupations have intensified as econom ic activity has picked up with the re
turn of spring. Payroll em ploym ent increased by 700,000 (seasonally adjusted)
between January and April, while the number of unem ployed fell by almost
200,000. Seasonal factors would increase the demand for experienced workers
in the months to come, even if production and em ploym ent were not increasing
beyond seasonal expectations.
M any employers state that available workers do not have sufficient qualifica
tions, that costs are increasing as a result of inefficiency and heavy overtim e assign
ments, and that high turnover is multiplying the expenses of recruitment and
training.
Labor turnover data provide circumstantial evidence corroborating the claim
that the manpower situation has tightened. After no change in February
turnover rates in manufacturing rose sharply in March. The quit rate in manufac
turing in M arch 1966 was 2.7 percent, up from 2.4 percent in February and 1%
times the rate in M arch 1965. The rate of accessions (or hirings) rose from 4.8
percent in February to 5.1 percent in March.
The highest M arch quit rates were in the furniture and fixtures, leather and
leather products, lumber and w ood products, textile mill products, and apparel
industries. These are the lowest paying of all manufacturing industries, and
historically have had difficulty in retaining workers in times of rapid econom ic
growth and abundant jo b opportunities.
It is im portant to avoid exaggerated impressions of the degree of stringency,
however. Other econom ic indicators show that the expansion is being taken in
stride so far as manpower supply is concerned. In periods of severe stringency,
average weekly hours and overtime hours increase sharply as employers stretch
out the utilization of their existing workforce. Although hours of work rose in
1965 and early 1966, this trend has halted in the past tw o months. The factory
workweek in April remained the same as in March— 41.5 hours— and was 0.1
lower than in February. Average overtime hours were unchanged for the third
straight month.
In durable goods manufacturing, hours rose by .1 hour to 42.4 in April, but
were the same as in two of the previous three months. Average overtime hours
also rose in April (by .2 hours to 4.6) but were no higher than in February.
Among individual manufacturing industries, average weekly hours were highest
in machinery manufacturing and paper and allied products (43.7 hours), and
transportation equipment (43.5). The only significant increases over the month
were in ordnance (.6 hour), and transportation equipment (.5 hour). In most
other industries, hours either rose slightly or actually declined last month. (The
foregoing data on labor turnover, hours and overtime are seasonally adjusted.)
If inefficient, marginal workers were as numerous as some are inclined to believe,
the productivity statistics would be affected. Y et output per man-hour in the
total private econom y was 3.0 percent higher in the first quarter of this year than
in the first quarter of 1965. This is about equal to the trend rate of productivity
increase during the entire postwar period. The productivity advance since the
last quarter of 1965 has been 0.6 percent, equivalent to an annual rate of 2.4
percent.
Again, if manpower shortages were holding up production on a significant scale,
the ratio of unfilled orders to m onthly shipments in manufacturing would be
stretching out. In a few industries such as metalworking machinery, order back
logs have been accumulating. But for durable goods industries as a whole, the
average backlog of 3.2 months in M arch 1966 was the same as a year previously.
For all manufacturing industries the ratio of orders to shipments was 2.7 months
in March, the same as in February and up 0.3 month from M arch 1965.
II. THE MANPOWER SITUATION IN LABOR AREAS
One-third of the 150 m ajor labor areas for which m onthly unem ployment meas
ures are available are now classified as having relatively low unem ployment (un
employm ent rates of about 3 percent or less).
Am ong the largest metropolitan areas (1,000,000 or more population), the
areas of low unem ployment are Washington, D .C .; Atlanta, Minneapolis-St.
Paul, Cleveland, Houston and Milwaukee. Other areas with relatively tight job
markets are Hartford, Connecticut; Wilmington, Delaware; Cedar Rapids and
Des M oines, Iow a; Lansing and Saginaw, M ichigan; Rochester, New Y ork ;
50
JOB VACANCY STATISTICS
Lancaster, Pennsylvania; Ft. Wayne, and Indianapolis, Indiana; D ayton and
Columbus, Ohio; Dallas, Texas; Richmond, Virginia; and M adison, Wisconsin.
Although regular reports from the m ajor employm ent centers show a continued
downward trend in unem ployment rates, there are as yet no m ajor areas which
warrant classification in the overall labor shortage category, defined by an un
em ploym ent rate of 1.5 percent or less.
The number of m ajor areas having substantial unemployment rates (6 percent
or more) has fallen from 29 in February 1965 to 18 in February 1966. Relatively
large areas with substantial unemployment are Fresno, San Bernadino, San Diego
and Stockton, California; Fall River, Lawrence-Haverhill, Lowell, and New Bed
ford, Massachusetts; Duluth, M innesota; Atlantic City, New Jersey; Altoona,
Scranton, and Wilkes-Barre-Hazleton, Pennsylvania; Charleston, HuntingtonAshland, and Wheeling, West Virginia. The cities of Miami, Newark, Oakland,
Philadelphia, Pittsburgh, and San Diego are designated as *‘persistent unem
ploym ent areas” for purposes of Federal procurement preference.
III.
THE MANPOWER SITUATION IN SELECTED INDUSTRIES
The most serious manpower problems of any industry continue to be found in
machinery manufacturing. The machinery industry will need 150,000-200,000
additional workers in 1966, and im portant occupations are already in short supply.
M anpower problems in the machine tool segment of the machinery industry
intensified during April, although other segments appear to be holding their own.
Seasonal fluctuations have the most pronounced effect on the construction
industry. Em ploym ent increased between M arch and April, although strikes
and poor weather held the increase below seasonal expectations. Skilled w ork
ers— particularly electrical workers, plumbers, pipefitters, iron workers, carpen
ters, bricklayers and sheet metal workers— are in short supply in many sections
of the country. California, New Y ork and Pennsylvania still have high rates
of unem ploym ent in construction, however.
The industry most affected b y placement of new defense contracts is aircraft.
Several im portant new contracts on the West Coast, both defense and nondefense,
are creating unusually heavy demands for engineers and metal workers.
IV . OCCUPATIONAL SHORTAGES
The occupations in shortest supply are still those reported last m onth— engineers,
draftsmen, nurses, other health service workers, machinists, tool and die makers,
machine tool operators, and pattern and model makers. There are nationwide
shortages in these occupations regardless of the situation in particular labor areas.
Even unskilled industrial workers are difficult to hire in some of the tighter areas.
An indication of the nature and trend of shortages in specific occupations is
provided by interarea recruitment data of the Federal-State em ploym ent service
network. Over the M arch 1965-M arch 1966 period, there was a substantial
increase in interarea requisitions for every major occupational group. Needs for
professional workers increased by almost a third, those for production workers
nearly tripled, those for clerical and sales workers nearly doubled and those for
service workers jum ped over 75 percent.
Am ong professional workers, the largest numbers of jobs placed in interarea
recruitment were in engineering and health specialities. The engineering special
ties in greatest demand are mechanical, electrical and industrial engineering.
Needs for draftsmen— mostly in aircraft, shipbuilding, ordnance, contract con
struction and engineering and architectural services— have increased sharply
during the past few months.
Am ong production workers, demand continued to increase in skilled occupations
such as metalworkers, particularly machinists, tool and die makers, and machine
tool operators. Similarly, interarea recruitment for pattern and model makers
multiplied significantly in the first quarter of 1966, as recent large contracts in
the aircraft industry pushed hirings for these workers. Dem and for semi-skilled
metalworkers also increased in many areas. The recent step-up in aircraft has
prom oted additional interarea recruitment for semi-skilled workers needed b y this
industry.
V. PROGRAMS TO COPE WITH SHORTAGES
R eports on hiring practices show a further readiness among employers in most
sections of the country to reduce hiring specifications, eliminate arbitrary barriers
to em ploym ent, and accept trainees who have completed Federally-financed
training programs.
JOB VACANCY STATISTICS
51
The U.S. Em ploym ent Service is also taking new actions aimed at resolving
problems of skill shortages. (1) State agencies have been advised to inform em
ployers of developing skill shortages and to provide guidance about adjustm ents in
hiring specifications, job simplifications, and the establishment of training pro
grams. (2) M D T A training is being redirected to accom m odate itself both to the
need for equipping the disadvantage for available job s in the community, for
which 65 percent of the training funds will be allocated; and, for meeting skill
shortages in local areas, which will absorb the remaining 35 percent. (3) Close
cooperation with other agencies through a series of task forces and special work
groups, is mobilizing resources in the Departments of Commerce, Defense and
Health, Education, and Welfare for im provement of inform ation and policies.
M D T A training activities are also playing an increasingly important role in
providing additional manpower for hard-to-fill openings. Approximately 345,000
persons have been enrolled in M D T A institutional training programs since the
inception of the program in 1962. Almost one-third of these persons have
enrolled since mid-1965. Three-fifths of all institutional M D T A training under
taken through 1965 has been geared to training workers in shortage occupations.
O n-the-job training programs under M D T A are helping to meet the heavy demand
for trained workers. In the last few days of April, for example, on-the-job
training programs were approved for more than 6,000 workers. These included
one contract with the Social Developm ent Corporation of Washington, D .C . to
establish hospital training programs for 1,775 workers in eight subprofessional
occupations— nurses’ aides, ward clerks, orderlies, dietary aides, psychiatric
aides, computer operators and programers, and medical secretaries. Sixty-five
other contracts were approved for the training of workers in occupations ranging
from apprentice draftsmen and surgical technicians to machinists and apprentice
tool and die makers.
Government-sponsored programs can only meet a small part of the N ation’s
training needs. It is essential that employers increase their own upgrading and
training activities in order to correct manpower imbalances as they develop.
Chairman P r o x m i r e . I am wondering about the validity of these
reports, in view of the strong case you make here for getting this
additional data.
Mr. Ross. Well, I don't want to fall between two stools in replying
to you, sir. I will say that we are able to identify the areas which are
the tightest. We are able to identify some of the professional and
skilled labor and service occupations where the jobs are hardest to
fill. We are able to identify certain industries, such as a machinery
and construction, which are having problems in that locality-----Chairman P r o x m i r e . Y o u can't get much of it quantitatively?
Mr. Ross. We can't make it quantitative.
We do have the data on unfilled Employment Service openings.
We know that while some employers use them, some do not.
Chairman P r o x m i r e . That is about all they have in foreign
countries, isn't it?
Mr. Ross. Yes.
Chairman P r o x m i r e . S o we go as far as they go now?
Mr. Ross. Except that some of the foreign countries probably
have a much better—in some of the foreign countries the employers
use the Employment Service more than in this country, so that their
employment service records would be more complete than ours would.
Chairman P r o x m i r e . Have y o u discussed this at all with labor
leaders?
Mr. Ross. Yes.
Chairman P r o x m i r e . The reason I raise this point is that we are
having Mr. Goldfinger tomorrow, and it would be very helpful to me
if I could anticipate his viewpoint. Could you give us what you
think might be said, to the extent that they may have reservations or
objections?
52
JOB VACANCY STATISTICS
Mr. Ross. Yes. I would not want to speak for him, of course.
We have been very aware of the misgivings on the part of some labor
and management people as well. And we think that we have been
able to answer them. For example, labor has expressed the concern
that some gross estimate of vacancies wmild be deducted from the
gross total of unemployment, and then the subtrahend would be
represented as the true or net figure of unemployment. As I said
in my statement, I do not think that, any responsible labor economist
would do such a thing. And that is not our intent.
Chairman P r o x m i r e . I am sure that experts such as you would
not do it. But it is a perfectly natural thing for many, many people
to do, if these statistics are available. And I am wondering what
we can do to minimize that. In my judgment that certainly should
not, prevent this information from being available, but how can we
prevent the misuse?
Mr. Ross. I think what we do to minimize it is to associate the
vacancy data with occupational detail, industry detail, location,
area detail—in other words, a vacancy has an occupation, it has an
industry, it has an area, it has a wage rate, just as a person has arms
and legs and eyes and a nose. And I think we should discourage as
much as possible promiscuous use of undefined or unqualified vacancy
totals.
Now, as you say, there is always a possibility that people will use
them that way. And I think the advantages of the data are great
enough that we should go ahead in any case and do all we can to
prevent misuse.
Chairman P r o x m i r e . Maybe it will take us time to develop this,
but I certainly appreciate hesitation in this area.
Mr. Ross. Yes, I do too.
Chairman P r o x m i r e . Because if, for example, we had 3.7 percent
unemployment, and you found that there were 2 percent vacancies,
and the charge was made that the true unemployment is less than 2
percent for this reason, it obviously would be wrong. But also it
would be very tempting for those who want to minimize unemploy
ment decisions.
Mr. Ross. Yes. I think it would not be justified to say that under
those circumstances the unemployment rate is really 1.7 percent.
I would see no validity in such a statement. Of course, someone
might make it, just as people misinterpret price statistics and wage
statistics. We would do all we could to discourage misinterpretations.
But it is a free country.
Chairman P r o x m i r e . Congressman Curtis has very courteously
permitted me to go ahead. And I appreciate it. I have just a couple
of more questions.
Would this information be useful, for example, for those of us who
are somewhat critical of the NASA program, those of us who feel
that it is draining too many engineers out of the civilian economy,
and call for manpower studies, to at least evaluate what the impact of
that is? Would we be in a better position with this kind of a study
to evaluate what effect governmental programs of various kinds have
on the availability of people with certain skills?
Mr. Ross. Well, I think it would to some extent, Mr. Chairman.
We do know, at least we have estimates, of the involvement of engi
neers, both in-house and in private industry in the NASA program.
JOB VACANCY STATISTICS
53
If we had information on the need for engineers elsewhere it would
be possible to make some judgments there. Of course, there are
many kinds of engineers, as you know. And we would have to
distinguish between civil engineers, aeronautical engineers------Chairman P r o x m i r e . Would you do that in your understanding of
how this program would work?
Mr. Ross. Well, we have a section in my statement on the problem
of occupational detail. This is something that needs more study.
And we have not yet fully solved the question of how much specificity
we could have as to the different types of engineers, different types of
physicians, different types of teachers, and so on.
Chairman P r o x m i r e . H o w about the alternatives? To what extent
might existing programs be strengthened to achieve some of the same
independence as the vacancy data? I have in mind the Employment
Service listings, the NICB index of help wanted advertising, and
various professional mediums.
Mr. Ross. Well, I think the openings on file at the Employment
Service do provide a helpful source of information. Manpower
Administrator Ruttenberg and I are using them in our analysis of
shortages and in the reports we are making. I might say that hence
forth Mr. Ruttenberg and I will make the shortage reports jointly,
although I have been doing them previously.
More can be done to get that information up to Washington speedily.
But the fact still remains that from one locality to another there are
differences in the degree to which employers in different industries
use the Service and file their orders. And there are some industries,
such as construction, that do not operate through the Service. They
generally get their employees in the union hiring halls. So typically
at least the union employers would not file any orders with the
Employment Service. And there are some of the professions and the
white collar occupations that generally go to the private employment
agencies organization, or—-then, of course, many jobs, as you know,
are filled informally by word of mouth, by personal contacts, and so
on. And they would not turn up in any formal reports from either
private or public employment offices.
The index of help wanted advertising by the National Industrial
Conference Board is an indicator—I do not think it is a precise statis
tical indicator, nor that it ever could be. And I do not think that it
would ever give us a good breakdown of demand as between occupa
tions. For one thing, the really hard jobs to fill are advertised in
many communities, or in many papers. You will find, let's say,
Lockheed needing aeronautical engineers.
Chairman P r o x m i r e . There are a lot of duplications?
Mr. Ross. Yes; there would be great duplication.
Chairman P r o x m i r e . It is hard to eliminate it, too.
Mr. Ross. And again you might have a thousand vacancies for
engineers and a thousand vacancies for restaurant cooks, and you
would get much bigger display ads for the former than for the latter,
or more ads all together. But there are still a thousand vacancies for
each. I do not disparage the NICB help wanted advertising index at
all, and I am very much impressed by their experimental job vacancy
surveys, but you should not think the help wanted ads will give us
the information we need.
54
JOB VACANCY STATISTICS
Chairman P r o x m i r e . Just one final question. Would the devel
opment of job vacancy data on the scale contemplated represent the
optimum allocation of our statistical resources?
Mr. Ross. I think that we need the data sufficiently at this time
as to justify the allocation of the money we are requesting for this
purpose.
Chairman P r o x m i r e . Would this be the optimum, would this be
the best place we could put our money?
Mr. Ross. Well, we are, of course, seeking additional money for
a number of purposes. I believe that this justifiably has a high
priority among all the purposes for which data are needed.
Chairman P r o x m i r e . Thank you very much.
Congressman Curtis?
Representative C u r t i s . Thank you, Mr. Chairman.
And thank you, Mr. Ross, for what I regard as a very excellent
presentation. It is most timely, because we are way behind where we
should be in this area.
I am reminded that I slipped over here from the Ways and Means
Committee, where we are putting the finishing touches on the unem
ployment insurance bill. After months of hearings, we realize how
weak our data are. Our knowledge of this problem of employment
and unemployment is inadequate. If we had job vacancy information
available to give us some better idea of the profile of the unemployed,
we would be much further ahead. One of the basic points is whether
we should simply extend the benefit period under unemployment
insurance, or attack long-term unemployment by the manpower
training approach. I was terribly disappointed when the witnesses
of the administration, in their prepared statements, failed to relate
the two. But, of course, the training of the long-term unemployed is
dependent upon having some statistics on the jobs available.
I want to point up other areas where these statistics would be
valuable, not from the standpoint of discounting what you have
said, but to help myself get a survey. I have remarked that our
new immigration laws were really turning over to the Department
of Labor the problems of immigration, related as they are to the needs
for the skills of particular immigrants in our society.
How does the Department of Labor now operate these new pro
visions of the immigration laws without having more sound statistics
on job vacancies? Is that in your area?
Mr. Ross. That is more in Mr. Chavrid's area in the U.S. Employ
ment Service.
Would you care to have him answer at this time?
Representative C u r t i s . I will wait to come to him.
But from your standpoint, would you agree that if we were going
to do a job here these kinds of data would be invaluable to us?
Mr. Ross. Yes. Well, for many purposes, including immigration,
employment services policy, interarea recruitment, and so on, we do
need to have lists of short occupations. And of course if vacancy
data are available, it will give us a more accurate list. It is always
important when you come to labor shortages to base it on evidence
rather than on the loudness of complaints, because the employer
who complains the loudest is not necessarily the one who is really
the shortest of labor.
JOB VACANCY STATISTICS
55
Representative C u r t i s . Incidentally, in your exchange with Senator
Proxmire on the possible misuse of statistics, I was thinking that it
also might help to prevent the present misuse that is going on, at least
in my judgment, where the cause of our unemployment problem has
been identified as a lack of aggregate demand, rather than structural
and frictional imbalances. I would be willing to risk the misinterpre
tation of the statistics if I could help to dispel some of what I think is
a very shortsighted approach to the nature of our unemployment.
I have observed that you can get roast pork by burning the barn
down, but it is a poor way of getting it. I think this country through
out its history has been short of labor, and we have filled a great deal
of our labor needs through immigration. But I think we are still in
this period of shortage of labor.
The second area is rehabilitation. There has been some develop
ment in this area, and I am wondering whether or not these data
might not be available for this purpose. It is not automatically rec
ognized. But the great need there would be to identify what kind of
jobs there are for people who might have a handicap. Is there enough
of that kind of data available to you and do you avail yourself of it
now?
Mr. Ross. Again, I think that is something which Mr. Chavrid of
the Employment Service could better comment on, because it is an
operating question so far as the rehabilitation is concerned.
Representative C u r t i s . A third area in this business of meeting
the demand for skills isn't to take the unemployed, who tend to be
untrained and semiskilled and train them for the difficult skills that
are in demand. It is the much more difficult problem of taking a
person with a job and upgrading his skill, and then leaving his good
but less-skilled job to be filled by someone who requires training.
With this in mind, I was interested in a study where an estimate was
made of the amount of on-the-job training done by private industry.
And the figure came out, as I recall it, to $14 to $16 billion a year.
But again, in this area, there should be some way of identifying what
is going on. An employer has told me that he does not even advertise
for these people, because he knows they are in such short supply. Or
it is a skill that just has come in, and there is no training going on, so
many employers do not advertise at all. They simply set up a method
of on-the-job training. Now, I do not know whether there is any way
that we can get these data, or whether they are in such shape that
they can be used. But if they were, I think they would be very
important in our understanding of this problem.
Mr. Ross. Mr. Curtis, I have not seen the $16 billion estimate.
I have discussed this with industry people. And I am not aware of
any comprehensive or reliable study of the amount of in-plant train
ing. And I agree with you that the upgrading in this plant training
has tended to deteriorate somewhat during the past decade, because
there was a lot of unemployment throughout, and employers were
able to find what they needed more often than not on the open market.
Now, the point Mr. Cassell made—and it may have been before
you arrived, sir—was that as an employer he found that he was often
in doubt as to whether to wait and see if the person would show up in
response to an ad or a job order, or whether he would have to take
one of his existing people, give him more training, and put him on a
more responsible job. I think the vacancy job would help employers
to make that decision more rapidly .
56
JOB VACANCY STATISTICS
Representative C u r t i s . I fully agree with you.
Incidentally, the quit ratio seems to always go up during periods of
economic upturn.
Mr. Ross. Yes, indeed.
Representative C u r t i s . That is an accurate statement of mine?
Mr. Ross. Yes; it is.
Representative C u r t i s . This indicates a healthy thing. People
who possibly m a period of relatively high unemployment will stick
with their jobs rather than look around for something better, or up
grading skill, or whatever. Is this a ratio that is of any value in indi
cating job vacancies?
Mr. Ross. Well, it is affected by quite a few things. Naturally,
when a person quits, ordinarily the employer wants to replace him.
So that, itself, would indicate a vacancy. We find, however, that the
quit rate is highest in the low-paying industry.
Representative C u r t i s . Well, that would indicate an upgrading o f
skills, then, would it not?
Mr. Ross. Yes. For example, right now in manufacturing, which
is the only place we have quit rates, the quit rates are highest in lum
ber, apparel, textiles—-well, other of the higher paid manufacturing
industries, they are not so high, let's say, in machinery, where we have
some of the more noted occupational skills.
So it is a good general economic indicator, but it does not tell you
where your difficult shortages are.
Representative C u r t i s . N o w , just to mention again the problem
of military manpower utilization. I am speaking not only of the
utilization of people they put in uniform, but those that they might
have in the civil service, under the military, plus the munitions in
dustry, which is a fluctuating thing.
Now, we have or had— and I think it is still in existence—a Man
power Utilization Board, that is the exact title it had during World
War II. And I think it continued in existence. Its job was to try to
anticipate the military needs in relation to the tight skills from the
military standpoint, at any rate, in the civilian sector. And in our
draft laws right now there is a provision where people in certain occu
pations are deferred.
It would be interesting to know what data they have, how they col
lect it, and where these shortages are. Have you reached into that
area at all?
Mr. Ross. Mr. Curtis, I think it might be best if we would file a
statement for the record on that question of critical occupations in
relation to draft policy.
Representative C u r t i s . And anything further that you might
want to supply along the lines of the questions that I asked before on
this?
Mr. Ross. Yes, sir.
(Following statement was later supplied for the record:)
In determining if there is a shortage of workers in a particular occupation—
one of the criteria used in placing an occupation on the List of Currently Critical
Occupations— inform ation is developed primarily from indirect evidence, and
follow ing a detailed examination of available data. The number o f jo b openings
on file w ith the State Em ploym ent Services provides some evidence. Analysis
of current and past trends in empkwm ent and unemployment in the particular
occupational group provides some information. Discussions w ith unions, trade
associations, employers, educators and training authorities are also helpful.
JOB VACANCY STATISTICS
57
Other information examined relates to wage levels, hours of w^ork, recruiting
efforts, and training activities. In no occupation, however, are there direct
incontrovertible facts such as might be provided by long-term information on the
level and trend of job vacancies. Without such direct evidence, determination
of an occupational shortage must rest on circumstantial evidence.
Representative C u r t i s . Incidentally, here is an in-house operation
of the Federal Government where they control both the demand and
are in the business of trying to meet this demand. And I suspect that
they have got over a period of years on an empirical basis a lot of
data. And yet it is not available. I can't even get the study that
the administration was supposed to have been making—which was
reported as being available, I think, in April of 1965—into this whole
problem of the draft as a method of procuring manpower for the Mili
tary Establishments and its impact on our labor market. I think I
know what is in that report, or I have an idea of some of the
conclusions.
But to have a report of that importance suppressed—and that is
apparently what it has been—to me is a rather shocking thing.
The final area that I want to touch on—and you have already
answered to some degree—relates to the shift that is continuing to
go on in our society from rural to urban occupations. I felt that
a great deal of the problem of the Negro has been erroneously iden
tified as a Negro problem when really it is the age-old problem that
occurs whenever there is a shift of the rural dweller into the urban
areas. And there is great need, I would argue, for updating our Fed
eral vocational educational program, which was heavily oriented
toward retraining or training in the agricultural skills, toward the
kind of skills that are in the urban areas. But I think we have got
to reach into the agricultural area, and not separate it out. There
needs to be this coordination between the Department of Agriculture
and the Department of Labor.
I said that was my final observation. I do have one other before
I ask a very specific question.
It seems to me that as a society, as our economy advances, it moves
heavily from the manufacturing or product area into distribution and
services of all sorts. And certainly in the job area the demand for
skills seems to be increasingly in the service and distributive area, if
we include in service things like medicine, education, and government,
local and State.
This, I think, is an important factor, because so much of our
statistics seems to be oriented toward those which are easiest to
collect, which is in manufacturing. Would you say that is a fair
observation?
Mr. Ross. Yes, I think that goes for employment statistics, wage
statistics, fringe benefit statistics, turnover statistics, almost every
kind. And it is part of my program at the Bureau to extend the pro
gram throughout the economy into the service industry, and those
that are harder to collect.
Representative C u r t i s . The specific question I had in mind was
this. When I asked Secretary^ Wirtz earlier this year—and I think
you were on the panel at the time—why we had not moved forward
more rapidly in this area of getting our job vacancy statistics, he
responded by saying it was because Congress had not given the
administration the money. And I asked him how much he was
58
JOB VACANCY STATISTICS
talking about, and he said $2.5 million. And I was frankly a little bit
shocked to think that this program had been held up for any reason
of that nature. What are you asking for now in the way of additional
funds?
Mr. Ross. The program which we are requesting would be $2.5
million.
Representative C u r t i s . It is that same figure?
Mr. Ross. Yes.
May I comment?
Representative C u r t i s . Certainly.
Mr. Ross. I think it is on a little different point, so perhaps I had
better do it later.
Representative C u r t i s . I certainly hope that the administration
presses this, and alerts the leaders of the Congress to the need for it.
I certainly will do everything I can on my side of the aisle, as I have
stood ready to all along. And to me it is a shame, really, that we
have held back developing this kind of data because of sums of money
of this size. We would have this money back a thousand times over,
in my judgment, in getting better programs and unemployment in
surance, manpower training, immigration—-and you have named a
great many others.
Thank you.
Mr. Ross. Mr. Chairman, one question that you raised I don't
think I answered very well. Could I add a little bit to it?
Chairman P r o x m i r e . By all means.
Mr. Ross. You asked what were the misgivings which labor had
expressed about the program, and whether I thought we had
been able to meet that. One of them I did mention was the possi
bility of misuse of gross vacancy data against gross unemployment
data. I have commented on this. And we, of course, w^ould not do
it.
A second misgiving which has been expressed is whether employers
understand the concept of vacancy, and whether they might not
exaggerate their vacancies. We find upon our response checks that
most of them do understand very well, and rather than an exaggera
tion, there is an understatement.
Thirdly, a point raised by some of the labor people goes to the
definition of a vacancy. Our current definition would count as a
vacancy a job which has become open for any period of time. Now,
the labor people argue that since one is ordinarily not counted as unempkn^ed unless he was unemployed throughout the survey week,
therefore a vacancy should be counted for our statistics when it has
been open for a week.
We have counted them both ways. And we find that it does not
make an awful lot of difference, in that most of the vacancies reported
have been open for a week. But I think that is an open question,
and one which we will want to do more thinldng about. We have
not arrived at a dogmatic position concerning the question of how
long a vacancy should be open before being counted.
Next, the labor people have expressed concern as to whether sub
standard wages or other unfavorable working conditions—■
—•
Chairman P r o x m i r e . This agricultural question, where you are
asking people if they will work for 50 cents an hour?
JOB VACANCY STATISTICS
59
Mr. Ross. Yes. Of course, the program does not contemplate
coverage of agriculture yet, but the labor people have been appre
hensive lest a vacancy which could not be filled because the wages
were too low be given legitimacy and authenticity by being counted
with other legitimate vacancies. Our feeling about that is that it is
very important to know the wages, and if possible other important
conditions of work, and to keep them always in mind in analyzing
these vacancies. Vacancies that cannot be filled because the wages
are too low should be identified as such, for operating purposes.
Chairman P r o x m i r e . Y ou would count it as a vacancy?
Mr. Ross. Our program does count reported vacancies as such.
Now, we did make a study, a first pilot study of the wages—you see,
in a couple of areas we also collected wage rates, although not in all
the areas— excuse me, may I correct that—in five areas the wage
information was also collected. And my statement does show-----Chairman P r o x m i r e . Would this have a very big effect on the
statistics? Supposing you excluded all those vacancies where the
wage was obviously too low?
Mr. Ross. No; it shows that the great majority of the vacancies
do offer at least the going entry rate for that occupation in that
community. But it is a problem. And I feel that one should not
lose sight of the wages that are being offered.
Finally, the labor people have felt, and I agree, that it would be a
mistake to engage in a large scale publication, a published interpreta
tion program too quickly before we know what we are dealing with.
This is true with any new statistical program. And it is our feeling,
as I say in my statement, that we would want to look at the informa
tion carefully and to study it, at least to discuss it with our industry
advisory committee, our labor advisory committee, to discuss it with
academic experts, and so on, before we decide what types of publica
tion would be appropriate.
Thank you.
Chairman P r o x m i r e . Thank you very much.
Our last witness is Mr. Chavrid, Director of the Office of Manpower
Analysis and Utilization.
You have an excellent statement here, Mr. Chavrid; very detailed.
I am sure it will be helpful to us. I had a chance to go through it.
You may present it any way you wish. Do you wish to put the whole
statement in the record and summarize it?
STATEMENT OF VLADIMIR D. CHAVRID, DIRECTOR, OFFICE OF
MANPOWER ANALYSIS AND UTILIZATION, U.S. EMPLOYMENT
SERVICE
Mr. C h a v r i d . Yes, Mr. Chairman.
Mr. Cassell described some of the operating uses of the data.
Mr. Ross reviewed some of the analytical uses of the data. I would
like to briefly describe— and this is summarized in my statement—■
the experiment that was conducted in fiscal year 1965 in some 16
labor market areas.
The 16 areas that were selected for the experiment were based on
criteria relating to the size of the area itself, the industrial composition
of the area, its geographic location, as well as the extent of unem
ployment in these areas so that we could get as good a cross section
60
JOB VACANCY STATISTICS
as possible, in terms of geography, industry, and the economic well
being of the area.
The studies were done by the U.S. Employment Service and the
affiliated State employment services in cooperation with the Bureau
of Labor Statistics. All of the work done in the field was done through
the State agencies, except the response analysis which was done by
the Federal staff of the Bureau of Labor Statistics.
The emphasis in the experimental program was operational in
nature, in an attempt to improve the functioning of the job market.
A variety of information was collected in the experimental program.
We obtained information on the total number of job vacancies at
a given date, those that were open for 7 days or more, and those that
were open 30 days or more.
We have also obtained information on part-time job vacancies and
temporary job vacancies in two areas. We have obtained data in
five areas, as was mentioned earlier, on wage rates, to see what the
extent of substandard wages would be.
In some areas, we also have divided the sample establishments
selected into two groups with half the area receiving a standard
collection schedule and the other half a variant schedule. The addi
tional data collected on the variant schedule included part-time and
temporary vacancies and wage data. This was done in an attempt
to see if the attempt to collect some additional information would
result in a lower or higher response on the part of the employers to
our questionnaires.
As it turned out this did not have an important bearing on the
willingness of the employer to cooperate in providing the information.
There were, however, differences in the response by different areas.
In the first survey that was conducted in the last quarter of 1964,
some 78 percent of the employers who were asked for the information
responded. In the April 1965 survey there was an 81 percent response.
In some areas, the response was unusually large. In fact, in Mil
waukee the response was close to 95 percent. And so it was in Port
land, Oreg.
I sometimes think that the kind of response we had was perhaps
based more on the specific timing of the survey and the ability of the
local office to get enough staff to do the job at that moment, rather
than the reluctance of the employers in one area or another to provide
the information.
In both rounds of the survey 99 percent of the schedules returned
by the employers were usable.
Now, briefly, on the nature of the job vacancies, there were, all told,
some 20,000 employers selected in the sample in the 16 areas. This w^as
on a sampling design that provided for the inclusion in the sample of
large establishments, medium-sized establishments, and small estab
lishments.
A question was raised of what was meant by a large establishment or
a medium establishment. These varied from area to area, depending
on the size of the area. In the case of New York a large establishment
was defined as one having more than 2,000 workers. The mediumsized establishment was defined as that having between 100 and
2,000 workers. A small establishment was that having between 4 and
99 workers. No establishment below four employees was in the
sample.
JOB VACANCY STATISTICS
61
In the case of Richmond, which was a smaller area, a large estab
lishment had 200 or more workers, a medium sized from 200 down to
88 and a small one would employ between 4 and 87.
The employment in the so-called medium and large establishment
which were to be covered in the sample was to represent 50 percent
of the total nonagricultural employment in a given area.
Employers were asked on the schedule to list their vacancies by
plant job title, which were then converted by the local office employ
ment service staff into the codes and nomenclature defined in our
Dictionary of Occupational Titles. The people who coded the sched
ules were experienced in this area of work. If there were to be emerg
ing occupations, as Mr. Curtis has brought out, that currently are not
in the Dictionary of Occupational Titles, this would appear on the job
vacancy questionnaire that the employer would fill out. If the occu
pation or the job title listed there was not known to the local employ
ment service, they would send out an experienced occupational ana
lyst, who would then try to assign some existing code or develop a
new one for that particular occupation.
So through vacancy surveys it is possible to learn about new and
emerging occupations as they develop, which might not happen as
stated before, if only help wanted advertising were used to identify job
vacancies.
Representative C u r t i s . Could I interrupt?
Your sample covered all kinds of occupations, then. It was not
just manufacturing?
Mr. C h a v r i d . The sample covered all kinds of industries, including
government.
Representative C u r t i s . Very good.
Mr. C h a v r i d . Including nonprofit institutions, the total in a com
munity. The only thing that it did not cover was an establishment
which had less than four workers. We felt that we were reaching a
diminishing return by going that low.
Now, the Employment Service job openings are reported for all
skill levels.
I might mention that, what we call here the so-called broad skill
groups are occupational groups like professional, managerial, clerical,
and sales, service, and then it goes down to the skilled, semiskilled,
and unskilled. There were nine such groups. These 9 one-digit
groups include 580 so-called three-digit occupations. And within
these there are all told 9,000 different occupational codes. All of these
are coded in the Dictionary of Occupational Titles.
So it is necessary for experienced staff to identify all of these occupa
tions in the individual areas and for the Nation as a whole if the data
are to serve some useful purpose.
In both rounds of surveys, vacancies were found at all skill levels.
Clerical and sales occupations accounted for nearly 30 percent of the
vacancies in both rounds, a larger number than for any occupational
group. The second largest representation was among professional
and managerial workers.
While the vacancies were reported at every skill level, the magnitude
and distribution varied widely from area to area.
Demand for certain occupations was general, while for others it was
concentrated in specific areas. Almost all areas reported a demand for
nurses, for example. In New York City there were almost 1,000
62
JOB VACANCY STATISTICS
vacancies for industrial engineers, while in 13 other areas there was a
combined total of only 300 vacancies for this particular occupation.
So tiaere were very wide differences among the areas.
More vacancies existed in manufacturing industries than in any
other segment. But this was to be expected because, in most of the
15 areas, factory jobs were the largest segment of the employment.
About 30 percent of nonfarm employment in those areas was in
manufacturing, and about the same proportion of vacancies was also
found in manufacturing.
A separate analysis was made in our paper of the so-called hard-tofill jobs. We have more or less arbitrarily defined hard-to-fill jobs as
those vacancies which exist for 30 days or more. And about 50 per
cent of all of the openings fell into this category, that is, they had been
open for 30 days or more.
By occupation, as we expected, the proportion of the so-called
hard-to-fill vacancies ranged from slightly over 70 percent in the
professional and managerial and skilled groups to about 45 percent in
clerical and sales groups.
While the overall proportion of long-duration vacancies was high,
certain detailed occupations were considerably higher. Among the
professional and skilled groups, for example, 9 out of 10 vacancies for
trained nurses and for tool and die makers were open 30 days or longer,
while 80 percent of the vacancies for civil and chemical engineers and
tailors were hard to fill.
Now, a major purpose of the job-vacancy experiment was to deter
mine the extent of the Employment Service participation in the local
area job market, that is, the percent of total vacancies in the area
which was accounted for by Employment Service unfilled openings at a
single point in time. The comparison revealed that of the total job
vacancies in the community, the Employment Service held about
one-third. This is a very impressive figure, and much higher than
we in the Employment Service or people on the outside thought. It
is nearly comparable with some of the more advanced countries or—
let's put it differently—some of the countries in which it is believed
that the Employment Service is much more advanced than in the
United States.
There were differences, of course, between areas. This would
depend upon the effectiveness of the Employment Service in the area,
as well as the extent of labor shortages. The more shortages there
are in the community, the more likely the Employment Service would
get the larger share of such openings.
There was, on the whole, a general correspondence between the
unfilled job openings that were listed with the public Employment
Service and the job vacancies, except for certain occupations such
as bricklayers and carpenters and other construction occupations,
in which placements are made primarily through the unions.
I might at this point touch on the immigration question. There
is no question that the additional information on the job vacancies
would provide a more conclusive tool for determining the total occupa
tional shortages at any one time. But the decision to admit immi
grants or not is based on needs in the individual communities.
We have two lists. One refers to the professional occupations,
which by law we must admit to the United States. The second list
JOB VACANCY STATISTICS
63
is of those occupations for which very little training is needed. So
the Secretary of Labor compiled a list of those occupations, that is,
those that could be met by available American workers.
But in between there is a large grey area of nonprofessional occupa
tions which require considerable training, which may be short in
Milwaukee, but are surplus in Scranton, Fa., or elsewhere. So the
determination for this is based on the extent to which there is a need
for them in individual communities.
This is one of the operating uses in the job vacancy data.
Some of the others were in the development of training courses under
the Manpower Development and Training Act.
In Baltimore and Kansas City, for example, the job-vacancy
information was used as a basis for establishing vocational training
courses in the school. The Oregon agency reported that the data
from the first round of job-vacancy surveys was used by the State
department of education to develop vocational education and other
post high school courses for Portland Community College.
In Wisconsin, the job-vacancy survey aided in the development
of mass training programs for 3,000 disadvantaged youths to be trained
for the next 2 years.
At the national level we have used this information to guide the
State agencies in the development of training programs.
In terms of using this information for developing additional jobs
in which the Employment Service could place unemployed applicants,
its use at this time was not too great. Perhaps the most important
reason for this was that the kind of occupations that empk^ers re
ported as having been unfilled for any duration of time were pretty
much the kind of occupations that the Employment Service is having
a very difficult time in filling. But in the higher skilled occupations
where there are applicants, this could be a pretty useful tool.
These studies were used for the Employment Service to structure
the visits to the employers.
I might say here that there had been a lot of apprehension at the
inception of the survey that because the studies were made by the
Employment Service, the employers would not report the data,
because the Employment Service would run out and start developing
jobs. This was an erroneous assumption, and the final results were
a very pleasant surprise to me and others in the Employment Service.
There has been no difference in reporting attributable to this. In the
first round certain areas attempted to develop jobs right after that
first round. In the second round the response in these areas was the
same. In addition, areas where there was no job development had
about the same response rates as those that did. So there was no
effect of that at all.
To begin with, we did not attempt to develop jobs in any establish
ment unless we have an applicant in that occupation. And I am sure
that if we had an applicant in a particular occupation which the
employer has been short for 30 days, he would be most happy if we
were able to provide a qualified applicant to him.
The experimental surveys clearly indicated that the labor supply
in virtually all areas exceeded labor demand by relatively large margins
some 12 and 15 months ago. We don't know’ what the situation is
64
JOB VACANCY STATISTICS
now. These analyses were based on the number of applicants by
occupations that were registered with the public employment office,
and the total job vacancies that this survey provided.
I might mention here that on a national basis the Employment
Service job registrants account for some 80 percent of all workers who
suffer a spell of unemployment. So in terms of the occupational
detail for some 80 percent of the unemployed there is really no better
source of information than the public employment office applicant file,
because the characteristics of these people, such as age and educational
level, are obtained by a very competent person who is trained to do
this.
In Charleston, W. Va., the job applicants outnumbered the job
vacancies by about 10 to 1. For every job vacancy there were 10
job applicants. In only one area—and that was Richmond, Va.—were
the number of job vacancies somewhat higher than the number of
job applicants.
Of course, there was wide variation by occupation from area to area.
Now, in fiscal 1966 we will continue this experimental program in
roughly the same areas with some modification of our schedule.
Wage data will be obtained in all the areas. We will drop the collec
tion of vacancies existing for 7 days but less than 30 days, because
this information was not useful for operations. The number of
vacancies open 7 days was almost the same as the total number of
vacancies. We will retain the collection of vacancies in existence
30 days or longer.
Chairman P r o x m i r e . You say you are dropping the first category
of those where the vacancy is open for less than 7 days?
Mr. C h a v r id . N o ; we will get all of the vacancies regardless of
the duration. And in addition, vacancies that are open 30 days or
longer. So that we will have total vacancies.
Chairman P r o x m i r e . This would be one of labor’s objections,
that you require 7 days for a person to be classified as unemployed,
but that is not symmetrical, because you do not require it to be open
7 days before you fill it; is that correct?
Mr. C h a v r i d . I don't think that is their objection.
Chairman P r o x m i r e . I say that one of their concerns is that it
is not symmetrical.
Mr. C h a v r i d . We are not asking for 7 days or more this year, we
are asking for all of the openings which the employer has now. In
getting this additional information for 7 days or longer, we found
that it was almost the same as the total job vacancies, over 80
percent of all the job vacancies were open 7 days or longer. So it
would be an additional burden on the employer to provide this
information. It is not useful for the operations, if I may say so. It
may be useful for some statistical analysis. But we in the Employ
ment Service are concerned primarily with these data in terms of the
manpower operations.
Now, on the surveys that are now underway we have obtained
some insight as to how they compare with the earlier ones. This is
very preliminary. There has been an increase in job vacancies. Also,
we have asked a specific question of the employer, if he would like
Employment Service assistance to fill his job. And over 60 percent,
on a preliminary basis, said, “ Yes, we would be most willing to
cooperate with you.’7
JOB VACANCY STATISTICS
65
Mr. C h a v r i d . This is about all that I wish to say. And if there
are any questions, I would be more than happy to answer them.
Chairman P r o x m i r e . Thank you very much, Mr. Chavrid.
(Mr. Chavrid’s prepared statement follows:)
P r e p a r e d S ta t e m e n t o f V la d im ir D . C h a v r id , D i r e c t o r , O f f i c e o f M a n
p o w e r A n a l y s i s a n d U t i l i z a t i o n , U.S. E m p l o y m e n t S e r v i c e
It is my pleasure and privilege to appear before you and to place in the official
record some of the wealth of job market information derived from the N ation’ s
first major effort in the field of job vacancies.
Y ou have already heard M r. Frank Cassell, D irector of the United States
Em ploym ent Service, testify on job vacancy data as an operations tool in the
administration of manpower policy. M r. Arthur Ross, Commissioner of Labor
Statistics has approached the subject in conceptual terms with emphasis on
vacancy data as an econom ic analysis tool. M y purpose in appearing before this
Com mittee is to provide the basic statistical support underlying both of these
positions and at the same time to show what these data mean to the on-going and
prospective programs of the Em ploym ent Service.
It is a relatively rare circumstance to participate in the developm ent of a new
avenue of work, much less to witness its emergence as a useful tool in the pursuit
of knowledge and in application to existing problems. The job vacancy program
has provided such an opportunity. I have been privileged to share with my
colleagues the satisfactions arising out of this new program whose potential, I
believe, will eventually permit it to take its place alongside those basic econom ic
barometers which measure the number of people who work and those who are
seeking work.
I have observed that there are no passive onlookers to the job vacancy program .
Such is its attraction and power to m ove the imagination and so manifold its uses
that it has created both a host of devoted adherents on the one hand, and equally
interested partisans on the other, not necessarily of the same size, who wish to
m odify its form in one way or another.
The initiation of the job vacancy program in late 1964 and early 1965 was unique
in at least three respects. First, it was an avow edly experimental effort full o f
unknowns and administrative difficulties yet the scale on which it was tried— the
cities covered accounted for one-fourth of the N ation’s work force— presumed a
very high level of sophistication. Second, the time allotted for its completion,
including conceptual thought, preparation of how -to-do-it manuals, staff training,
gathering, processing, and analyzing the data, and applying the findings to job
market problems, was only one short year. Lastly, the acceptance of this pro
gram by employers, educators, manpower agencies, economists, and the general
public appears to have few parallels in previous efforts of this kind.
The jo b vacancy program marked another milestone. It showed for the first
time, by relatively precise measurement, the im portant role that the Em ploym ent
Service plays in the jo b market. At any point in time, unfilled job orders on hand
in the Em ploym ent Service are estimated to constitute about 30 percent of existing
jo b vacancies in the Nation. This jo b market exposure compares favorably with
some of the highly regarded em ploym ent service systems abroad and occasioned
considerable surprise among domestic and foreign manpower experts.
In assessing the facts revealed by these surveys, I would like to convey a m ajor
caution. The program is so new and the techniques so experimental that we
cannot altogether tell the difference between fact, atypical variation, or perhaps
plain sampling error. Figures will appear throughout this report which cannot
be explained entirely by known econom ic phenomena. Consequently, we suggest
that the data shown be regarded not as precise measurements but as approxi
mations within which the probable answers will tend to cluster. The program
will need to become periodic and the results inspected over time before more
definitive conclusions can be formed.
background
In fiscal year 1965 the United States Em ploym ent Service of the Bureau of
Em ployment Security, in cooperation with the affiliated State employm ent
security agencies and the Bureau of Labor Statistics, conducted an experimental
job vacancy program in 16 metropolitan areas. These areas accounted for approx
imately a fourth of the Nation’ s nonagricultural employm ent. Tw o surveys, the
66
JOB VACAXCY
STATISTICS
first spread over the last quarter of calendar year 1964 and the second in April
1965, were conducted in each of the following areas:
Baltimore
Birmingham
Charleston, S.C.
Charleston, W . Va.
Chicago
Kansas City
Los Angeles
M iami
Milwaukee
M inneapolis-St. Paul
New Orleans
New York
Philadelphia
Portland, Oreg.
Providence
Richm ond
The primary emphasis of the experimental job vacancy program was operational
in nature in an attem pt to improve the functioning of the jo b market. The public
Em ploym ent Service has a large volume of manpower inform ation available on a
local area basis. This information consists of current em ploym ent and unem
ploym ent statistics, listings of job openings filed by employers, estimates of future
occupational needs as estimated by the area skill surveys and training needs
surveys and inform ation on jo b applicants registered with the Em ploym ent
Service. Inform ation on current jo b vacancies would be an im portant supplement
to this existing body of local area data and is needed for a more efficient and
effective matching of workers and jobs.
In the 1965 experimental program, information on total job vacancies, vacancies
existing one week or longer and vacancies existing one month or more were col
lected on what will be referred to as the standard collection schedule.
In addition, several types of variant schedules were used. In New Y ork the
schedules did not collect information on vacancies existing one week or longer.
In Chicago, Miami, Baltimore and Minneapolis-St. Paul, the standard collection
schedule was expended to provide for collecting wage data for the vacancies
reported.
Still another experiment was to determine what effect the request for different
types of inform ation had on employer response rates and which type of schedule
was most successful in obtaining the data. Five of the labor areas surveyed
collected the job vacancy information on a split area basis (i.e., half the sample
received the standard schedule and the other half received a variant schedule).
T he data collected in the half of the sample which received the variant schedule
were tem porary vacancies in Kansas City, part-time vacancies in Charleston,
W . Va., w^age information in Los Angeles, all three of these items in New Orleans,
and in Philadelphia, inform ation on total vacancies only.
Tests were also made to determine whether employer response was affected by
conducting the job vacancy surveys entirely by mail from the central office of the
State em ploym ent service as compared with conducting the surveys from local
offices which involved employer visiting, in addition to mailing, both to introduce
the program and to follow up on nonresponse. Accordingly, three of the area
surveys were conducted from the State office. T o the extent possible, the other
13 areas attem pted to tie operating techniques such as job order developm ent and
placem ent to the collection of vacancy data.
SAM PLE
AND
RE SPON SE
Com plete inform ation on the nature of employer response is available for 15 of
the 16 areas included in the experiment. Although completing the survey, the
data from New Orleans are om itted because of technical problems encountered
early in the study.
Approxim ately four out of every five employers included in the sample of the
remaining 15 areas provided inform ation on their job vacancies. The overall
response rate was 78.4 in the first round and 81.1 percent in the second round.
In no area was there less than 60 percent response in either round, and three areas
in the first round and four in April had response rates exceeding 90 percent.
In both rounds of surveys, slightly more than half of the total nonagricultural
em ploym ent in the 15 areas was included in the sample and the percent of sampled
em ploym ent in the responding establishments was generally very high. In fact,
it was close to 100 percent in Milwaukee and Portland, Oreg. (Table 1).
67
JOB VACANCY STATISTICS
T a b l e 1 .—
Percentages of establishments and employment in sample and responding
firms, 2d round {April 1965) of job vacancy survey, by area
Universe
Number
of estab
lishments
Baltimore__________ . . . .
Birmingham..^ ______
__ _
Charleston, S.C____ _ . . . __
Charleston, W . Va_____ _
Chicago____
_
. ____
Kansas City .
_ _ _
Los Angeles____ _ _ _ _
Miami.. __________ __ _____
Milwaukee..
_ _ __
Minneapolis-St. Paul___ __ .
New York __ _
_ _
Philadelphia. __ _____ __ __
Portland___ _
______
Providence__ _____
____
Richmond. ______
.
_ __
12,807
4,462
1,506
1,647
47,800
8, 570
57, 520
9, 790
8,829
13,109
114,920
30,559
8,140
7, 579
3,999
Employ
ment
541,321
178,905
51, 457
68,539
2,163,300
535,436
2,203,070
309,598
414,875
547, 510
3, 336,690
1,349,104
249,897
222,687
168, 526
Percent of universe in
sample
Respondents: Percent
of sample
Establish
ments
Establish
ments
6.3
12.8
35.6
37.1
4.3
8.1
3.4
11.0
9.4
5.5
4.9
4.8
12.2
10.0
20.1
Employ
ment
57.4
54.3
66.8
77.5
51.1
53.3
50.6
59.2
60.7
52.1
50.4
41.3
59.9
54.2
67.8
85.4
92.7
83.6
82.8
64.5
85.7
93.8
88.1
93.9
86.4
72.3
87.1
97.8
60.2
89.3
Employ
ment
84.4
94.2
92.5
96.2
71.3
87.4
88.5
85.7
98.6
69.3
74.2
76.9
97.1
65.4
94.1
In both rounds of the survey, over 99 percent of the schedules returned by
employers were usable. In the first round, 7 areas, and in April, 11 areas, re
ported that all schedules submitted were usable. Furthermore, in the “ split”
areas which used both standard and variant schedules, there was no appreciable
difference in the percent of employer response. In no case was there more than 4
percentage points between the groups using the different schedule, either in terms
of response rates or percentages of respondents reporting vacancies.
Prior to initiating the F Y-1965 job vacancy program, a test survey conducted in
Chicago probed the reaction of employers to furnishing vacancy data to the local
employm ent service office. Despite the overwhelmingly favorable response,
there still was a residue of apprehension that employers might feel their right to
hire through whatever channels they wished would be placed in jeopardy and,
furthermore, that the confidentiality of the reported data could be violated.
Subsequent results of the broader 16 area study merely confirmed the earlier
Chicago finding. An analysis of the data from the two surveys indicated that
em ployer response was one of the most favorable in the history of data collection.
T o avoid any violation of privacy, the employm ent service in the respective areas
were instructed not to try to fill any jobs unless they had available applicants
on hand and had obtained prior employer permission to make referrals.
Milwaukee and Portland, two areas which mounted strong operational followup
efforts, had response rates well above 90 percent in both rounds of surveys. There
was no attem pt at job order development during the first round of surveys in five
areas (Charleston, S.C., Chicago, Los Angeles, Miami and Minneapolis-St. Paul).
In the second round, two of these areas (Los Angeles and Minneapolis-St. Paul)
included a jo b order development effort as part of the experiment. The lowest
response rate in round one and the second lowest in April was in Chicago which
had no operational followup. The other two areas which did not do any opera
tional followup were in the middle third, in terms of response.
n a t u r e o f jo b v a c a n c ie s
About one-fourth of the almost 20,000 employers sampled in each of rounds one
and two of the fiscal year 1965 surveys reported job vacancies. There were,
however, distinct differences in proportions of respondents reporting vacancies in
terms of size of establishment. In both surveys some 57 percent of all large
establishments reported job vacancies as compared with about one-fourth of the
medium and one-tenth of the small establishments.
Differences in the proportion of firms reporting vacancies varied widely in the
areas surveyed, ranging from about 10 percent of the responding establishments in
68
JOB VACANCY
STATISTICS
Charleston, W. Va., to almost two-fifths of responding establishments in
Milwaukee.
Employers were requested to list their vacancies by plant job titles which
were then converted by ES personnel into the codes and nomenclature defined in
the 1949 edition of the Dictionary of Occupational Titles. Occupations are classi
fied in the Dictionary into nine m ajor categories which are given a one digit code.
These nine m ajor categories are: professional and managerial, clerical and sales,
service, agricultural-forestry-fishing, skilled, semiskilled, and unskilled. The nine
m ajor categories are also defined more specifically in terms of 3-digit, occupational
codes, of which there are some 580 and then 6-digit codes of which there are some
9,000 different occupational titles.
In both rounds of the survey, vacancies were found at all skill levels. Clerical
and sales occupations accounted for nearly 30 percent of the vacancies in both
rounds, the largest number for any occupational group. The second largest
representation was among professional and managerial. (Table 2.)
While vacancies were reported at every skill level, their magnitude and distri
bution varied from area to area. For example, approximately half of all job
vacancies in Charleston, W est Virginia, were for professional and managerial
workers, while 5 percent were for semiskilled workers. On the other hand,
roughly half of the job vacancies in Providence were for semiskilled workers, and
vacancies for professional and managerial personnel made up less than 5 percent
of the total.
T o ta l, all o c c u p a tio n s . _____________
B a lti
m ore
B ir Charles Charles
m in g
ton,
ton,
ham
S .C .
W . Va.
C h i
cago
M in n e K ansas
Los
M ia m i
M il
apolisC ity A n geles
w aukee
St.
Paul
N ew
Y ork
Ph ila
d e lp h ia
P ort
land
P rov i
dence
R ic h
m on d
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
21.8
27.3
9.6
1.2
15.1
18.3
6.7
13.1
24.8
12.3
.3
17.8
21.0
10.7
24.4
22.5
13.8
.2
13.2
11.8
14.1
16.8
15.3
7.9
.6
32.5
16.1
10.8
44.0
34.2
9.2
1.4
4.4
5.0
1.8
20.8
27.1
8.6
.8
23.3
15.3
4.1
19.7
38.1
4.4
2.4
19.6
13.3
2.5
27.0
24.2
10.8
.2
8.4
24.6
4.8
10.7
25.2
16.6
.1
14.2
26.2
7.0
16.8
15.5
16.3
2.0
20.7
14.4
14.3
26.4
35.7
5.5
14.9
4.1
9.5
3.9
26.4
34.6
6.4
0
10.6
15.4
6.6
18.2
19.6
12.9
.4
22.7
20.8
5. 4
19.6
29.9
18.0
.9
13.2
10.5
7.9
3.5
11.0
12.3
.1
18.1
45.4
9.6
14.5
26.2
9.4
.3
14.1
16.7
18.8
STATISTICS
Profession a l_________________ _____ _ _______
C lerical and sales, _________
________
S e r v ic e .-. __ _ __ _
________
__ __
A gricultu ral, forestry, e t c ___ _ _ _ _ _ _
_
S k illed _________________________
_ ___
S em isk illed____ _
U n sk illed . _________ __ _ _ _ _ _
T o ta l15
areas
VACANCY
O ccu p ation a l groups
Percent distribution of vacancies by major occupational group and area, April 1965
JOB
T a b l e 2 .—
O
CO
70
JOB VACANCY STATISTICS
Demand for certain occupations was general, while for others it was concentrated
in specific areas. Alm ost all areas reported a demand for nurses. In New Y ork
City alone, there were almost 1,000 vacancies for industrial engineers while in
13 other areas there was a com bined total of only 300 vacancies for this occupation.
Similarly, about 1,600 vacancies for draftsmen were reported in New Y ork City
as against 1,000 in 13 areas combined. Because of the im portance of the apparel
industry, there were many vacancies for tailors and tailoresses in Philadelphia
and New York, but in the other areas the demand for these was quite small.
Furthermore transportation laborer openings were a sizable portion of unskilled
vacancies in New Y ork City. Elsewhere the need for this type of laborer was
relatively light.
In general, the vacancies reflected the industrial mix of the areas. M ore
vacancies existed in the manufacturing industries than any other segment, but
this was to be expected because in m ost of the 15 areas, factory jobs are the
largest segment of employment. A bout 30 percent of the non-farm em ploym ent
in these areas were in manufacturing, and about the same proportion of vacancies
was also found in manufacturing.
Am ong the individual areas, about 40 or more percent of the vacancies were in
manufacturing in highly industrialized areas like Providence and Chicago. The
bulk of the vacancies in Charleston, S.C., the State capital and site of many
Federal governm ent activities, was in the government sector. Alm ost half the
vacancies in Miami, a resort area, were in trade and service.
Seasonal differences affected the industrial pattern of vacancies. This was
especially true in the trade sector where there was a significant drop in the number
of vacancies between rounds one and two. In the earlier surveys, a great many of
the vacancies reflected the need for additional help for the Christmas season.
H A R D -T O -F IL L
V A C A N C IE S
The job market is always in a state of flux as people enter and withdraw from
the labor force, or change jobs, and as employers attem pt to replace the workers
who leave or are offered new positions. Therefore, no matter what the state of
the econom y in an area at any given time, there are always jo b vacancies avail
able. However, many of these vacancies are filled in a relatively short period of
time depending upon the attractiveness of the job offer, the availability of labor
supply, the tightness of the job market, and other factors. On the other hand,
some vacancies may take a long time to fill because of a lack of workers with
requisite skills for the open positions, or because of low wages, poor working
conditions, poor plant location, unrealistic hiring specifications, etc. These
hard-to-fill vacancies, rather than the gross figure of vacancies, are the primary
tool to help identify manpower shortages and aid in manpower program action
to match employer demands with the occupational supply of workers available.
The average amount of time required to fill a vacancy varies, of course, with
the occupation involved, the econom ic situation in the area, wages offered, and
many other factors. In the Chicago feasibility study previously mentioned,
another of the questions which employers were asked was to indicate by occupa
tion an average length of recruitment time. It was presumed that if the vacancy
extended beyond this period, it would pass into the category of “ hard-to-fill” .
Employers reported that the longest period of time— one month— was needed
to fill the professional, technical, and managerial positions and the shortest period
of time— one to tw o weeks— was for unskilled occupations. Therefore, to ac
curately define hard-to-fill vacancies, a different period o f time is needed for
each category of skill. However, it was conceded that sim plicity of form ou t
weighed other considerations and the decision was made to define “ hard to fill”
as a vacancy which existed for 30 days or longer. Although the resultant list
of jobs vacant 30 days or longer is a most useful indicator of manpower shortage,
further research needs to be devoted to this field in view of the known variation
in average period of recruitment.
In both rounds of the F Y -196 5 surveys, over half of all vacancies had been
unfilled at least one m onth and were therefore considered hard to fill. The
percentages of jo b vacancies that had been open for 30 days or longer in April
1965 ranged from a low of a little less than one-third in Minneapolis-St. Paul
to a high of nearly nine-tenths in Baltimore. Due to the pioneering effort o f
the survey, it is not yet known whether the wide variation on these percentages
resulted from factors relating to employer reporting, econom ic differences between
areas, or survey procedure.
71
JOB VACANCY STATISTICS
B y occupational group, the proportion of hard-to-fill vacancies ranged from
slightly over 70 percent in the professional and managerial and skilled groups to
about 45 percent in the clerical and sales groups. The proportion of vacancies
that was hard to fill in the other occupational groups may be found in Table 3.
While the overall proportion of long-duration vacancies was high, certain
detailed occupations w^ere considerably higher. Am ong the professional and
skill groups, for example, 9 out of 10 vacancies for trained nurse and for tool and
die maker were open 30 days or longer while 80 percent of the vacancies for civil
and chemical engineers and tailor were hard to fill. Occupations in the semi
skilled group which were in great demand and hard-to-fill were sewing machine
operators, chauffeurs and drivers, filer and grinder, and occupations in the treat
ment of metals and in the service group, there were hospital attendants, practical
nurses, janitors, porters, cooks, police and sheriffs and bailiffs. Workers in the
manufacture of clocks, watches and jewelry and transportation laborers were the
unskilled occupations with the greatest proportion of vacancies open 30 days or
longer.
There were also differences by area in the proportion of hard-to-fill vacancies
by occupational group. For example, in 3 areas in each round, 9 out of 10 of the
vacancies in the professional group had existed 30 days or longer. On the other
hand, the percent of long standing professional vacancies was quite low in Los
Angeles and New York.
T a b l e 3 .—
Hard-to-fill1 vacancies as a percent of total vacancies by area and occupa
tional group, A pril 1965
Area
Baltimore.._______ __ __ . . .
Birmingham____
,
Charleston, S.C_-_ _______ __ _
Charleston, W . V a ...
___ __
Chicago.. _ __ _
Kansas City__ __
_.
Los Angeles.. - _ . .
Miami
______ _ _
Milwaukee______ . . . ______
Minneapolis-St. Paul __ __ _
New York . . . . . .
_ _ __
Philadelphia... .
_ __
______ _
Portland. .
Providence___
__
Richmond___ _____
___
Total
Profes
all occu sional and Clerical
pations
mana
and sales
gerial
89.4
65.0
80.9
81.2
63.1
51.9
38.4
76.4
59.2
31.3
34.9
(2)
62. 2
84.2
86.7
94.8
75.8
88.2
92.0
81.6
79.9
50.6
76.6
81.6
59.7
35.3
(2)
81.3
91.8
83.1
Service
Skilled
94.2
60.4
79.3
76.7
45.5
3.2
31.0
77.2
41.1
34.4
27.0
(2)
32.6
100.0
71.5
92.6
81.1
79.5
48.3
77.5
64.7
26.1
84.5
63.3
46.5
36.1
(2)
82.5
86.7
91.2
95.7
40.0
76.9
83.0
50.8
34.9
22.3
61.5
41.6
27.5
22.8
(2)
52.3
57.6
82.8
Semi
skilled
85.9
66.2
82.0
12.1
57.8
74.3
54.1
90.6
61.6
15.7
49.4
(2)
68.1
87.4
92.4
Unskilled
64.6
74.4
72.5
81.0
31.6
2.9
6.0
58.4
60.3
12.6
68.6
(2)
83.8
71.5
93.9
1 Vacancies in existence 30 days or longer.
2 Not available.
CO RRESPOND ENCE
BETW EEN
O P E N IN G S
AND
V A C A N C IE S
A m ajor purpose of the job vacancy experiment was to determine the extent
of penetration of the ES local offices into area job markets, i.e., the percent of
total vacancies in the area which were accounted for by ES unfilled openings at
a single point in time. The comparison, by detailed occupation of vacancies and
unfilled job openings, revealed a surprising degree of penetration, namely, that
ES unfilled job openings at the time of survey were equivalent to about one-third
of the aggregate total of job vacancies in the surveyed areas. Jobs in domestic
service were excluded from both arrays.
On an individual area basis, however, there were wide differences in the extent
of ES participation in the vacancy total. In the April survey, over 40 percent
of the job vacancies were listed in the local ES offices in Kansas City, Portland,
Milwaukee, Minneapolis-St. Paul, Birmingham, and Los Angeles, while less
than 25 percent were listed in Charleston, West Virginia, Charleston, South
Carolina and Richm ond (Table 4).
72
JOB VACANCY
STATISTICS
4 . — Total job vacancies, unfilled employment Service job openings, and job
applicants registered with the Employment Service, by area, A pril 1965
T a b le
Number of
unfilled
Number of Employ
ment
job
vacancies Service job
openings
Job
openings
as a
percent of
vacancies
Employ
ment
Service job
applicants
2,407
866
211
97
9,725
1,712
9,348
998
3, 464
3,305
18, 727
4,239
1,3-54
0)
643
32.6
46.7
14.9
14.8
34.8
60.4
41.5
24.7
40.6
41.0
40.0
28.2
46.8
0)
16.0
25,095
6, 775
2,432
5,894
60, 576
15,306
0)
6, 506
11,519
21,697
65,469
2 61,300
12,990
0)
1,706
Baltimore
___________ _______________
Birmingham
_ _ __ ______ __________
Charleston, S.C.. .
__ ____- - __ _
Charleston, W . Va_ __ _ _
_____
Chicago ___
__ ____ _ _
- ___ _
Kansas C i t y ___ _______ __ __
__ Los Angeles__
___ - _ __ __ _____ ___________ __ _ _ ___ _. __ Miami
__
Milwaukee
__ _ _
Minneapolis-St. Paul _ _
_ __
___
New York________ _
_ ____
__ __ _
Philadelphia _ _____ __
___ __
Portland____ _________
___
Providence__
__ _ ~ _ - --Richmond... _ _ _ _
_ _ _ _ _
7,381
1,854
1,413
655
27,947
2,836
22, 506
4, 037
8, 542
8, 061
46, 780
15, 037
2,893
5,663
4,015
Ratio of
Employ
ment
Service
applicants
to job
vacancies
3.40
3.65
1. 72
9.00
2.17
5.38
0)
(0
1.61
1.34
2.69
1.40
4.08
4.99
.43
1 Not available.
2 Data on job applicants not entirely comparable with other areas.
There was a fairly close correspondence between unfilled openings and vacancies
on a m ajor occupational group basis (one digit D O T C ode). As can be seen in
Table 5, the distribution o f vacancies and openings for professional and man
agerial and clerical and sales was quite similar. On the other hand, in the skilled
group, vacancies accounted for a higher proportion of the total than openings,
reflecting the fact that hiring in some skilled occupations is traditionally done
through union hiring halls and other channels rather than through the E m ploy
ment Service.
Another measure o f similarity, and perhaps more im portant, is reflected in the
fact that only a relatively few detailed occupations account for the bulk o f the
vacancies and jo b openings. Thus, only 43 (3-digit) occupations comprise 66
percent o f the total job vacancies and these same occupations account for 58
percent o f the jo b openings. Table 6 expresses this relationship. The corre
spondence between vacancies and unfilled openings was n ot particularly close
for many occupations having small numbers of vacancies or unfilled openings.
Am ong the individual areas, the percentage distributions of openings and vacan
cies on a broad occupational basis were also quite similar, for some groups more
than others. Similarities were notably present in the service and clerical and
sales occupations.
5.— Percentage distribution of job vacancies and ES unfilled job openings by
major (1-digit) occupational groups, 1st and 2d rounds of job vacancy surveys
T a b le
1st round
2d round
Major occupational group
Job
vacancies
Total___ __
_ ________________________
Professional and managerial.
__ __ _ _____
Clerical and sales________ ______
...
_ _ _
Service 1_____ __ _________ _ _ _ ________
Skilled________________________________________
Semiskilled________ ____
_______
Unskilled.. . _________
. _ ______
O t h e r .._____ _______
______
_____
1 Excludes domestic service occupations.
Unfilled
openings
Job
vacancies
Unfilled
openings
100.0
100.0
100.0
100.0
21.0
27.5
8.2
16.9
17.3
8.8
0.3
21.9
28.2
10.2
12.7
20.2
6.7
0.1
22.0
29.0
9.5
14.1
16.2
7.7
1.5
23.8
27.6
10.7
11.3
18. 5
7.8
0.3
73
JOB VACANCY STATISTICS
T a b l e 6 .—
Percent job vacancies and unfilled, openings in selected 3-digit occupations
representative of each major occupational group
Occupations (for specific listing, see next page)
Percent job
vacancies
Total, all groups, 43 occupations-____________________________________
Percent
unfilled
openings
66
58
62
78
62
58
71
2 56
65
64
47
50
64
27
Professional and managerial, 11 occupations (including 6 categories of engiClerical and sales, 11 occupations___ _________ __ _ _____ ______ ________
Service,1 4 occupations_____ _ __________________________ ________________
Skilled, 8 occupations____
____ __________ ______________________________
Semiskilled, 5 occupations _ _______________________________________ _______
Unskilled, 4 occupations _ ________________________ ___ ___ __ _____
1 Excludes domestic service occupations.
2 May be some distortion because of sample inflation.
L is t
of
43 S e le c t e d 3 -D ig i t O c c u p a tio n s
Professional and Managerial: engineers (metallurgical, chemical, civil, electrical,
industrial, and mechanical), social and welfare workers, trained nurses, drafts
men, laboratory technicians, healers and other medical occupations.
Clerical and Sales: bookkeepers and cashiers, clerks-general office, general industry
clerks, office machine operators, secretaries, stenographers, and typists, stock
clerks, salesmen-insurance, salesclerks, salespersons, sales-except to consumers.
Service: waiters and waitresses, kitchen workers, attendants-hospital, porters.
Skilled: machinists, tool and die makers, machine shop occupations, welders and
flame cutters, carpenters, plasterers, mechanics and repairmen-motor vehicle,
mechanics and repairmen-other.
Semiskilled: textile products occupations, machine shop occupations, occupations
in the treatment of metals, chauffers and drivers, occupations in laundering,
cleaning, etc.
Unskilled: occupations in meat products, packing and filling occupations, transpor
tation equipment laborers, warehousing and related occupations.
o p e r a t io n a l u se s o f t h e d a t a
The primary emphasis of the experimental job vacancy program was for Em
ployment Service operational use in the job market. In the first tw o rounds of
the surveys, the most extensive use of the jo b vacancy information was in con
nection with manpower training programs. The surveys disclosed unmet needs
for workers in a wide range of occupations, many of them suitable for M D T A
institutional or apprentice training. Data on vacancies open for one month
or more (defined as hard-to-fill vacancies) rather than on total vacancies were
most useful in determining training needs.
In Baltimore and Kansas City, the job vacancy inform ation was used as a basis
for establishing vocational training courses in the schools. The Oregon agency
reported that data from the first round of jo b vacancy surveys were used by
the State Departm ent of Education to develop vocational education and other
post high school courses for the Portland Com munity College. In Wisconsin,
the job vacancy survey aided in the development of a mass training program
for 3,000 disadvantaged youth to be trained over the next tw o years. Because
of the job vacancy survey, the number of adults to be trained in"Milwaukee was
expanded in 12 occupations, including industrial laboratory assistant, production
foundry worker, machine shop inspector, nursing assistant, cook, electric appliance
repairman, and repairman for lawn and boat motors.
The Urban League o f New Orleans had been endeavoring to establish an onth e-job training program, to be financed jointly by M D T A funds and a nonprofit
organization. W ith the use of jo b vacancy data, the Louisiana Em ploym ent
Service was able to furnish the Urban League with a list of about 3 0 occupations
which were suitable for on-the-job training. On the basis of inform ation obtained
from the New Orleans jo b vacancy survey, a second list o f 5 0 occupations with
potential for apprenticeship training was developed and made available to the
Louisiana Board of Apprenticeship. As a result, several on jo b training (OJT)
projects were planned. Tw o of these were for a group of specialized occupations
in the dye industry and for auto repairmen.
74
JOB VACANCY STATISTICS
A t the national level, the USES used the jo b vacancy data to prepare guides
for State agencies to follow in the development of manpower training programs.
Local employm ent offices in eleven of the surveyed areas attem pted to develop
new jo b orders from the vacancy data. Placements as the result of the jo b
developm ent efforts were relatively small. M any vacancies were already filled
before the Em ploym ent Service could attem pt jo b order development. Job
order developm ent was limited by the fact that many of the vacancies were
already on file in the ES local offices; in attributing placements to the job vacancy
program, States were instructed not to count placements for an occupation until
such placements began to exceed the jo b order for that occupation already on
file in the ES from the employer. In addition, a number of vacancies were
closed to ES assistance due to union hiring rules or the existence of civil service
regulations and thus reduced the number of vacancies available for jo b develop
m ent and placement activities.
Another reason for the limited number of placements was the fact that so many
o f the vacancies were in jobs for which there were no available applicants. Fi
nally, although the States were instructed to attem pt jo b developm ent and place
m ent, the newness of the program, the need for initiating complex administrative
and analytical procedures, and the short time permitted to get the program under
way, results tabulated and reported, of necessity kept the States from devoting
the time that would be given if the program were operational over a period of time.
Vacancy data were useful, however, in helping E m ploym ent Service staff to
structure their contacts with employers more effectively. Even on the basis of
tw o surveys, agencies found that the inform ation suggested new areas for p ro
m otional activities and led to increased em ploym ent possibilities for applicants.
In New York, for instance, the Em ploym ent Service learned, as the result o f the
surveys, that job vacancies existed in establishments not known by the agency
to have vacancies in such occupations.
Several areas participating in job vacancy surveys used the job vacancy find
ings in counseling. Some of these areas were Portland, Los Angeles and R ich
m ond. Furthermore, the Wisconsin agency used the job vacancy data along
with other jo b market information in a bulletin on vocational outlook in M il
waukee. National occupational outlook information would be com bined with
local data in the form of an occupational monograph.
Although some of the participating agencies expressed the opinion that the jo b
vacancy data would be useful in worker mobility, there was not sufficient time
to integrate the existing vacancy program into the established interarea clearance
machinery to take advantage of job imbalances between areas. It is intended
in future surveys, to summarize the data in a manner useful for specific m obility
projects and for the whole Em ploym ent Service interarea recruitment program .
W A G E A N A L Y S IS
In much of the pre-experimental discussion concerning the validity of the
vacancy figures, it was advanced by some economists that many of the vacancies
reported would carry substandard wage rates. This, they added, would tend
to overstate the econom ic significance of the vacancy data or conversely, render
them meaningless as a gauge of job opportunity within the area. Consequently,
it was decided that in five areas, the vacancy experiment would attem pt to collect
wage data. For each of the occupations for which an array of wage data was
available from employers, Em ploym ent Service placem ent specialists were asked
to provide the going rate for the occupation in the area.
The job vacancy surveys demonstrated that (1) employers are able and
willing to furnish wage rates in conjunction with their vacancy inform ation; and
(2) that wages offered, by and large, were in line with going rates for those occu
pations in the area.
In round one, four reporting areas (Chicago, Los Angeles, Miami, and M inneapolis-St. Paul) provided wage data for approximately two-thirds of the total
vacancies reported in these areas. By area, the proportion of vacancies with
wage data ranged from 55 percent in Los Angeles to almost 85 percent in M iami.
In round two, employer response im proved in all areas where wage data were
collected except in the Minneapolis-St. Paul area w'here it remained at a very
high level. Five areas collecting wage data in round 2 (the above four plus
Baltimore) reported that 83 percent of the vacancies included wage information.
The proportion of vacancies with wage data to total vacancies reported ranged
from 70 percent in Minneapolis-St. Paul to over 90 percent in Miami.
JOB VACANCY STATISTICS
75
The sample design made it possible to measure in another way the im pact
on reporting of the collection of wage data. In Los Angeles wage data were
collected for one-half the sample and not for the other half. There was virtually
no difference in the employer response rate of the wage sample as against the
other half of the sample for which no wage data were requested.
In general, the wages listed for job vacancies were in line with entry rates as
determined from local office records. Of the vacancies for which comparable
entry wage rates were available, approximately 15 to 20 percent of the vacancies
offered wages below the usual range of entry rates for the occupation in the area.
Because of the survey design, these percentages were based only on actual
vacancies and wages reported rather than on an inflation of the vacancies to a
universe total. If it were to be subsequently demonstrated that smaller firms
pay lower wages than the larger firms, it would have the effect of understating
somewhat the percentage of substandard wage vacancies. This earlier limitation
will be corrected in the fiscal year 1966 experimental surveys. All areas are
requested to collect wage data and inflation procedures will permit a more accurate
measure of both substandard and above-average wage rates attached to the job
vacancies.
U N E M P L O Y M E N T , JO B A P P L IC A N T S , A N D JO B V A C A N C IE S
Econom ic policy in the United States has been designed to maximize em ploy
ment opportunities. While in sympathy with the policy, various elements
have been in disagreement as to the best methods for achieving this goal. Does
joblessness arise from a lack of aggregate demand, that is, are there not enough
jobs for all those who want them? Or does unem ploym ent stem primarily from
structural imbalances, that is, an inability to match existing jobs with existing
workers because of skill differences, geographical or commuting problems, or
such factors as wages, restrictive hiring specifications, and the like? We hoped
the experimental vacancy program, would provide some more insights on some
of these questions, but much more needs to be done along these lines.
As part of this effort, inform ation on ES job applicants was tabulated for each
area and related to the corresponding job vacancy data by detailed (3-digit)
occupation. On a national basis, Em ploym ent Service jo b registrants account
for some 80 percent of all workers who suffer a spell of unemployment.
The two experimental surveys clearly indicated that the labor supply in vir
tually all areas exceeded labor demand by a relatively large margin 12 to 15
months ago. Here again, there existed marked variations by area. A t one
extreme, Charleston, W. Va., reported a ratio of 10 applicants for each jo b
vacancy. At the other extreme, Richm ond was the only area in which the number
of job vacancies exceeded the number of applicants.
The variations in the relationship of vacancies and jo b applicants unem ployed
by occupation was even more marked. For the 15 areas as a whole, there were
fewer applicants than openings in the professional and managerial category.
There were about twice as many applicants as vacancies in the clerical and sales
and skilled categories, and 3 or 4 times as many applicants as vacancies in the
semiskilled and service groups. As expected, surpluses were most pronounced
in the unskilled occupations, with over six times as many applicants for every
vacancy in round 1 and almost 5 applicants to 1 vacancy in round 2.
The comparison of vacancies and applicants by occupation indicated that
there were definite skill imbalances in the econom y. Job vacancies were most
often found in the higher skilled occupations, while the bulk of the applicants
were located at the lower end of the skill spectrum.
Despite the overall excess of applicants, there were many shortages for specific
occupations in these areas. In Kansas City, there were 4 times as many vacancies
for nurses as there were applicants, and 3 times as many openings for welders as
there were applicants. Even in Charleston, W. Va., where joblessness was
quite high, several hundred engineers were needed but only 9 applicants were
available.
Clerical and sales workers appeared to be in short supply in New York, yet
there were 3 applicants for bookkeeping jobs for each vacancy. In Milwaukee
there were more vacancies than applicants in the service group, although applicants
for janitor and porter and barber and beautician jobs exceeded the number of
vacancies.
Although there were 10 times as many skilled applicants as vacancies in M inneapolis-St. Paul, there were almost 3 times as many vacancies for machinists
as there were available applicants. In Milwaukee where applicants were slightly
76
JOB VACANCY
STATISTICS
greater than vacancies in skilled occupations, there were fewer applicants than
vacancies for tool and die maker, mechanics, masons and welders.
There were also several cases of interarea occupational imbalances. For in
stance, while a surplus of applicants for welders and mechanics existed in Charles
ton, W. Va., there was significant shortage in these occupations in Milwaukee.
In enumerating these many variations by area, I recognize that I may have
indulged in excessive detail. M y purpose, however, was deliberate. I wished
to show that the use of aggregate figures alone, either for the 15 areas as a whole
or for a larger conglomerate, would conceal vital differences between areas and
effectively limit the application of manpower program actions.
JO B
VACANCY SU RV EY
IN 1966
The fiscal year 1966 experimental program is being conducted in two sections.
In 12 labor areas (including the entire State of Wisconsin), the experiments are
designed primarily to develop job vacancy data for use in conjunction with em
ploym ent service operations. The areas will be the same as those surveyed in
1965 with the exception of New Orleans, Providence, and Philadelphia. In tw o
experimental surveys, in M aryland and Connecticut, the collection procedure will
be tied to the cooperative labor turnover program. In both cases, the data will
be collected in the field by the State em ploym ent services.
A revised sampling design was developed for the 12 area phase of the program
to im prove the accuracy of the estimates of job vacancies by occupation. Firms
were sampled by 2-digit SIC industry and stratified into six specific size groups.
These six groups will be the same for each area surveyed. The revised sample
design will increase the number of small firms included in the surveys and will
better represent the individual area’s industrial mix.
The survey data collection form was also redesigned for easier understanding
by the employers. The definition of vacancies open 30 days or longer was
clarified because there were a number of instances in the first tw o surveys where
employers thought this referred to the duration of the job itself rather than the
duration of the vacancy. Data on vacancies of one week's duration proved to
be of little operational or analytical value and is not being collected in the current
program.
Because the experiments for collecting wage data proved feasible and because
of the great interest expressed in these figures, wages offered for the reported
vacancies will be collected in all surveyed areas in 1966.
There will be no ‘ 'split” areas using both a standard and a variant schedule as
in the F Y 1965 program. Instead, the same survey form will be used for all
firms in an area.
Since data on part-time and tem porary vacancies proved inconclusive last
year, these figures will be gathered once again in tw o areas to test their feasibility
of collection. Inform ation on tem porary vacancies will be collected in Miami
and part-time vacancies in Chicago. In addition, the Wisconsin agency will
collect data on a Statewide basis and for Milwaukee separately, on ‘ ‘other job
openings,” i.e., positions for which a firm is actively seeking workers, but which are
not immediately available (anticipated or future jo b vacancies). The standard
definition for a job vacancy includes only those jobs immediately available.
On the 1966 schedule, a box is now provided in which the em ployer can indi
cate his desire for em ploym ent service assistance in filling the vacancies listed.
This is expected to facilitate jo b development without impairing relations with
the cooperating respondents since jo b development will be conducted with firms
specificalh7- requesting it.
Some preliminary findings of the April 1965 experiment are available at this
tim e:
Because of the new sampling design, the sample of firms was larger in all
areas. It was expanded approximately 2 to 3 times in the largest areas in the
survey— New York, Chicago, and Los Angeles— and to a lesser extent in the
smaller areas.
In 8 areas and the State of Wisconsin, preliminary tabulations are available
on the number of respondents reporting vacancies. Although about the same
proportion of responding firms reported vacancies as last year— 1 out of every
4— there was a large variation among the areas. In Charleston, W est Virginia,
where unem ployment is relatively high, only about 19 percent o f the respondents
reported vacancies. On the other hand, the proportion of employers with vacan
cies was about 43 percent in Richm ond, Minneapolis, and Chicago, each low
unem ployment areas.
77
JOB VACANCY STATISTICS
Approxim ately three-fifths of the employers reporting vacancies requested em
ploym ent service assistance in filling them. The proportion of employers asking
for ES aid in filling their vacancies, however, ran as high as 75-80 percent in Los
Angeles and Richm ond.
The preliminary estimate o f reported jo b vacancies is almost double those
reported in April 1965 for the same areas. Although some of the increase in
the number of vacancies can be attributed to the fact that more emplo.yers were
surveyed this year, the increment since last April was so large as to indicate that
at least a portion of this was due to im proved econom ic conditions and a tighten
ing of the jo b market. It should be remembered, however, that these data
relate to reported vacancies only, and inform ation is not available at this time
concerning the number o f vacancies when statistically inflated to represent
area-wide totals.
The 1966 surveys, as the Committee recognizes, are being conducted in a job
market which has tightened considerably from the previous year« The national
unem ploym ent rate between April 1965 and April 1966, tw o of the months in
which vacancy surveys were made, had declined from 4.8 percent to 3.7 percent.
As sizable inroads are made into the N ation’ s available manpower pool, the
operational problems of placement agencies grow more difficult. A t the same
time, the need for specific knowledge about labor demand and job shortages
grows more insistent. Although far short of the total picture, the 1966 job
vacancy surveys, by covering 25 percent of total nonfarm wage and salaried
workers, will nevertheless provide information on a substantial segment of the
job market.
The Com mittee will be interested to know that in response to continuing public
demand, the United States Em ploym ent Service and several of the State agencies
participating in the experimental effort have prepared articles on their job vacancy
program findings. These articles were printed in the Employment Service Review
and subsequently reprinted under a separate form at which is attached to this
testimony.
I appreciate the opportunity given me to place the results of this pioneering
effort into the official record.
JOB V A C A N C Y S U R V E Y S
Job
V acan cy
D a ta
fo r
an
A c tiv e
M anpow er
P o licy 1
(B y Louis Levine, Director, U.S. Em ploym ent Service)
The current and lively interest in job vacancy information in the United
States is closely related to the emergence of an active manpower policy and to
the relatively recent passage of manpower legislation implementing this policy.
Although there are obviously many interpretations of the meaning of an active
manpower policy, one definition in its broadest sense is ‘ ‘those actions and pro
grams wrhich will produce the kind of labor force which can and will respond
to the inevitable challenge presented by man’s accelerating accumulation of
knowledge,” 2
Achieving this goal calls for manpower policies and operations which, among
other devices, are ultimately designed to deal with employers and workers and
their needs in a specific market place. Job vacancy information is part of a
comprehensive occupational program which will provide detailed information
on the demand side of the job equation for im proving the mechanism of the
employm ent process.
The emergence of interest in job vacancy inform ation was the natural con
com itant of manpower programs whose purpose was to identify job opportunities
and to prepare workers for them.
Am ong the more im portant claimants for job vacancy inform ation is our public
Em ploym ent Service system. The availability of such information woulid con
tribute to its effective operation. Job vacancy inform ation could im prove the
mechanism of the job market by facilitating the matching of workers and jobs.
Frictions in the job market result in unem ployment even when job opportunities
are available. There were 3.5 million unem ployed workers in December 1964,
and 1.5 million of these were drawing unem ployment insurance benefits. Reduc
1 Reprinted from Employment Service Review, April 1965 issue, publication of the U.S. Employment
Service, Bureau of Employment Security, U.S. Department of Labor.
2 Prepared statement of John F. Henning, Under Secretary of Labor, hearings before the Subcommittee
on Employment and Manpower of the Committee on Labor and Public Welfare, U.S. Senate, 88th Cong.,
1st Sess., pt. 2.
63 - 947— 66 ------------ 6
78
JO B VACANCY STATISTICS
ing the duration of unem ployment by just 1 week for only 10 percent of these
insured unem ployed would, among other things, save more than $5.6 million in
unem ployment benefit payments— benefits financed by taxes on employer
payrolls.
By providing a better insight into the nature of available job opportunities, the
job vacancy program may assist the Em ploym ent Service in designing training
programs for the retraining of workers with obsolescent skills or those who have
been adversely affected by automation and technological change. There are
many unfilled jobs at the semiskilled and unskilled levels which should offer
em ploym ent opportunities for the disadvantaged “ poverty” groups in our society.
Such data would also be useful in counseling workers and students and directing
them to occupational choices which will provide better prospects for employment.
The value of this program in carrying out the purposes of the Area Redevelopm ent
Act, the M anpower Developm ent and Training Act, the Vocational Education
A ct of 1963, and the Econom ic Opportunity Act is self-evident. In fact, the
collection and use of these data can be of great assistance to the Em ploym ent
Service in implementing many of its functions under an active manpower policy.
Existence of a job vacancy program would further help to im prove the job
market mechanism by stimulating employers to plan and take an organized
approach to their manpower needs, just as they do in the rationalization of their
production. Asking them to look at their vacancies would cause them to think
about recruitment, form al on-job training, intraplant transfers, and a prom o tionfrom -w ithin policy. Exclusive concentration on the problems of the unem ployed
in the work force may result in a tendency to ignore the 95 percent who are
em ployed and to many of w hom advancement within the establishment is a
param ount need. M oreover, upgrading the employed would open more entrylevel jobs to be filled by new entrants or less qualified workers. One of the ob
stacles to absorption of new entrants into the labor force has been a sharp dwin
dling in the entry-level jobs. Thus, an expanded on-the-job training program by
employers to upgrade the skills of their workers would serve the dual purpose of
filling some of the hard-to-fill, more skilled jobs, and at the same time making
available more entry-level jobs.
Apart from other considerations, the placement process is a key element in
E m ploym ent Service operations. In its matching of workers and jobs, the E m
ploym ent Service must take account of job vacancies. Its social and legislative
base for existence is an unqualified mandate to function as a labor exchange. T o
deny it access to inform ation on job opportunities at the same time that it is
charged with responsibility for registering and seeking jobs for the unem ployed is
to seriously impede its operating effectiveness and to add considerably to the
limitations influencing an already imperfect employm ent market.
A job vacancy information program can and should result in better services to
employers. If the Em ploym ent Service has locally qualified applicants for
existing vacancies, particularly when the job is hard to fill, the em ployer should
be so inform ed and given the option of asking for a referral. An em ployer and
local office staff can sit down together to discuss the factors affecting his com
petitive position in the job market, a practice that might well lead employers to
im prove their personnel and hiring practices. In addition, the vacancy program
can increase the possibility of recruitment of hard-to-find workers through an
interarea exchange of lists of available candidates or, conversely, can encourage
the unem ployed to seek jobs in areas where opportunities exist.
Since 1963, interest in a job vacancy inform ation program in this country has
grown rapidly, partly as a result of the report of the President’s Com m ittee to
Appraise Em ploym ent and Unemployment Statistics. Interest has increased
also because of the growing emphasis on the Em ploym ent Service as a man
power service agency and its responsibilities for the development of occupational
training programs geared to labor demand under manpower legislation passed in
the first half of the 1960’s.
An experimental program of job vacancy surveys is now being conducted by the
United States Em ploym ent Service and its affiliated State agencies in cooperation
with the Bureau of Labor Statistics. The surveys are designed to provide de
tailed data on a current basis, by area, industry, and occupation on unfilled
openings in individual establishments, and the kinds of skills needed to fill these
jobs. M any of the country’s largest em ploym ent centers are included in the
group of pilot areas. A number of small and medium-sized areas are also in
cluded to assure that the survey is representative in terms of geographical dis
persion, industrial characteristics, and the nature of employm ent conditions.
Altogether, the 16 areas account for about one-fourth of the nonfarm em ploym ent
in the country.
JOB VACANCY STATISTICS
P R E L IM IN A R Y
79
F IN D IN G S
As of the close o f January 1965, relatively complete results of areawide jo b
vacancy studies had been compiled and subm itted by four areas— Birmingham,
Milwaukee, Portland (Oreg.), and Providence. These areas had initiated their
experimental surveys in October 1964. The completed studies are historically
significant as the first successful attempts to collect such inform ation for detailed
occupations in large metropolitan areas, permitting the inflation of results to a
universal basis.
Seven other areas initiated their collection programs during N ovem ber 1964.
While the final results are not yet available, inform ation has been subm itted by
these areas on the employer response rate and other lim ited types of data.
Em ployer response exceeded expectations. Of the first 10 areas conducting
their surveys in October and November, 8 areas reported an em ployer response
of approximately 80 percent or more. Three of these indicated response rates
of 95 percent or better! In the tw o areas where the response rate was 63 and 70
percent, there were special circumstances which accounted for the lower participa
tion.
Approxim ately one in four establishments reported one or more vacancies.
Specifically, in 6 out of 10 areas, the proportion of the establishments with
vacancies ranged between 20 and 30 percent. Am ong the other four, the propor
tion ranged from lows of 12 and 19 percent to a high of 34 and 35 percent.
The inform ation on jo b vacancies is useful for identifying shortage occupations
and establishing training courses. A significant number of the vacancies have
been unfilled for at least a month, providing clues to A R A and M D T A training
and vocational education opportunities. Am ong those which were general to
the first four cities surveyed are registered nurses, draftsmen, laboratory tech
nicians, physical therapists and licensed practical nurses, stenographers and
typists, machinists, tool and die makers, machine shop and related occupations,
and mechanics and repairmen. There are also many local hard-to-fill vacancies,
depending on the nature of the local econom y.
There also appear to be occupational opportunities for the disadvantaged for
which short-term training courses may be indicated. This observation is particu
larly timely as administrators begin to appraise and im plement the Econom ic
Opportunity A ct and indeed all earlier legislation in the training field. While
such hard-to-fill jo b opportunities were not too numerous, there were nevertheless
a significant number of semiskilled and unskilled occupations. Am ong these, for
example, were: Chainmen and rodm en; lumbermen; foundrym en; and a series of
occupations in local industries, including textiles, confections, plastics products,
metalworking, and watches and jewelry. These findings have not been evaluated
as to the part played by wages, working conditions, and seasonality in recruiting
workers for these hard-to-fill listings.
A paramount objective of the experimental design is to gain insight into the
extent of structural unemployment as reflected in the matching, or mismatching,
of available skills and available jobs. The vacancies, of course, represent the
demand side of the equation. The active file of registrants in the local em ploy
ment office represents the supply side. Results of the comparison clearly illustrate
the imperfect balance between needed and available skills. On a broad occupa
tional basis for the first four areas, the most conspicuous imbalances occurred in
the skilled and semiskilled groups. Thus, for example, the relative number of
skilled vacancies was twice as large as the proportion of skilled among the unem
ployed; moreover, the number of such vacancies was actually larger than the num
ber of workers registered and seeking work. As might be expected for the occupa
tional array, there was an inverse relationship between level of skill and extent of
supply. These data on a broad occupational grouping are only suggestive of the
nature of the problem. Operational usefulness is predicated on a matching of
specific skills, and the data reveal that comparisons of detailed occupations show
even greater imbalances of supply and demand.
The experimental jo b vacancy program should provide the Em ploym ent Service
with the means for exploring its penetration of the jo b market as reflected in the
relationship between existing job vacancies and job openings obtained by local
offices. This aspect of the surveys is discussed in greater detail elsewhere in this
issue. In general, however, the tentative conclusion appears warranted that the
jo b openings which the Em ploym ent Service receives approximates a cross section
of the vacancies in the area, with the possible exception of skilled occupations.
A bout half the total vacancies reported in the four areas were considered hard
to fill, that is, they were still vacant following a m onth’ s recruitment effort by
80
JOB VACANCY STATISTICS
employers. It was not unexpected that the proportion of liard-to-fdi jobs was
higher among professional and managerial and skilled wTorkers. Somewhat sur
prising, however, wTas the difficulty in filling a relatively large number of jobs in the
other occupational groups. Concerning the latter, the results are still unevaluated
with respect to the part played by wages and working conditions.
C O N C LU SIO N
In view of the encouraging results of this work and the widespread interest and
need, the Em ploym ent Service, with the technical collaboration of the Bureau of
Labor Statistics, has proposed the initiation of a job vacancy survey program
covering a total of 150 Standard M etropolitan Statistical Areas during 1965-66.
These areas, representing two-thirds of the country’ s nonfarm employm ent,
would be surveyed quarterly.
Advancem ent and extension of the job vacancy surveys represent a much needed
next step in the analysis and diagnosis of job market behavior and the adminis
tration of job market forces designed to achieve maximum developm ent and
utilization of manpower resources. In this connection, vacancy surveys are
a natural outgrowth of the increasing awareness and acceptance of the need
for an active manpower policy in the United States. The large volume of newly
enacted manpower legislation since 1961 in the fields of education and training,
concerned with the em ploym ent and unem ployment of manpower resources and
occupational rehabilitation, makes a more comprehensive knowledge of jo b
vacancies on a detailed occupational and geographic basis an absolute necessity.
M uch of the manpower emphasis in recent years has been devoted to the supply
side of the jo b market problems. It now becomes increasingly im portant to
direct attention to the demand side. In the final analysis employability must lead
to employm ent. The real test of the effectiveness of recent manpower legislation
will be in terms of jobs. One of the outstanding results of the pilot jo b vacancy
surveys has been to reveal the imbalances between the kinds of jobs which remain
unfilled and the skills of the unemployed.
As greater emphasis is given to providing manpower assistance to the econom
ically disadvantaged and educationally and culturally deprived sectors of the
population and labor force, the importance of job vacancy inform ation increases.
For numerous operations, it is needed not only by occupation and locality but
also by specific establishment. Over a period of time, such inform ation can
contribute greatly to the job market operation research of the public Em ploy
ment Service, to the training and competence of Em ployer Service representatives,
and to the im provement of manpower services to employers.
E m p lo y m e n t S e r v ic e
O p e r a tin g
D a ta
as
a
M easu re
of
Job
V a c a n c ie s 1
(By Vladimir D . Chavrid and Harold Kuptzin)
In calendar year 1964, the public Em ploym ent Service received som e 8.2
million nonfarm job openings from over 1 million employers— a total which,
on the basis of the best available estimates, represents nearly one-fourth of all
new hires for the country as a whole. About 6.3 million of these openings—
equivalent to nearly 16 percent of the new hires— wTere filled by workers referred
and placed by the Em ploym ent Service. Because use of the public E m ploy
ment Service in this country is optional for employers, some question has been
raised as to whether the openings listed at public employm ent offices are repre
sentative of all job openings in terms of their occupational and other character
istics. The new7 Departm ent of Labor experimental job vacancy program throws
some new light on this subject.
Under this experimental program , the term “ job vacan cy” is defined as:
. current job openings in an establishment which are immediately available
for occupancy for workers outside the firm and for which the firm is actively
seeking workers.” The job vacancy figures to be developed as part of this pro
gram are intended to represent, with a few exclusions, the sum total of all current
job opportunities. The more important of these exclusions are vacancies for
which there is currently no active recruitment (jobs held for employees who will
be recalled, jobs to be filled by transfers or promotions, jobs for which new workers
have already been hired but have not yet reported, jobs vacant because of labor
i Reprinted from Employment Service Review, April 1965 issue, publication of the U.S. Emolovment
Service, Bureau of Employment Security, U.S. Department of Labor.
JOB VACANCY
81
STATISTICS
disputes), and vacancies which are not available for immediate occupancy (jobs
for which recruitment is now in progress— such as teaching positions— but where
personnel hired are not expected to report for work until the beginning of the
fall semester).
D IM E N S IO N S
AND
O B J E C T IV E S
OF T H E
PROGRAM
The follow ing points should serve to recall and highlight some of the more
im portant aspects of the experimental program.
1. The experimental job vacancy information surveys are being conducted
twice during this fiscal year in the 16 pilot areas selected. These 16 areas include
most of the country’s largest employm ent centers, and a group of smaller and
medium-size areas to assure representation in terms of geographical dispersion
and the nature of em ploym ent conditions.
2. The surveys are being conducted by State Em ploym ent Services affiliated
with the U.S. Em ploym ent Service of the Bureau of Em ploym ent Security in
cooperation with Bureau of Labor Statistics of the Departm ent of Labor.
3. D ata are being collected in considerable occupational detail by requesting
employers to list the plant titles of their vacant jobs. These plant titles will be
translated to standardized occupational classifications, as listed in the Dictionary
of Occupational Titles, by Em ploym ent Service occupational analysts or other
qualified staff.
4. The primary focus of the experimental program is to help reduce unem ploy
ment by using job vacancy information as a basis for a more effective matching
of men and jobs. In most areas, employers will be offered assistance by the local
office in filling their job vacancies if suitable applicants are available, and if the
employer wishes to give a specific job order to the Em ploym ent Service.
5. The data are also being used to supplement existing inform ation on training
opportunities and for counseling and guidance. The possibility of using such
information as an econom ic indicator will also be explored.
6. The information collected from employers is being compared with existing
Em ploym ent Service operating data on the skills of the unem ployed and the
nature of available job openings. These comparisons should provide additional
clues regarding the degree of skill maladjustments in the area.
7. Inform ation on jo b vacancies is being collected on the basis of a probability
sample, and the data will be inflated, by occupation, to represent area wide totals.
The sampling frame is designed to insure that employers representing more than
50 percent of total em ploym ent— including all large establishments— are covered
by the survey, and that it provides for representation in all m ajor local industry
groups.
8. Inform ation is being collected in most areas on duration of the job vacancies,
and these data also will be inflated to areawide totals, by occupation. Separate
information is being compiled for vacancies existing 1 week or more (for com
parability with national unem ployment estimates) and for those existing 1 month
or more (as an indicator of hard-to-fill vacancies).
9. As part of the experimental aspects of the program, inform ation is being
collected in some areas on wage rates offered for job vacancies, and in others on
vacancies for part-time and tem porary employment. In 3 of the 16 areas, infor
mation is being collected through the State central offices of the Em ploym ent
Service, rather than through the local offices. In these areas, the Em ploym ent
Service will not attem pt to develop job orders on the basis of the vacancy reports,
unless the employer initiates such a request.
10. Because of the experimental nature of the program, the timing of the
initial surveys conducted by the States was staggered from O ctober 1964 through
January 1965. Four area surveys were conducted in October, seven in N ovem ber,
four in December, and one in January.
11. The second survey for all 16 areas is scheduled for April 1965.
In terms of basic objectives, the experimental program was designed to provide
answers to some of the following questions:
1. Is it feasible to collect inform ation on job vacancies from employers, by
occupation, on a regular basis?
2. Can such information be translated into a standardized occupational system
which would be comparable among employers and areas, and comparable with
other information on local labor demand and supply?
3. What, if any, structural imbalances in labor demand and supply does the
job vacancy inform ation reveal?
4. Can employers provide valid information on the duration of vacancies, as
well as on wage rates, part- and full-time vacancies, and openings for tem porary
82
JOB VACANCY STATISTICS
workers? Will the request for such information limit the response with respect
to total vacancies?
5. Will contacts by the E m ploym ent Service to develop job orders for vacant
jobs, if qualified applicants are available, serve to limit employer response?
6. Can the job vacancy inform ation be used to im prove the placem ent and
other manpower operations of the public Em ploym ent Service, in terms of match
ing jobs and workers, in planning training programs, and in counseling and guid
ance activities?
Preliminary answers to some of these questions are now becom ing available
from initial reports of the State Em ploym ent Services covering the surveys con
ducted in October and November.
N ATU RE
OF E M P LO Y E R RE SPO N SE
When this article was being written (toward the end of January), summary
reports from the State Em ploym ent Services were available only for the first four
areas surveyed under the experimental program— Birmingham, Milwaukee,
Providence, and Portland (Oreg.)— and partial data for six of the seven areas
surveyed in Novem ber (Baltimore, Charleston (S.C.), Miami, Minneapolis-St.
Paul, New York, and Richm ond). Thus, many of the present conclusions must
be regarded as tentative and subject to revision when additional inform ation
becomes available and is analyzed in detail.
One conclusion that appears likely to be sustained, however— on the basis o f
the inform ation already available— relates to the feasibility of collecting job
vacancy information by specific occupation from employers. The experience in
the first 10 reporting areas appears to demonstrate that the collection of such
inform ation from employers is indeed feasible— at least on a one-time basis.
Em ployer response exceeded 60 percent in each of the 10 areas, reaching 80 percent
in 6 areas and 90 percent or more in 3 areas (Milwaukee, Portland, and R ich m on d)..
An examination of the schedules returned, and a response analysis survey of a
subsample of the establishments contacted, conducted by the BLS in several o f
the pilot areas, indicated that the employers generally understood the instructions,
definitions, and forms, that they were willing to report the data requested, and
that they reported their vacancies in an accurate manner. Whether this high
rate of response can be sustained if the surveys are conducted on a periodic basis
is, of course, not yet certain. The second round of surveys, to be conducted in
April, should provide a further indication of this.
Em ployer response did n ot appear to be significantly affected in the first 10
areas by the possibility of additional contacts by local employm ent offices to
explore whether the employers wished Em ploym ent Service assistance in locating
workers to fill their job vacancies. As a matter of fact, the four areas reporting
the highest response rates— Birmingham, Milwaukee, Portland, and Richm ond— ■
are areas where such job order development efforts were part o f the experimental
program.
The proportion of firms reporting job vacancies of any kind during the survey
month ranged from about 12 percent in Charleston, S.C., to 35 percent in M il
waukee. In most of the remaining areas, about one-fifth to slightly more than
one-fourth of aill firms surveyed listed job vacancies.
A large proportion of the vacancies reported were of relatively long duration.
In 9 of the 10 areas (data not available for New Y ork City), more than two-thirds
of the vacancies had been in existence at least 1 week, and about half in all 10
areas had been open for a m onth or more. Five of the 10 areas (Baltimore,
Charleston, Minneapolis-St. Paul, Providence, and Richmond) reported that more
than half of the job vacancies listed were in the relatively hard-to-fill category
(in existence for 1 month or longer).
Surveyed employers in all 10 areas reported job vacancies in a broad of occu
pational classifications, including some in lower skill categories. The number o f
separate 3-digit occupational classifications for which vacancies were listed ranged
from 79 in Charleston, S.C., to 214 in New Y ork City. The occupational dis
tribution of the vacancies reported by employers is shown in the table on page 7.
The pilot areas in the experimental program were also requested to summarize
the information on job vacancies, on an inflated area wide basis, in relation to their
local office unfilled job openings as of the end of the month, and also in relation
to the job applicants registered at the local offices. Such data were available for
three areas— Birmingham, Milwaukee, and Portland— as of this writing.
Job openings listed with the Em ploym ent Service represented around 29 per
cent of the total number of job vacancies in these three areas. The proportion in
JOB VACANCY STATISTICS
83
professional and managerial occupations was over 41 percent, however, and it
exceeded 30 percent in the clerical and sales, service, and semiskilled groups.
Em ploym ent Service job openings represented about 20 percent of the total
number of unskilled job vacancies in these areas, and about 16 percent o f the
skilled.
Turning from the demand to the supply aspects of these data, unem ploym ent in
Birmingham, Milwaukee, and Portland (Oreg.) was relatively m oderate at the
time the survey was taken. Approximately 90 percent of the unem ployed were
registered at the local offices in these areas at the end of October. The number of
applicants classified as skilled was slightly below the vacancy total (inflated to
represent area totals), however, and professional and managerial applicants
exceeded the number of vacancies in that category by a relatively small margin.
In contrast, the number of applicants was about 2}i times or more the number of
vacancies in the clerical and sales, service, semiskilled, and unskilled categories.
As another example, the inflated job vacancy data showed that jo b vacancies
for skilled welders in M ilwaukee in October were about 1J4 times the number of
job applicants with this type of occupational background (187 to 112). A t the
same time, Portland had only 12 vacancies in this occupation and 100 applicants,
and Birmingham 89 vacancies and 53 applicants. Y et if the number of vacancies
and applicants for welders were shown on an aggregate basis for all three areas,
the data would indicate that demand and supply for this occupation were nearly
in balance (288 vacancies compared with 265 applicants).
Subject to congressional approval, the Departm ent is now planning to introduce
a job vacancy information program in fiscal year 1966 on a quarterly basis in the
150 m etropolitan areas for which regular labor area surveys of em ploym ent and
unem ployment are prepared by the USES and the affiliated State Em ploym ent
Services. Such an expansion would, of course, take full cognizance of the experi
ence accumulated in the operations of the experimental program in terms o f
schedule design, methods and types of information collected, and procedures for
the effective utilization of these data in im proving the operations of the jo b market.
Percent distribution of job vacancies, by major occupational group, 1st 4 surveyed
areas,1 1964
Total
vacancies
Vacancies existing 1 month
or more
Major occupational group
Total
_ _
_ _
Professional and managerial________ _______ __
________
Clerical and sales____________ ___________ __________________
Service2_____
________ ________ ____ ___________
_ _ __
Skilled______________________________________________________
Semiskilled__ ____ __
_ ________ _ _ _ _____ ___ ____
U n sk illed ...____ ______________ ______________ __
___
Percent
distribution
Percent
distribution
100. 0
11.1
22.2
8.6
24.7
21.0
12.4
100.0
12.4
14.2
6.2
29.1
26.4
11.7
Percent
of total
59. 3
66.3
38.0
43.2
69.9
74.8
55.1
1 Job vacancies reported by employers, inflated to represent areawide data, for 4 areas surveyed in
October—Birmingham, Milwaukee, Portland (Oreg.), and Providence.
2 Excludes vacancies for domestics, but includes a small number of vacancies in agriculture, forestry, and
fisheries.
Even when in full operation, however, the jo b vacancy program will provide
only a single dimension of a sound program for the compilation, analysis, and
utilization of inform ation on occupational developments and outlook. The jo b
vacancy program is designed to yield a detailed snapshot, as of a single point in
time, of occupational skill requirements in a local community. A series of such
snapshots, accumulated over time, will make available essential inform ation on
the changing characteristics of occupational needs. However, the jo b vacancy
data by themselves do not represent a complete fulfillment of the requirements for
the kinds of occupational job market information necessary for a proper appraisal
of manpower requirements and resources in local labor areas.
Three additional elements seem to be needed for a comprehensive occupational
job market information program for local areas. These include:
1.
A relatively current benchmark or inventory of em ploym ent by occupation
in a community— to be updated periodically, perhaps once every 2 years. While
some information on employm ent by occupation is available from the decennial
84
JOB VACANCY STATISTICS
census, these data soon becom e obsolete, and the occupational detail provided is
not precise enough for use in local manpower operations.
2. A detailed analysis of relatively long-range (2 or 5 years) occupational re
quirements in relation to probable resources. The job vacancy inform ation—
which provides data on current occupational needs— can be used as a basis for
updating this forecast in terms of changing developments as they occur.
3. A broader program for the translation of these materials into appropriate
counseling and guidance tools, and for the utilization of the inform ation in planning
com m unity development and training programs, im proving the functioning of
the jo b market, and facilitating interarea recruitment and m obility of workers.
Plans are being developed for the implementation of such a broader occupational
inform ation program. Some of the basic tools and approaches needed for this
program are already in existence— developed through a number of years of oper
ating experience in conducting comprehensive area skill surveys, the preparation
of local occupational profiles, occupational guides, and related Em ploym ent
Service programs, and in implementing the provisions of recent manpower
training legislation.
In d ia n a Job V a c a n c y S u r v e y s 1
(B y Em ily Hawk and Evelyn T oliver)2
Indiana conducted seven area job vacancy surveys in September and O ctober
1964, preceding the pilot surveys conducted by State Em ploym ent Services in
areas in various parts of the country. Indiana’ s purposes and approach differed
somewhat from those in the Departm ent of Labor’s pilot program. Indiana was
particularly interested in conducting job vacancy surveys and in beginning
them early, partly for use in connection with the re-em ploym ent program in
South Bend, and in order to have current occupational inform ation when its
General Assembly convened in early January 1965. Vocational training was
expected to be an im portant issue in the Assembly, and the inform ation obtained
from job vacancy surveys would be useful in documenting the overall need for
training. Also, job information, by area, was needed for the use of vocational
education departments in the public schools and new area vocational schools.
As was true in the D epartm ent’s pilot program, we also were interested in
determining whether em ploym ent office orders were representative of the occupa
tional demand situation in general. Thousands of unem ployed people were
registered for work in local offices while thousands of unfilled job orders were on
file in em ploym ent offices and an undetermined number of other job openings
were known to exist. How’ numerous and how different were these job openings
which were not listed from those on file with employm ent offices? Were there
vacancies in occupations for which local offices had qualified applicants? D id
employers and local offices agree about the kinds of jobs that were hard to fill?
We wanted to know, too, how employers would react to attempts to collect job
vacancy inform ation and how successful we would be in developing and filling job
orders obtained from employers reporting job vacancies.
We elected to sample employing establishments in our seven largest areas,
contacting all firms employing iOO or more and a sample of smaller ones in all
industries. The samples were drawn from the latest available (M arch 1964)
tabulation of firms covered by the Indiana Em ploym ent Security Act, and the
largest noncovered firms, consisting of hospitals, government agencies, and non
profit organizations, were added. The samples were not inflated to area totals.
Concurrently with the employer surveys, local offices reviewed their job orders.
At the time of the survey, employm ent in the seven areas was at or near the
peak for the year. In each area, except South Bend, nonfarm em ploym ent was
above the corresponding period in 1963 and the unemployment rate was below the
year-ago rate. The State and area unemployment rates, except for South Bend,
were all below the national rate.
In most areas, vacancies per 100 employees in responding firms showed an
inverse relationship to unem ployment rates. Differences in response rates among
areas and the relative stability of em ploym ent were factors which caused some
distortion in this relationship.
Em ployer response to requests for job vacancy information was very good in
terms of percentage of firms returning the survey forms, but considerable effort
1 Reprinted from Employment Service Review, April 1965 issue, publication of the U.S. Employment
Service, Bureau of Employment Security, U.S. Department of Labor.
2 Emily Hawk and Evelyn Toliver are analysts in the Research and Statistics Section, Indiana Employ
ment Security Division.
JOB VACANCY STATISTICS
85
was expended to get replies and there was reason to question that all vacancies
were reported. About one-fourth of the firms responding listed vacancies.
Of the 2,737 firms contacted, 96 percent responded to the requests for informa
tion. However, 4,312 contacts were made, including 769 personal visits and
1,416 telephone calls, to obtain the responses of 2,634 firms. A bout half of the
firms responded to the first request; another 40 percent answered the second
request; and the remaining 6 percent replied to a third contact. In general,
personal visits were made on the initial request only; these were much more
effective than initial contacts by mail, yielding nearly three times as many re
sponses. One office which contacted some firms by mail the first time and fol
lowed up with personal visits obtained information during the visits from 57
percent of the firms. M ost followups were by telephone, and these also were
very effective.
M any of the firms contacted the second time explained that they had not
understood that the form was to be returned regardless of whether job vacancies
existed in their firms, although this w~as stated in the instructions accom panying
the form. Others stated they had not replied because they felt that their vacan
cies could not be filled by the local office, leading us to wonder whether all vacan
cies were listed by the firms that replied to the first request. Em ployers con
tacted the second time also indicated that mail requests may not reach the proper
company official. Survey forms were addressed to the company, usually with
no individual specified. In some cases, the forms had not reached the same
person later contacted by telephone. There was great variation in titles of those
completing the forms; a review of 150 survey forms in one area revealed 35
different titles, ranging from assistant cashier to president. The size of the
firm appeared to be the determining factor as to who furnished the information.
Specific reasons for not participating in the survey were given in only a few
cases. M ost of the nonresponse was among firms in which interviewers were
unable to contact the proper official on foil own p or among firms which stated
they had not received the form, had returned the form, or would return the form
but did not. A few replied that they had no time for such surveys, and one of
the largest firms in the seven areas (branch of one of the largest in the country)
stated that giving such information was “ contrary to its philosophy.” This firm
also expressed concern that the local office would consider the job vacancies as
job orders. Although assured that the office would make no referrals, the firm
did not participate. However, other branch establishments of the same firm
did cooperate.
The best response was from firms with over 100 employees; this group also
received the greatest percentage of initial personal visits; listed most of the total
vacancies; and had a higher percentage of their job orders on file in local offices
than did smaller ones.
The kinds of unfilled jobs reported on job vacancy survey forms corresponded
closely in all seven areas with the kinds of orders on file in local offices. There
was general agreement, also, about the kinds of jobs that are hard to fill. T w o
thirds of the five- or six-digit occupations reported as job vacancies were repre
sented in local office files, although local offices had identical orders from only
one-third of the job vacancy survey employers. The occupations not repre
sented in local office orders were not unusuai occupations; nearly one-half were
in the professional and skilled occupational groups and had been on file in local
offices at one time or another. N o entirely new occupations were reported.
The amount of job development undertaken by local offices was limited by the
fact that 60 percent of the job vacancies reported were described b y employers as
hard to fill, and most of the occupations represented were those carried on local
office unfilled job orders. Interviewers reviewed job vacancy forms as they were
returned and checked the files for qualified applicants. When it seemed possible
that good referrals could be made, the interviewers attem pted to develop job
orders. W ith a few exceptions, employers had no objections to listing job orders
with the em ploym ent offices. However, many of the easier-to-fill vacancies were
already filled. In all, fewer than 14 percent of the new vacancies listed by em
ployers were developed as jo b orders, and only 3 percent o f the new jo b vacancies
had resulted in placements at the close of the study.
Employers reported the greatest number of new vacancies in the clerical and
sales group; this group represented the second highest number of local office orders.
M ost of the job openings developed were in this group and the second highest
number of placements resulted. A bout half of the vacancies reported by em
ployers in this group were considered hard to fill; local offices reported to a slightly
higher percentage as hard to fill. Three percent of the new vacancies in this
group resulted in placements.
86
JOB VACANCY STATISTICS
Professional and related occupations represented the second highest number of
new vacancies reported, but job orders in this group represented one o f the lowest
proportions of total job orders. Ninety percent of these job vacancies and job
orders were considered hard to fill. The smallest number of job openings was
developed in this group and the fewest placements made. Obviously, the demand
for qualified applicants for these kinds of occupations cannot be supplied currently
from any source.
A bout the same situation existed for the skilled occupations; slightly fewer new
job vacancies were reported; more job orders were on file in local offices. C on
siderably more job developm ent was undertaken, but placem ent results were little
better than for the professional group. About 80 percent o f the skilled job vacan
cies were considered hard to fill.
Semiskilled and unskilled occupations wTere considered by employers a,« the
easiest to fill, and it was in these groups that the highest percentages of place
ments were obtained from job development. Although unskilled occuaptions
represented the highest proportion of total job vacancies and job orders, this was
because of situations in tw o areas: The Gary-H am m ond-East Chicago area
figures were affected by intensive recruitment for steel mills, and the Fort W ayne
area by heavy demand for the electrical machinery industry. In most other
areas, fewer job vacancies were listed for the unskilled than for any other group.
Local offices filled slightly fewer than half of the unskilled job openings they
developed.
There were many job vacancies reported in occupations suitable for training
under the M anpower Developm ent and Training Act, but in most cases training
courses in these occupations had already been proposed, were in progress, or had
been com pleted by the end of the survey period. Openings for welders were
reported in five areas, although M D T A training courses have been in operation
since 1963 in the areas. The survey indicated that continuation of training in
this and other courses is necessary. Present apprenticeship training appeared
to be inadequate; many of the vacancies listed were for such occupations as
machinist, tool and die maker, skilled machine operators, and electricians. H ow
ever, this information was not different from that reported on local office job
orders for which openings appeared in these same occupations.
Several conclusions were drawn from the Indiana job vacancy studies:
Job openings on local office orders do represent the general occupational demand
situation.
Because the kinds of job orders corresponded to the kinds of job vacancies,
and because employers and local offices were finding the same types of jobs hard
to fill, job development opportunities were limited.
The number of placements resulting from job development attempts was small
com pared with the effort involved.
It is possible to obtain job vacancy inform ation; however, collection by mail
w ithout follow up will yield a much lower response than collection by personal
visit and follow up calls.
There is a tendency am ong small firms not to report when they have no
vacancies. This may lead to some bias wThen conclusions are drawn.
P o r tla n d
S tu d ie s Job V a c a n c ie s 1
(B y Wesley E. Zellner)2
The proposed study of job vacancies in the Portland, Oreg., area was received
with enthusiasm by the Oregon Em ploym ent Service. The copy of the letter to
the President from the Secretary of Labor, which accom panied the survey in
structions, played an important part in generating interest and enthusiasm.
The proposed study plan provided for surveys in October 1964 and January and
April 1965. This October to April period coincides with the employm ent dip and
recovery phase experienced each year, although in 1964 the area had been char
acterized by a relatively tight jo b market situation. It was anticipated that
inform ation concerning jo b openings over and above that received b y the local
offices in their normal operations would yield valuable insights into the disparities
1 Reprinted from Employment Service Review, April 1965 issue, publication of the U.S. Employment
Service, Bureau of Employment Security, U.S. Department of Labor.
2 Wesley E. Zellner is Supervisor of the Research and Statistics Division, Oregon Department of
Employment.
JOB VACANCY STATISTICS
87
between labor supply and labor requirements which prevail during a slack em
ploym ent season.
Subsequent cancellation of the January survey was dis
appointing, in that some of the hoped-for data on seasonal changes in demand
would not be available, although the April survey is expected to yield useful
information on this aspect of demand.
Preparations for the October portion were initiated during early September.
Preliminary discussions were held with the local office managers to acquaint them
fully with the procedures for obtaining job vacancy information and job develop
ment. In the meantime, central office work (selecting a sample of employers to
be interviewed, printing record cards and questionnaires, etc.) was being com
pleted. These preparations proceeded sm oothly. A news release was issued and
an announcement subm itted to television stations, but there was no widespread
campaign to reach employers, labor organizations, or the general public.
Interviewing and mailing proceeded without serious difficulty. The total
number of employers responding without a followup represented something less
than half of those selected. A week was allowed between the due date for the
schedules and the first followup calls. During this interval, completed schedules
were processed and, whenever applicable, referred to the local offices for job
development. Then the first followups were started. It developed that m ost
of the employers who had not returned the schedule had forgotten about it.
Quite often, the information was then obtained over the phone. Only 15 of
994 employing establishments refused to cooperate in the study, although an
additional 31 persisted in “ forgetting” to forward their completed schedule.
Seven others did not respond because data were not available.
In total, 1,173 job vancancies were reported. Of these, 408 were already
on file in the local offices, and 765 were vacancies that had not previously come
to the attention of the Em ploym ent Service. Because no suitable applicants
were registered with the local office, no effort was made to obtain job orders on
119 of the 765 new vacancies reported. Another 181 could not be filled either
because of union hiring restrictions (47) or the requirement that the applicant
be on a Civil Service list (134). There were, therefore, 465 job vacancies, not
previously known to the Em ploym ent Service, available for possible referral
of job applicants. It was found, however, that slightly more than half of these
openings had been filled or for some other reason were no longer extant by the
time employer contacts were made.
Only a small proportion— 36 of 227— of the newly reported job vacancies
were subsequently listed with the Em ploym ent Service. The im pact of the survey
upon placement was negligible; only 15 placements resulted from the survey.
It appears that timing was an im portant contributory factor to the unsatis
factory placem ent results. The local offices indicated that the time lapse between
the date the opening first became available and the date it was referred to the
Em ploym ent Service interiewer was too great. Part of this delay was caused
by the failure of employers to return the report on the due date, and part was due
to time lost during the initial processing of the com pleted reports. Another
problem stemmed from what some employers interpreted as a breach of the
confidential nature of the report. T o avoid recurrences of this sort, it should
be made clear to employers when they receive the schedule that any job vacancy
information they report will be made available to the local offices.
A bout 35 percent of the job vacancies reported were already on file in the local
offices. M ore often than not, reported vacancies had gone unfilled for a month
or more. This was particularly true of the shortage occupations. Long-term job
vacancies were apt to occur in smaller firms. Studies of hiring channels have
dem onstrated that these employers do not generally resort to methods capable
of reaching large numbers of potential applicants when they are trying to fill a
jo b opening.
Between October 15 and Novem ber 15, unem ployment in the Portland m etro
politan area showed a seasonal rise of about 4,000, an increase of over 30 percent.
The study did suggest that much of this increase could have been avoided or
postponed if there were some way to channel qualified persons into the jo b vacan
cies which existed in the area. It is not possible, however, to quantify the effect
that filling these vacancies would have or to determine how the skill com position
of the unem ployed group would be altered.
An im portant short-run solution would appear to lie in im proving and speeding
up the process by which qualified applicants are directed toward existing job
vacancies.
88
JOB VACANCY
M ilw a u k e e
E m p lo y e r s
(B y H.
STATISTICS
C o op erate
in J o b
Survey 1
J. J a c k s o n ) 2
Milwaukee was one of the first areas to conduct a job vacancy survey. Em
ployer cooperation was excellent; 95 percent of the firms surveyed returned ques
tionnaires, all of them usable.
Staff of the Wisconsin State Em ploym ent Service decided to make their initial
approach to employers on an individual basis, rather than through an employer
association, and to stress the practical uses of the information, especially its use as
an indicator of area training needs. About a week before the survey began, letters
were sent to firms in the survey sample, asking for cooperation and emphasizing the
need for local training information. Four days before the survey began, there was
a meeting of the special antipoverty committee of the Social D evelopm ent Com
mission of Greater Milwaukee. A t this meeting, the director of theM ilwaukee
local office announced the jo b vacancy survey and pointed out its relevance to the
developm ent of youth training programs which were of particular interest to the
group. There was extensive local newspaper coverage of the meeting and the
planned survey.
Additional strength was given to the project by the assignment of E m ployer
Relations Representatives to introduce the program to employers with whom they
had had previous contact. Undoubtedly, this contributed to the high response
rate.
Milwaukee and Waukesha local office staff, assisted by administrative office
research analysts, visited 387 firms. Questionnaires were mailed to another 409
employers, and 269 followup telephone calls were made. The excellent response
from employers indicates that it will probably be feasible to conduct the next
Milwaukee job vacancy study by mail. This will result in considerable savings in
cost and will permit a greater emphasis on job development efforts.
Members of the survey crew were instructed to undertake job developm ent
efforts only if they would not jeopardize future employer cooperation. A total
of 285 job openings and 78 placements resulted directly from the survey. This
com paratively unimpressive record is in part attributable to the fact that in this
first round of surveys, establishment of the program was the primary concern,
and job development was not stressed as it will be in future surveys.
The long-range effect of the job vacancy program upon placement could be
substantial. Job vacancies were reported in almost 200 different 3-digit occupa
tions. In more than two-fifths of these occupations, there were more estimate
vacancies than there were applications in the Em ployment Service active files.
The location of the estimated vacancies has important implications for future
Em ploym ent Service operations. A time series of this inform ation can tell us
what industries to visit or telephone for orders, when to contact them, and what
kinds of orders to expect. For example, 57 percent of the estimated vacancies
in the Milwaukee SM SA were in nonmanufacturing and 60 percent were in
firms with fewer than 144 employees.
Occupations in which vacancies are reported, when compared with jo b orders
received from employers, can help us evaluate employer acceptance of our services.
For example, openings on file were low compared with vacancies in skilled trades
but high relative to vacancies in professional occupations.
When evaluated in connection with the applicant supply, reported vacancies
can tell us what applicants we need to seek and what training to encourage.
Our applicants fell far short of the estimated number of job vacancies in profes
sional, semi-professional, managerial, clerical and sales, and skilled occupations.
Further evidence of shortage occupations is given by vacancies reported as
open for a month or more. There may be some bias in this figure since employers
may normally take longer to fill certain types of jobs than others. Yet, there are
obviously shortages wiien 67 percent of the vacancies for engineers and scientists,
85 percent of the vacancies for nurses, and over half the vacancies for skilled
machine shop occupations had existed for a month or more.
In total, the M ilwaukee job vacancy study was a success. The Milwaukee
and Waukesha offices sent each employer a letter thanking him for his cooperation
and preparing for the next survey. The Em ploym ent Service usually cannot
create jobs, but it can reduce unemployment through filling available jobs faster
and through providing the information needed to train people to match jobs.
Job vacancy studies can, in many ways, help to do both.
1 Reprinted from Employment Service Review, April 1965 issue, publication of the U.S. Employment
Service, Bureau of Employment Security, U.S. Department of Labor.
2II. J. Jackson is Supervisor, Manpower Analysis, Wisconsin State Employment Service.
JOB VACANCY STATISTICS
P rogress
R e p o r t— Job
89
V a c a n c y S tu d ie s 1
(By Norman M edvin and James H iggins)2
Growing attention has been directed in recent years toward the development of
inform ation on job vacancies. For many years, a great deal of inform ation has
been available on the N ation7s em ployed and unemployed. Lacking, however, are
precise data showing wrhere the jobs are and the kinds of skills needed to fill them,
which could be used to facilitate the placement of workers in these jobs.
Lack of inform ation on jo b vacancies has given rise to a number of provocative
questions among manpower analysts and economists seeking to unravel the
factors causing dislocations in the jo b market. The divergent views evidenced
by the “ expansionists” and the “ structuralists” point up some of the difficulties
that arise in dealing with econom ic questions when insufficient knowledge is
available. The two schools differ primarily in the emphasis and importance
assigned to the various factors. Proponents of both agree, however, that infor
mation on job vacancies on both an occupational and area basis, along with simi
larly detailed data on the characteristics of the unemployed is essential for a precise
definition of both employm ent and unemployment problems. Analysis of these
data wTould provide the basis needed to develop im proved programs to com bat
the problems.
The need for a count of jo b vacancies by area and by occupation and an ac
tive program of placing workers in these jobs wras also pointed out by the
President’s Com mittee to Appraise Em ployment and Unemployment Statistics
(the “ Gordon Com m ittee” ). In the words of the Com mittee report: “ It is
doubtful that any suggestion for the improvement of knowledge about the
N ation’s job markets was more frequently voiced to this Com mittee than that
calling for jo b vacancy statistics.” While there are undoubtedly many technical
difficulties in obtaining valid jo b vacancy data, the Com mittee noted that the
need for such information is becoming continually more acute and recom mended
that it be obtained for analytical use and for operational and administrative
purposes.
The need for job vacancy data for effective Em ploym ent Service operations
has long been recognized by the U.S. Em ploym ent Service and its affiliated
State agencies. T o meet its prime responsibility for the placem ent of jobseekers
and the recruitment of qualified applicants to meet employer staffing require
ments, the solicitation of job orders from employers has always been an essential
activity of the public Em ploym ent Service. However, it was not certain that
job orders obtained by the local office were representative of total job vacancies
in the community. For this reason, as long ago as W orld War II, and later
during the Korean conflict, pilot programs were initiated for the collection of
job vacancy data by occupation and industry in major metropolitan centers in
anticipation of possible areawide manpower shortages. W hile these programs
were discontinued after the end of hostilities, job vacancy data have been col
lected since that time on a voluntary basis by some 15 State em ploym ent se
curity agencies on a statewide or area basis, but in all instances except one, without
any occupational breakdown.
Research in the area of jo b vacancy information has also been conducted
by the Bureau of Labor Statistics. A survey made in 1956 to determine whether
individual establishments keep job vacancy records indicated that employers
do not keep records of vacancies which are comparable in accuracy or detail
to their payroll records, and that the job vacancy data they could furnish (at
that time) would not be equal in quality to that obtained in the em ploym ent
statistics program. Additional research by the Bureau o f Labor Statistics was
conducted in early 1964 on the job vacancies reporting programs in foreign
countries.
Beginning in 1963, interest in a job vacancy inform ation program in this
country grew rapidly, partly as a result of the increased emphasis on the E m ploy
ment Service as a manpower service agency, and its responsibilities for the de
velopm ent of occupational training programs geared to job market demand
under such legislation as A R A , M D T A , VEA, and the EO A. It also reflected
the continuing pressures arising out of the report o f the President's Com mittee
to Appraise Em ploym ent and Unemploym ent Statistics, the recommendations
1 Reprinted from Employment Service Review, March 1966 issue, publication of U.S. Employment
Service Bureau of Employment Security, U.S. Department of Labor.
2 Norman Medvin is Chief and James Higgins is a Labor Economist in the Branch of Skill and Industry
Surveys Office of Manpower Analysis and Utilization, U.S. Employment Service.
90
JOB VACANCY
STATISTICS
of the Senate Subcomm ittee on Em ploym ent and Manpower, and the active
support of the Governors of California, Illinois, and New York, among others.
In response to the widespread demand, as well as to meet its own needs, the
USES initiated a program of pilot jo b vacancy studies, in cooperation with
State em ploym ent security agencies, to determine whether vacancy information,
by occupation, could be collected on a labor area basis. The first pilot study
designed to test the feasibility of collecting job vacancy data by detailed occu
pational classification, was conducted in the fall of 1963 by the Illinois Bureau
of Em ploym ent Security at the request of the G overnor’ s Com m ittee on Un
employm ent and in cooperation with USES staff.
This feasibility study covered 62 firms in the Cook County sector of the Chicago
m etropolitan area. The m ajor findings of this study indicated that the collection
of jo b vacan cy data by occupation was feasible and that about four-fifths of the
em ployers could provide this inform ation on a continuing basis. In addition, a
large proportion of the employers reported that they could identify hard-to-fill
vacancies.
While the Illinois study tested the collection of jo b vacancy data b y occupation
from a relatively small sample of employers, a second survey in Buffalo (conducted
b y the affiliated New Y ork Division of Em ploym ent in cooperation with the
Buffalo area Cham ber of Commerce) studied the possibility of compiling such
inform ation on a mass basis for a m ajor industrial area. A Cham ber of Com m erce
mailing list was used as the basis for the Buffalo survey and covered 2,687 firms
em ploying a total of over 150,000 workers.
An experimental program of jo b vacancy surveys on a more am bitious scale,
taking into account the knowledge gained from the Chicago and Buffalo experi
ments, was initiated in the fall of 1964 by the Bureau of Em ploym ent Security
and the Bureau of Labor Statistics in the U.S. D epartm ent of Labor. This pro
gram was form ally announced by President Johnson during his press conference
of August 8, 1964. Sixteen labor areas were included in the jo b vacancy surveys
and accounted for approxim ately one-fourth of the N ation’ s em ploym ent. T w o
surveys were conducted in each of the 16 areas, the first spread over the last
quarter of calendar year 1964 and the second in April 1965. The 16 areas sur
veyed were as follow s:
Baltimore
Birmingham
Charleston, S.C.
Charleston, W . Va.
Chicago
Kansas City
Los Angeles
Miami
Milwaukee
Minneapolis-St. Paul
New Orleans
New Y ork
Philadelphia
Portland, Oreg.
Providence
Richmond
The fiscal year 1965 jo b vacancy program was experimental in nature. There
was no certainty that employers would furnish the data, that the great volum e of
detailed occupational inform ation could be processed in meaningful form , or that
the data so gathered would be useful in im proving the mechanism of the jo b
m arket.
Numerous technical concepts had to be mastered. The definition of a job
vacancy, a sampling technique structured to the absence of occupational bench
marks, the design of an effective collection form, preparation of instructions
covering the gathering, processing, and application of the data for use in E m ploy
ment Service operations, editing procedures, and response analysis evaluation
were but a few of the problems encountered in planning the surveys.
As part of the experimentation, 5 of the 16 areas were split into two groups for
control purposes. One-half of the surveyed employers in each area was given a
standard schedule and the other half a variant schedule. The purpose was to
measure response under varying reporting requirements and to test the feasibility
of collecting special types of inform ation such as wage data and part-tim e and
temporary vacancies.
The primary emphasis of the job vacancy program was determined to be for
E m ploym ent Service operating use in the job market, with statistics as a b y
product of the operational effort. Operational uses were directed toward in
creased placem ent of workers, vocational counseling and guidance,, establishment
of training classes under M D T A and vocational education, and encouragement
of worker m obility through matching of inter-area surpluses and shortages. Of
course, since com pletion of the first year’s experimentation, the job, market has
grown progressively tighter, and it may well be that subsequent uses of the p ro
JOB VACANCY STATISTICS
91
gram will focus on occupational shortages for defense and essential nondefense
industries and program policy on how to overcome or ameliorate them.
The most im portant finding of the surveys was that it is both feasible and
practicable to collect job vacancy information by detailed occupation. A very
large proportion of employers— approximately 80 percent of those sampled by
the USES and affiliated State sei vices in the 16 pilot areas— wrere able and willing
to provide these data. In addition, a series of controlled tests in some of the
16 areas indicated that employers appeared willing to furnish other tj^pes of data
relating to their vacant jobs, such as wage rates and inform ation on part-tim e
and tem porary vacancies. Although the high degree of cooperation in both
rounds of surveys occasioned some surprise, it bore out very closely the results
of the pre-test of feasibility conducted by the Illinois em ploym ent security agency
in Chicago in 1963.
A bout one out of every four employers responding in the 16 areas reported
at least one vacancy, although this varied by size of establishment. A bout one
of every tw o large firms had some vacancies, while in the smaller establishments,
1 in every 10 reported at least one vacant job.
The tw o experimental surveys indicated that the level of jo b vacancies fell
short of the number of unem ployed at the time of the surveys. In the 16 areas,
there were about 2.4 job applicants registered for work in the ES files for every
job vacancy. There were wide variations among areas, however, ranging from
an approximate balance in Richm ond and Milwaukee to a 10 to 1 relationship in
Charleston, W . Va.— long an area of high unemployment.
The problem of structural unemployment is clearly demonstrated by the im
balance between the types of workers needed, which were m ostly in the higher
range of skills, and the types of workers available, wdiich were mostly on the bot
tom rungs of the occupational ladder. The lowest ratios of applicants to vacancies
wTere in professional and managerial occupational classifications, where aggregate
vacancies exceeded applicants, and in the clerical and sales and skilled occupations,
where the number of applicants averaged about double the vacancy total. In the
service, semiskilled and unskilled groups, the applicant-vacancy ratio averaged
about 4 and 5 to 1.
Although vacancies occurred in a broad spectrum of occupations, the heaviest
demand was concentrated in a relatively narrow band. For example, nurses
accounted for one out of every four professional vacancies. Sales persons, clerks,
and stenographers and typists accounted for two-thirds of the vacancies in clerical
and sales occupations. Over half of the jo b vacancies in the service occupations
were for waiters and waitresses, kitchen workers, practical nurses, and hospital
attendants. One-fifth of the vacancies for unskilled workers were in warehousing.
Vacancies for skilled workers were more widely dispersed. Obviously, each area
reflected its unique industrial pattern.
A surprisingly high proportion of the jo b vacancies— about half— were hard-to*
fi.ll, as indicated by the fact that they had remained vacant 1 month or longer.
The identification of these hard-to-fill jobs in relation to total vacancies, by occu
pation, promises to be one of the most useful tools of the program. Over a period
of time, such data can be a guide to training opportunities, assuming other
variables such as wages and working conditions are not adverse.
TH E
E M P L O Y M E N T S E R V IC E
IN
TH E
P IC T U R E
The experimental jo b vacancy program indicated that the Em ploym ent Service
plays a far more im portant role in the total picture of jo b vacancies than was
previously recognized. In the cities surveyed, some 30 percent of total jo b
vacancies, excluding domestics, were listed with the Em ploym ent Service. There
is a reasonably close correspondence in the occupational distribution between
total job vacancies and unfilled Em ploym ent Service jo b openings. This cor
respondence suggests that the types of openings which the Em ploym ent Service
cannot fill readily are reasonably representative of hard-to-fill jo b Opportunities
for the jo b market as a whole.
The results of the surveys clearly demonstrate that, with respect to vacancies
as with other jo b market measures, the labor demand-supply differences by area
are very substantial. If the program becomes operational, an area-by-area
approach will be required to make the job vacancy program meaningful for man
power planning and operations.
USES
FO R D A T A
It is apparent that jo b vacancy data will have many operating uses. Probably
the most extensive use to date has been in connection with manpower training
92
JOB VACANCY STATISTICS
programs. Seven of the participating State agencies have used the inform ation
to establish training programs, for M D T A , OJT, vocational education, and in
special programs aimed at disadvantaged youth. Other uses by local offices
have been for counseling and jo b development.
The jo b vacancy program, through its listing of many hard-to-fill jobs in the
lesser-skilled occupations, has operational implications for the “ pov erty ” pro
gram. A bout half of the vacancies for unskilled workers listed in the first survey
had been open for at least a month. Prominent among these were laborers for
the metalworking, transportation equipment, warehousing, and construction
industries. Nearly two-thirds of the vacancies for semiskilled workers also had
been available 30 days or longer; among them were jobs in the machine shop,
construction, and textile manufacturing industries, and for chauffeurs, drivers,
and routemen. Of the vacancies in service occupations, approximately one-half
were hard-to-fill. M any of the vacancies were for practical nurses, hospital
attendants, porters, waiters and waitresses, and kitchen workers. The available
wage data do not show any special concentration of less-than-prevailing wage
offers for the lesser skilled jobs.
Inform ation on the value of vacancy information for job placement purposes is
inconclusive thus far. Concentration on technical and mechanical problems in
getting the program underway inhibited intensive local office use of these data
for jo b order development, in addition, many of the vacancies reported were
already listed with the Em ploym ent Service. Although there had been some
apprehension that job development and placement efforts flowing directly out
of this program might jeopardize the reporting, such job development efforts as
were made did not appear to have any effect on the willingness of employers to
supply data. A proposed revision of the data collection schedule will invite the
ask for recruitment assistance from the ES local office, if he desires
16 areas in the first survey, and in 5 of the areas in the second survey,
a response analysis was conducted jointly b y the Bureau of Labor Statistics and
the State ES agencies as part of the overall job vacancy survey program. As
described b y the Bureau of Labor Statistics, the objectives of the response analysis
survey were to: (1) determine how well the definition of a jo b vacancy was understood b v employers; (2) measure how wrell the instructions were interpreted by
employers; (3) provide inform ation on job vacancy recordkeeping practices of
employers; and (4) determine the reasons for nonresponse. In each of the six
areas, BLS personnel interviewed a randomly selected sub-sample of the respond
ents to the job vacancy survey immediately after its completion. According to
BLS, nearly 9 out of 10 establishments correctly reported their total number of
job vacancies. Correct reporting w^as found to be slightly more prevalent among
smaller establishments than among larger establishments. The interviews also
revealed that only about 5 percent of the establishments erred in reporting their
total number of job vacancies. The reported total number of vacancies was
found to be understated by about 3 percent as a net result of the detected error
in reporting by responding establishments. Under-reporting of vacancies pre
dom inated in larger establishments, whereas over-reporting was more usual in
smaller establishments;
There may be some further undetected downward bias in reporting owing to
the lack of employer records from which job vacancies can be reported. It was
found that establishments wdiich had complete or partial records on jo b vacan
cies were more prone to report vacancies than those which did not keep records.
Small random samples of the establishments which were nonrespondents in
the area jo b vacancy surveys were also interviewed as part of each of the 11follow up surveys. The principal objective was to determine the extent and direc
tion of any bias that might be introduced in the estimated total job-vacancies
by assuming that the nonrespondent establishments had the same rate of jo b
vacancies as the respondents. These interviews disclosed that the nonrespondent
establishments had substantially higher levels of current jo b vacancies than did
the respondent establishments. This was true for both larger and smaller
establishments.
Despite the pioneering aspects and the magnitude of the fiscal year 1965 pro
gram, the actual conduct of the surveys revealed only a few technical flaws in
the overall survey design and structure of the program. The most im portant
of these was an apparent weakness in sampling procedures as they related to
small firms. A new samplng design has been developed to correct this in the
fiscal year 1966 surveys. Some changes have also been made in the employer
collection form and instructions on the duration of the vacancies and also to
provide for the collection of wage data in all areas.
JOB VACANCY STATISTICS
93
The jo b vacancy program proposed for fiscal year 1966 will take account of
the fact that available funds are the same as for the previous year.
Tw o rounds of job vacancy surveys will be conducted in fiscal year 1966, if
time permits, in 16 areas, virtually all of which participated in the fiscal year
1965 experimental program. The first round of surveys will be in April and the
other in June 1966. The second may be changed in scope and content, depending
on the findings of the earlier survey.
Chairman P r o x m i r e . In the first place, this is a very fine and
helpful statement. You have indicated that nine-tenths of the job
vacancies in Baltimore fell in the category of “ hard to fill,” and only
a third of the vacancies in St. Paul-Minneapolis were in that category.
That would suggest to me that maybe there was something different
in the way the surveys were applied rather than a startling difference
between two cities which are fundamentally fairly similar.
Mr. C h a v r i d . Right. We recognize that. And it is conceivable
that the kind of verbal explanations that were given to employers in
one area could have been different from those in another area.
Chairman P r o x m i r e . If this concept is going to be useful to us,
it seems to us that that kind of discrepancy ought to be explored
pretty carefully.
Mr. C h a v r i d . It will be explored especially in terms o f the current
survey that is going on, that they are completing now. We will
compare to see whether such large discrepancies exist. We must
remember that this was the first attempt-----Chairman P r o x m i r e . Yes. I appreciate your reporting on this.
Mr. C h a v r i d (continuing). And there are lots of things that do not
seem to make complete sense. And something has to be done in
order to reconcile them.
Chairman P r o x m i r e . In discussing Milwaukee, Wis., you said that
a number of training programs were expanded so that some 3,000
underprivileged youths would have jobs in the next 2 years. Two
questions in connection with that. Perhaps you did not have control
over this program, but this seems to me to be a rather long period in
view of the experience that Milwaukee has had, where we have had
heavy unemployment in the last few years, and now labor shortages.
What would lead one to conclude that you will need people in the
next 2 years in a category in which the vacancies now exist?
I take it that most of those vacancies do not require training?
Mr. C h a v r i d . N o , they do not. We are faced in the training pro
gram with this kind of a problem. Here we have a significant number
of unemployed workers on the one hand, and on the other, we have
demands. So these represent short-range training requirements.
These are not the long range, the kind that the vocational education
normally gets.
Now, these data are not the only ones that are used in determining
what the future outlook might be in occupations. The job vacancy
information merely indicates that you are having shortages as of
now. You take these data and go to pretty much the same employers,
or to larger ones which reported these data, and you ascertain from
them whether they expect expansion in this particular occupation
before you start a training program. This training program, Mr.
Chairman, must look forward. And they all do. It cannot be based
entirely on the job vacancies as of now, unless you have this program
going over a period of 4 or 5 years. Then you can be sure when the
same occupation keeps reappearing as a shortage occupation.
63 - 947— 66 ------------ 7
94
JOB VACANCY STATISTICS
Chairman P r o x m i r e ;. Was the survey there used to eliminate any
particular training programs? You say there were 12 that were
expanded. Did you find that there weren’t job vacancies, and
therefore you would not proceed with any of the training programs?
Mr. C h a v r i d . I w ill h a v e to c h e c k t h a t o u t.
Chairman P r o x m i r e , Did I understand you to say that the sample
represented 50 percent of the employment in most areas?
Mr. C h a v r i d . Yes, sir.
Chairman P r o 5x m i r e , It was that large?
Mr. C h a v r i d . Yes. Well, the reason why it was so large is be
cause we took large employers with certainty. It is not 50 percent of
employers. It is 50 percent of employment.
Chairman P r o x m i r e : Probably a relatively small percentage of
the smaller firms and the entire group of large establishments defined
differently in different areas. At one point you said that you did not
interrogate firms that employed four or less, and at another point I
thought you said three or less. Which is the point which you don’t
serve?
Mr. C h a v r i d . Three or less.
Chairman P r o x m i r e . Have you had opportunity to determine
what portion of total employment this would involve?
Mr. C h a v r i d . This could be determined. I don’t know offhand
right now. It is a very small figure.
Chairman P r o x m i r e . It may be a small figure. It can be small in
manufacturing, but rather big in some service industries?
Mr. C h a v r i d . Yes, by industry, it would vary, there is no question
about it.
Chairman P r o x m i r e . So this would not give you an accurate
picture in some areas. It would not be a problem in many others?
Mr. C h a v r i d . In most of the areas we have this would not present
a serious problem. Perhaps in a few small ones it would. But in the
large areas I do not feel that going to three or less would in any way
change the results, the conclusions of the surveys.
One other problem that we have is that our unemployment insur
ance coverage is for employers of four or more workers. This is
required by law now. We try to make the surveys uniform among the
States. So we took as a cutoff four or more, which provided uni
formity among all the States that were involved in the program.
Chairman P r o x m i r e . Do you think it would be necessary to survey
all areas and all groups four times a year? In other words, how rapidly
does the structure of job vacancies shift as distinguished from the
general level?
Mr. C h a v r i d . This is something that I keep asking myself all the
time. And if we had had more experience, at least one year doing it
every quarter, we might be able to answer in some knowledgeable
fashion. But anything that I would say now would really be guess
work and opinion.
Chairman P r o x m i r e . Y ou might even want to increase it and have
it every month, I suppose, depending on what it shows. We do have
unemployment data every month.
Mr. C h a v r i d . Well, personally I do not see a need for a detailed
survey of this kind every month. I think that our Employment
Service unfilled job openings, which now represent something like
thirty per cent of all the job vacancies, would be good enough for the
JOB VACANCY
STATISTICS
95
two months for which we would have no job vacancy survey to indi
cate any significant thing that might happen.
Chairman P r o x m i r e . But the basic problem of data would be
job listings of the Employment Service supplemented by separate
surveys? In other words, you have the Employment Service data on
the one hand, and then separate surveys to supplement it, would that
be the way to proceed?
Mr. C h a v r i d . This is a very good question. You are suggesting
that the employers list jobs with us, we should not go to them and
try to use that as the basic information. The problem here is that
we don’t know whether an employer who uses the Employment
Service lists all his jobs with us or not.
Chairman P r o x m i r e . But you eventually could?
Mr. C h a v r i d . Eventually we might be able to do something like
that.
Chairman P r o x m i r e . So you could streamline it and reduce the
cost that way?
Mr. C h a v r i d . Right. If we had more experience with trying to
see the relationships between the two sources of data; unfilled job
openings and vacancies.
Chairman P r o x m ir e . Could you indicate for the record now the
ways in which job vacancies and unemployment data are not sym
metrical? You have already indicated that they are not, inasmuch
as the job vacancies definition begins from the first reporting, and
unemployment as I understand it is seven or more days. But there may
be other areas which are not symmetrical, and not comparable, and if
there are, I would like to have them.
Mr. C h a v r i d . Well, the job vacancies, as I have indicated, Mr.
Chairman, are a single shot at whatever the employer has.
Chairman P r o x m i r e . I have no particular objection on that.
Mr. C h a v r i d . The other is a person that has a day’s employment
in a week is counted as employed and not unemployed. So to that
extent this is not comparable.
I don’t know that there are any other major differences, except
one deals with the labor supply, and the other deals with labor demand.
I might point out this, and I think this is a very important factor,
that job vacancies are not necessarily filled by unemployed people.
When a new plant moves into a community it is not the unemployed
that go to work at that plant. It is the employed that get the jobs
in a plant, people who might not have had promotions-----Chairman P r o x m i r e . They are not in the labor market, they are
neither employed or unemployed?
Mr. C h a v r i d . That is right, some are not in the labor force.
The unemployed might get the jobs of those who quit one plant and
go to a new plant because the wages and working conditions in the
new plant are better.
So a direct comparison between the job vacancies and unemployed,
unless it is through some administrative action would not have too
much meaning, and especially in a very broad group. What I have
mentioned was the case in Milwaukee and Portland, Oreg. And
there was another area, I think, Birmingham, Ala. In one area there
was a sort of balanced situation for a given occupation and the other
was short. And the third was surplus. When you add up all these
figures you solve your problems; by balancing shortages in one area, by
96
JOB VACANCY STATISTICS
surpluses or unemployment in another. On paper you have solved
the problem in Milwaukee and Portland. So that one must be very
careful in aggregating these figures.
Chairman P r o x m i r e . H ow much would a satisfactory national
program of regular checks of job vacancies by industry, area, and
occupation cost? I am assuming that the $2.5 million being requested
does not do the full job.
Mr. C h a v r i d . The $2.5 million requested was requested for some
80 or 90 areas, thereabouts. The total number of standard metro
politan areas is about 220. The number of what we call very large
standard metropolitan areas is about 150. I do not know what it
would cost, but it could well cost double the figure requested. But
this will have to be examined very carefully.
Chairman P r o x m i r e . Around $5 million. What additional bene
fit would the government derive and the public derive from this
additional expenditure?
Mr. C h a v r i d . One thing that we would have on tap would be
information on job vacancies in all of our areas. It is possible that
some system could be developed, or these things would be put in some
national data bank, as has been suggested by the Automation Com
mission. And the same thing could be done with the unemployed
in individual areas. This would all go into one bank. And then we
would try to see what matches we could make as between the areas of
surplus and shortage in individual occupations. But this recom
mendation has been made by the Automation Commission.
Chairman P r o x m i r e . Just one more question. Will it be possible
to specify the statistical accuracy of a series on job vacancies?
Mr. C h a v r i d . Yes, I think it can be done.
Chairman P r o x m i r e . Thank you very much, you did an excellent
job of summarizing a long statement and giving it to us briefly.
Mr. Curtis?
Representative C u r t i s . D o you attempt or have you been attempt
ing to find out what has caused the vacancies?
Mr. C h a v r i d . The vacancies are caused primarily by expanding
employers in these individual areas, and also I think in part by in
adequate in-migration of people from the outside. I say this-----Representative C u r t i s . Let me interrupt. I was not asking really
for a conclusion as much as I was to find out whether you attempt in
your questions to identify what created the vacancy?
Mr. C h a v r id . N o. We have not asked those questions from the
individual employers.
Representative C u r t i s . For instance, I was thinking of the fact
that every company of any size has a regular turnover, retirement,
and so forth. Now, the actual jobs that become vacant are not those
held by the person retired or dead. Frequently, they open up at
the lower end, but each company probably has a regular turnover
rate. You have not attempted to develop that statistic, then?
Mr. C h a v r i d . No. I think that this would be a very worthwhile
special project, to select a number of establishments in certain manu
facturing industries, in nonmanufacturing, and do it on a special basis
rather than ask all of the employers, because this could be quite a
bit of work for them.
Representative C u r t i s . I was trying to find out whether you did,
and then to get into whether it would be worthwhile. I suspect that
JOB VACANCY STATISTICS
97
certain industries have a big turnover, and certain localities probably
vary on this.
Well, thank you very much.
Chairman P r o x m i r e . Thank you, Mr. Curtis.
And thank you very much, Mr. Chavrid.
I think these are very valuable hearings. There will be inserted
in the record at this point a letter from John H. Aiken, Executive
Director, Federal Statistics Users’ Conference and the statement of
the Conference.
(The material referred to is as follows:)
F e d e r a l S t a t is t ic s
U sers’ C o n fer en c e,
Washington, D.C., M ay 17, 1966.
Hon. W i l l i a m P r o x m i r e ,
Chairman, Subcommittee on Economic Statistics, Joint Economic Committee, New
Senate Office Building, Washington, D.C.
D e a r S e n a t o r P r o x m i r e : The Board of Trustees of the Federal Statistics
Users’ Conference was pleased to learn that the Subcommittee on Econom ic
Statistics of the Joint E conom ic Com mittee decided to hold hearings on the
feasibility of regular collection and reporting of jo b vacancy statistics and their
potential usefulness in formulating manpower policy at the local and national
levels.
Because of the increasing interest in the need for, and feasibility of, collecting
and reporting this type of information, the Board of Trustees of the Conference
wishes to file a written statement supporting the efforts of the Subcommittee to
examine the problems involved and particularly the results o f pilot studies made
by the Bureau of Labor Statistics and the National Industrial Conference Board.
Enclosed is a copy of the written statement of the Conference wrhieh we re
spectfully request be included in the printed record of the hearings.
Respectfully subm itted.
J o h n H . A i k e n , Executive Director.
S t a t e m e n t S u b m it t e d b y t h e F e d e r a l S t a t is t ic s U s e r s ’ C o n f e r e n c e o n
F e a s ib il it y o f R e g u l a r C o l l e c t io n a n d R e p o r t in g o f J ob V a c a n c y
S t a t is t ic s
This statement is submitted on behalf of the Federal Statistics Users’ Con
ference which is an association of 152 member companies and organizations
comprised of business firms, labor unions, and nonprofit research groups. Con
ference members study and utilize all types of Federal statistics for many and
diverse purposes, particularly in planning and in making operating and policy
decisions. Thus, Conference members have a com m on interest in obtaining
useful, adequate, timely, and reliable inform ation from Federal statistical pro
grams. Its members are representative of a broad segment of econom y and are
engaged in such activities as advertising, banking, insurance, manufacturing,
retail trade, printing and publishing, econom ic and market research, etc.
The Board of Trustees of FSU C was pleased to learn that the Subcomm ittee
on E conom ic Statistics of the Joint Econom ic Com mittee decided to hold hear
ings on the feasibility of regular collection and reporting of job vacancy statistics
and their potential usefulness in formulating manpower policy at the local and
national levels.
Our purpose in submitting this statement is: (1) to support the efforts of the
Subcom m ittee in examining the results of pilot studies made b y the Bureau of
Labor Statistics and the National Industrial Conference Board, (2) to indicate to
the Subcomm ittee our interest in, and concern with, developing inform ation on the
need, uses, limitations, proper use, feasibility and benefits of jo b vacancy statis
tics, and (3) to offer our cooperation in the future, if practical, in further studies
or programs which m ay result from these hearings.
For some years now, the Conference has recognized the need for developing jo b
vacancy statistics, but has been concerned with the feasibility, uses and lim itations
of such statistics— problems which the Subcomm ittee on Econom ic Statistics is
now taking a look at in these hearings.
A t the Eighth Annual M eeting of FSUC, held in O ctober 1964, the R eport of
the M anpower Statistics Com mittee of the Conference said: “ Some comparisons
between characteristics of the unem ployed and characteristics of vacant job s is
'9 8
JOB VACANCY STATISTICS
needed. Attem pts to measure vacancies have been unsuccessful in the past, bu t
it would be worth exploring this field again.”
Since the initiation of the exploratory studies undertaken b y the Bureau of
Labor Statistics and Bureau of Em ploym ent Security in Fiscal Y ear 1964, the
Conference has supported budget requests of B L S -B E S to continue exploratory
studies at the same level. H owever, the Conference has been reluctant to support
any proposals for inauguration of the program on a large scale until results of the
exploratory work are made public and proper study and evaluation of the work
can be made.
M em bers of the Conference have been well aware of the problems involved in
developing adequate and useful job vacancy statistics and have been particularly
concerned about resolving conceptual and definition problems. They are par
ticularly anxious to learn what has been done about solving these problems.
Naturally, we are hopeful that these hearings will provide sufficient inform ation
to indicate the progress made, what problems have been solved, what problems
remain unsolved, and the direction and course which should be taken in the future.
While the demands for job vacancy information have come from a number of
theoreticians who study em ploym ent problems, there is no practical evidence
yet available to the public as to the feasibility of collecting reliable inform ation
or whether or not such information would be of material assistance in the form u
lation of public policy.
Nothing herein, of course, should be taken as expressing any opposition to the
expansion and im provement of the employment services from an operating
standpoint. W e recognize that there is a great need for a better matching of
people and jobs. However, there is some doubt in the minds of many whether
the collection and reporting of statistical data on job vacancies will help in this
task. It m ay be that an expansion and improvement of em ploym ent service
operations will bring forth additional operating statistics that will enable such
services to gauge their efficiency and cast more light on the state of the labor
market.
In conclusion, the position of the Conference on the problem of collecting job
vacancy statistics was stated in our testim ony given on M arch 24, 1966, before
the Subcomm ittee on Labor, Health, Education, and Welfare of the House
Com m ittee on Appropriations. It reads:
“ The Conference believes that continued and more intensive experimentation
at the present level would be wise; it also believes that inauguration of the pro
gram on a large scale as contemplated would be unwise and a significant departure
from past practice in dealing with statistical programs of like size and com plexity.
“ Further work on the resolution of conceptual and other problems should go
forw*ard. The effectiveness of a regular job vacancy program as a tool for
prom oting local labor market efficiency should be tested . . .
“ This approach is consistent with the BLS approach to im proving unem ploy
m ent statistics, where proposed changes have been subject to experimentation
on a large-scale over a period of years before introduction into the regular unem
ploym ent statistics program. It is consistent with the developm ent of statistics
o f em ploym ent by occupation which is proceeding over a period of several
years . . . ”
M em bers of the Conference will be following these hearings with keen interest.
Chairman P r o x m i r e . While it tends to cover some of the ground
covered by other material, I think it would be well to include in the
record an excellent summary article on the Labor Department’s pilot
studies presented at the conference held by the National Bureau of
Economic Research last year. The report of the conference entitled
“ The Measurement and Interpretation of Job Vacancies/’ which
covers some 600 pages contributed by a score of authorities, suggests,
incidentally, the wide and thoughtful concern generated by the
subject.
Since all observers and possible users of job vacancy statistics need
to know everything that we can know about them before their collec
tion and reporting is undertaken on a regular basis, I think that one
article in particular out of this symposium is a useful summary of the
techniques, collection methods and problems which have come up in
the pilot studies conducted by the Bureau of Labor Statistics.
99
JOB VACANCY STATISTICS
The article from the proceedings of the National Bureau Con
ference entitled “ Experimental Job Vacancy Survey Program of the
United States Department of Labor,” by Irwin F. O. Wingeard of
the Bureau of Labor Statistics follows.
E
x p e r im e n t a l
Job
V acancy
Su r vey P rogram
of L a b o r
of
the
U .S .
D
epartm ent
(By Irwin F. O. Wingeard, Bureau of Labor Statistics)
in tr o d u c tio n
The Departm ent of Labor has been concerned about the need for job vacancy
inform ation for many years. Its Bureau of Em ploym ent Security has experi
mented with a number of pilot programs, some of which date back to W orld War
I I and the Korean conflict.
In 1956, the Bureau of Labor Statistics conducted a pilot feasibility study,
which concluded that it was impractical to initiate a regular mail collection of
statistical data on job vacancies. This conclusion was reached, at that time,
because of the unavailability of employer records on job vacancies, and the in
ability of a large proportion of the employers to estimate their jo b vacancies.
Despite this earlier negative finding, the President’s Com m ittee to Appraise
Em ploym ent and Unemploym ent Statistics (known as the G ordon Com m ittee),
was impressed by the widespread interest in and the potential uses of meaningful
statistics on job vacancies. The Bureau of Labor Statistics was requested to
prepare a report for the Com mittee that would take a fresh look at the possibility
o f setting up a national statistical program. The report subm itted suggested
a program of experimental research that might be followed in attem pting to solve
the technical problems involved and to assess the feasibility of developing such
statistics.
The G ordon Committee recommended that the Department of Labor initiate
the program of research suggested (which was published as Appendix B of the
Com mittee report, “ Measuring Em ploym ent and U nem ploym ent7’).
This research was undertaken on a cooperative basis by the D epartm ent’s
Bureau of Em ploym ent Security and Bureau of Labor Statistics, beginning in
fiscal 1964. The modest program included research on conceptual and defintional problems, pilot feasibility studies in the Chicago and Buffalo areas, and a
survey of the nature and uses of job vacancy statistics being collected in foreign
countries (on which Shelton will report in his paper).
o b j e c t iv e s
Following this preliminary research, the Departm ent of Labor launched a more
comprehensive experimental program for the collection of job vacancy information
in the current fiscal year (1965). This program was conducted jointly by the
Bureau of Em ploym ent Security and the Bureau of Labor Statistics and the
analysis and evaluation of the results is their shared responsibility. However,
the employm ent security agencies in the states involved are actually conducting
the individual area surveys included in the program and are summarizing the
results, under the administrative direction of the Bureau of Em ploym ent
Security.
One of the program ’s tw o primary objectives is to evaluate the feasibility of
collecting reliable and meaningful job vacancy data by occupation. The other
is to assay the usefulness of the job vacancy information collected in furthering
the operations of the public employm ent service, particularly in helping to place
unem ployed workers in available job openings and identify occupational training
needs. Another long-run aim is to assess the value and practicability of even
tually using the job vacancy data gathered for purposes of econom ic analysis and
public policy.
problem s
In designing the experimental program, it was necessary to consider and resolve
an appreciable number of difficult problems relating to the collection of job vacancy
inform ation. Some of the more im portant of these were the following:
1.
T o design a survey that would yield inform ation to meet the needs of U.S.
Em ploym ent Service operations, and at the same time provide reliable and mean
ingful data for general statistical and analytical uses. Although operational uses
100
JOB VACANCY STATISTICS
were given predominant consideration, the opportunity was taken to experiment
with approaches relevant for statistical and analytical purposes, keeping in mind
the need for geographical and occupational detail.
2. The delineation of the scope of the industry and establishment coverage that
should be included in the area samples. Special problems arise in attem pting to
attain com plete coverage, because employer reports filed under state unem ploy
ment insurance programs are used as the primary source for universe listings of
establishments, and certain coverage exclusions are com m on in those programs.
Farms, private households, railroads, and nonprofit organizations are universally
om itted, and small establishments with less than four employees and state and
local governments are generally excluded. Special problems must be solved and
different techniques developed with regard to sampling, collection, and estimation
for farms, private households, and very small establishments.
3. T o determine what size of sample would be required to obtain a reasonably
accurate measure of job vacancies by occupation within a geographic area. This
is a m ajor problem because the variable being measured usually represents a very
sm all segment of the total employm ent in any one occupation in the area, and
generally occurs with widely scattered and highly fluctuating incidence among
the different employing establishments. A sample spread thinly across all
em ploying establishments may very well miss significant clusters of vacancies for
som e occupations, as w ell as all vacancies for other occupations.
4. T o devise an appropriate mechanism for the collection of job vacancy infor
m ation which would produce a high rate and quality of response and make timely
vacancy information prom ptly available for administrative and analytical uses.
There is a question whether job vacancy information can be collected b y mail with
the voluntary cooperation of employers. Another question is whether the collec
tion system should be an independent one or tied in with an existing regular collec
tion program, such as the B L S -B E S -state agency cooperative program relating to
labor turnover statistics. Another question is whether employers would respond
more willingly and accurately if the collection was divorced from the local em ploy
ment offices, or was made solely for statistical purposes and without follow-up jo b
order developm ent efforts by the local employm ent offices.
5. T o draft a suitable instrument for the mail collection of job vacancy inform a
tion b y occupation which would yield accurate, reliable, and consistent responses,
and give sufficient detail as to the characteristics of job vacancies for adminis
trative and analytical purposes. The report form must not be so complex or
overburdening as to discourage mail response, and its definitions, instructions,
and form at must be clearly understandable to all reporters to whom it is mailed.
A problem exists as to how to identify and measure so called “ hard-to-fill” vacan
cies which signify the existence of labor shortages. There is also a question as to
whether a request for the rates of pay being offered for vacant jobs would
strengthen the authenticity of reported job vacancies without damaging the
response and statistical validity of the surveys.
6. T o evolve a definition for a job vacancy measure which would be suitable for
analytical purposes, as well as for Em ploym ent Service operating uses. For the
former purposes, the job vacancy measure should be conceptually comparable with
the unem ploym ent measure used in the monthly labor force survey. For opera
tional uses, a less restrictive concept might be more suitable. For both purposes,
however, a vacancy must represent something more tangible than an unoccupied
job slot appearing in an organizational chart or table, and something more con
crete than mere intention or desire to recruit or hire a worker.
7. The establishment of procedures for the collection and classification of jo b
vacancies b y occupational detail. Difficulties are encountered because of the
absence of standard job classification systems and uniform job title nom encla
ture in industry. For Em ploym ent Service placem ent purposes, detail by indi
vidual occupations would be preferable, whereas for analytical usage less refine
ment would be adequate. For comparison with vacancy data, total unem ploy
ment figures are available only by broad occupational groups, and insured
unem ploym ent data are compiled at an intermediate level of detail (three digits
of the Dictionary of Occupational Titles coding structure).
8. The construction of a suitable technical m ethodology for inflating the area
sample results to estimated area universe totals by occupation.
G E O G R A P H IC
COVERAGE
Under the D epartm ent’s experimental program, pilot job vacancy surveys are
being conducted for two different time periods in each of the following sixteen
Standard M etropolitan Statistical Areas: Baltimore, Birmingham, Charleston
101
JOB VACANCY STATISTICS
(S .C .), Charleston (W .V a.), Chicago, Kansas City, Los Angles, Miami, Milwaukee,
Minneapolis-St. Paul, New Orleans, New York, Philadelphia, Portland (Ore.)
Providence, and Richm ond. The areas were selected so as to include m ost of
the country’s largest em ploym ent centers and to give some range of representa
tion to areas of different labor force sizes, geographic regions, industrial charac
teristics, and em ploym ent conditions. The areas were not selected as a repre
sentative sample of the nation nor of all Standard M etropolitan Statistical
Areas. Accordingly, the results of the sixteen area surveys cannot be com bined
to derive any over-all estimates
An area approach was follow ed mainly because of the D epartm ent’s primary
interest in seeing the data used to help place unem ployed workers in available
jo b openings. In addition, however, it was thought that the use of selected areas
was the most suitable and manageable way of testing collectibility and solving
response, definitional, and estimation problems across a broad industry spec
trum, as was desired under the experimental program. M oreover, it was recog
nized that geographic detail would sooner or later be essential under any system
that might eventually be established for the collection of job vacancy data for
any purpose.
E S T A B L IS H M E N T
SAM PLE
AND
IN D U S T R Y
COVERAGE
A prescribed method for the selection of a probability sample of employers in
each area was provided to the state agencies in order to make it possible to inflate
the area sample results to estimated area universe totals. The universe from
which the sample was selected included all establishments with four or more
employees which were covered by the State Unemployment Insurance Law in
the first quarter of 1964, supplemented by a list of known nonagricultural establish
ments employing 100 or more workers, which were not covered by the State Un
employm ent Insurance Law. The instructions permitted the inclusion of non
covered establishments with fewer than 100 employees where this was considered
necessary to assure adequate representation in certain nonagricultural industries.
The instructions also permitted the use of first quarter 1963 unemployment
insurance records where those for the first quarter 1964 were not available. The
universe, therefore, generally extended across all nonagricultural wage and salary
payroll employm ent, except for the very small establishments.
The scope of the industry coverage was very broad and generally consistent
with that covered by the establishment employm ent series published by the state
agencies in cooperation with the Bureaus of Labor Statistics and Em ploym ent
Security. This wide industry coverage was desired in order to gain experience in
collecting jo b vacancy data from as many different industries as possible.
Establishments with fewer than four employees, as well as farms and private
households were not included in this experimental program because they require
the development of special techniques with regard to sampling, collection, and
estimation. When a regular and full-scale program for collection of job vacancy
data is launched, selecting a sample will present some problems because farms and
private households are universally omitted, small establishments are generally
excluded from state unem ployment insurance coverage, and employer reports
under the unem ployment insurance program are used as the primary source for
universe listing of establishments. This problem extends also to the inclusion
o f railroads, state and local governments, and nonprofit organizations that are
usually not covered in state unemployment insurance programs, but a makeshift
effort was made to include them in the sampling frames for the sixteen area
experimental surveys. The eventual inclusion of adequate representation for
all of these excluded places of employm ent would be important to an established
full-scale system for the collection of jo b vacancy statistics, because they may
have many of the job vacancies that are available to and suitable for unemployed
workers. Recognizing the importance of m oving forward with research on this
front, the Department is conducting a separate one-time experiment to try to
find out the extent and nature of vacancies for domestic workers in the private
households and farms covered in the January 1965 labor force survey, conducted
as part of the Current Population Survey. A major limitation of this kind of
survey for use in a job vacancy statistical collection system, however, would be
its inability to provide results for individual geographic areas.
The sample size in each of the sixteen area experimental surveys was rather
large, since it included all of the larger establishments which, when arrayed by
size in descending order, had a cumulative employm ent total equal to 50 per cent
of the area employm ent universe. In addition to this rather large certainty
stratum for larger establishments, the sample included a fixed number of smaller
102
JOB VACANCY STATISTICS
establishments, amounting to 1,000 in each of the four largest areas and 500 in
each of the remaining areas. In order to assure a good industry and size mixture
am ong the smaller establishments, they were selected at random from a listing
of all of the smaller establishments in the universe (comprising the remaining
50 per cent of the area em ploym ent universe) arrayed by em ploym ent size within
each tw o-digit standard industrial classification. It is uncertain whether the
rather substantial sample sizes established for the experimental area surveys
are adequate to provide reasonably accurate results for the extensive occupational
and other detail wanted for each area.
Several considerations entered into the decision on the sample size; cost, the
aim for essentially equal precision in the results obtained for the various areas,
variation in size of labor force among the areas, and the desire for results that
w ould give considerable occupational detail in each area and be useful for jo b
developm ent and worker placem ent in most areas.
C O L L E C T IO N
M ETHODS
The experimental character of this program afforded an opportunity to test
various collection schedules and methods. One question to be tested was whether
em ployers wTould be more willing to respond if the survey was made solely for
statistical purposes and w ithout follow -up calls by the local em ploym ent office to
solicit jo b orders for vacancies reported. This issue appears to be one of the
m ost controversial aspects of the D epartm ent’ s experimental program. Concern
has been expressed from a number of quarters that the statistical results of the jo b
vacan cy survey would be biased because some employers, quite possibly including
large and im portant ones, would withhold their cooperation or report incom pletely
on account of their dissatisfaction with the tie-in of Em ploym ent Service jo b
order developm ent efforts.
On the other hand, some employers welcome Em ploym ent Service assistance in
filling their vacancies and appreciate being contracted for jo b orders. The
E m ploym ent Service view is that, where there are jo b openings available which it
can help to fill, it would be remiss in its obligations to both employers and workers
it it did not call this fact to the attention of the employers. It is socially and
econom ically desirable to im prove mechanisms to speed the m atching of workers
and job s and reduce “ time-lag frictional unem ploym ent.”
A test was made to determine the effect on response of a solely statistical survey
versus one directly related to Em ploym ent Service placem ent activities. In
three areas— Charleston, S.C., Chicago, and Miami— data collection was handled
by the central office of the state agency, and employers were assured, when asked
to report their jo b vacancies, that they would not be asked for jo b orders unless
they specifically requested that this be done. In most other areas, data collection
was conducted by the Em ploym ent Service local offices, and job vacancies reported
by individual employers were used as leads for possible solicitation of jo b orders
from those employers.
In all areas except the three where no jo b order solicitation was attempted,
personal visits were used to introduce and explain the significance of the survey, to
appeal for voluntary cooperation, and to deliver the schedule to a substantial
number of sample employers, particularly the larger ones. In addition, a letter for
these same purposes was directed to all sample employers in the three areas
without job order solicitation and to remaining establishments in all other areas.
State agencies were requested to make a special effort to direct the schedule to the
official in each establishment most likely to be able to complete it. They were
also urged to publicize the survey and its timing in advance, through informational
media and contracts with employers and civic organizations.
Tw o follow -up contacts with nonrespondents were to be made in all areas. In
the three areas without job order solicitation, the first nonresponse follow -up
contact was made entirely by mail, by the central office, and the second entirely
by telephone by the central office, or by the local office where it was impractical for
the central office to do so. In all other areas the first nonresponse follow -up con
tact was made entirely by telephone and the second by personal visit to the largest
establishments and by telephone to all other establishments.
The initial surveys in the various areas were conducted at different times
because of the lead time needed by the Department and the state agencies to
develop plans and procedures for the variety of experiments included under the
program. Accordingly, the month of reference was October 1964 for four areas,
N ovem ber 1964 for seven areas, Decem ber 1964 for four areas, and January 1965
JOB VACANCY
for one area.
below :
STATISTICS
103
The specific areas surveyed in each of these periods is shown
Month of reference:
Areas surveyed
October 1964____ Birmingham, Milwaukee, Portland (Oreg.), and P rovi
dence
N ovem ber 1964__ Baltimore, Chicago, Charleston (S.C.), Miami, M inneapolis-St. Paul, New York, and Richm ond
D ecem ber 1964__ Kansas Citv, New Orleans, Los Angeles, and Charleston
(W . Va.)~
January 1965____ Philadelphia
It is planned to have each of the areas repeat their initial survey in April 1965.
This repetition is desired to test the willingness and ability of employers to report
and the consistency of responses that would be obtained under a continuing
survey program.
C O L L E C T IO N SC H E D U L E S
T o further the testing desired under this experimental program, a standard
collection schedule and six different variant collection schedules were designed and
used to collect job vacancy information (see example copy at end of this paper).
All schedules requested that the total number of current job vacancies be reported
by occupation, using establishment job titles. The schedules differed, however, in
the extent of additional detail requested relative to the duration of the vacancies,
the expected duration and part-time or full-time status of the jobs to which the
vacancies related, and the rates of pay being offered for the vacancies.
The standard schedule called for reporting the number of current job vacancies
existing one work week or longer and the number existing one m onth or longer, in
addition to the total number of current job vacancies. All three items were to
be reported by occupation. The first variant schedule asked for rates of pay
being offered for the vacancies, in addition to all of the inform ation requested on
the standard schedule. The second variant was like the standard schedule in all
respects, except that it excluded the breakdown for the number of vacancies
existing one work week or longer. The third called for the reporting of only the
total number of current job vacancies by occupation. The fourth asked how
many of the total number of current job vacancies related to part-time jobs only,
in addition to all of the information requested on the standard schedule. The
fifth asked how many of the total number of current job vacancies related to short
term jobs that were expected to last not longer than three days and how many
related to temporary jobs that were expected to last over three days but not more
than four months, in addition to all of the information requested on the standard
schedule. The sixth variant schedule requested the reporting of all of the detail
called for on all of the other schedules, but in a different manner. It asked for a
separate line-item descriptive entry for each individual job vacancy, rather than
an aggregate count by occupation and other characterictics as all of the other
schedules did.
Each of the items of information on the survey schedules was thought to be of
considerable importance for meaningful interpretation, analysis, and use of jo b
vacancy data for administrative or analytical purposes. They were not included
on all schedules, however, because there was not enough experience to indicate
whether they were collectible or what effect their inclusion might have on the rate
and quality of employer response and on the level of job vacancies reported.
Vacancies existing one work week or longer were included primarily for analyti
cal usage, because of the concept held by many persons that such a measure
of job vacancies is most appropriate for making comparative analyses between
jo b vacancies and the unemployed. This concept implies that a job vacancy
must have been “ in search of a worker” for one week or longer in order to be com
parable to an unemployed worker who must have been without a jo b for one
week or longer and looking for work. For administrative purposes, no particular
importance is attached to vacancies existing one work week or longer because it
alone does not denote the so called “ hard-to-fill” openings which are of special
significance for those ends. Inform ation on job vacancies existing one work week
or longer were requested from the entire sample in fourteen areas. It was om itted
for the entire sample in New York, and in the Philadelphia area for the one-half
of the sample that was asked to report total vacancies only.
Vacancies existing one month or longer were included as a means of obtaining
a measure of so-called “ hard-to-fill” vacancies. Knowledge about such vacancies
was considered to be important for both administrative and analytical purposes,
104
JOB VACANCY STATISTICS
because such vacancies are believed to signify the existence of labor shortages
for particular occupations. Vacancy duration may not be entirely satisfactory
as a means of identifying “ hard-to-fill” job openings; other factors may influence
the length of time needed to fill openings, such as the urgency of the em ployer’s
need for workers, the kind and intensity of recruitment efforts, and the level of
wages offered. In addition, the customary length of time required to fill jo b
openings varies considerably according to the kind and extent of occupational
skill and training required. One month or longer may usually be required to fill
technical, professional, or other jo b openings requiring specialized skills and
training. Despite these limitations, it was decided to use a vacancy duration of
one m onth or longer to indicate “ hard-to-fill” vacancies because a suitable and
reasonably simple alternative criterion was not available. Inform ation on
vacancies existing one month or longer was requested in fifteen areas from the
entire area samples. In the one remaining area, Philadelphia, it was om itted for
the one-half of the area sample which was asked to report total vacancies only.
Separate breakouts of vacancies for part-time and tem porary jobs were included
in order to measure the extent of their significance in the total vacancies reported.
It was recognized that part-time and temporary work is of considerable im por
tance, particularly in certain industries, such as trade and service. It was thought
that inform ation on the scope of vacancies for these kinds of jobs would be of
value for both administrative and analytical uses because the character and
significance of the demand for this kind of labor differs materially from that for
full-time, permanent jobs. Inform ation on vacancies for part-time job s was
asked from one-half of the area samples in Charleston, W. Va., and New Orleans.
Similarly, data on vacancies for temporary jobs were requested from one-half of
the area samples in Kansas City and New Orleans. In New Orleans, the lineitem entry schedule was used to request both of these items, whereas in Charleston,
W . Va., and Kansas City, respectively, the ‘ ‘part-time jo b s” and “ tem porary
jo b s” schedule variants were employed.
Inform ation on the rates of pay being offered was included primarily as a
means of assuring the authenticity of the vacancies reported. Representatives of
labor felt that the request for rates of pay might very well curb the reporting by
employers of substandard and unauthentic vacancies, and might also permit some
evaluation to be made of the extent of substandard wage offers among the reported
vacancies. Others have voiced the fear that some employers might be unwilling
to report their wage offers to Em ploym ent Service local offices, or possibly to any
other agency, even though such offers are not suspected of being substandard.
If the inclusion of a request for pay rates offered strengthens the authenticity of
reported jo b vacancies without damaging the statistical surveys, it would con
tribute significantly to the value of the job vacancy data for both administrative
and analytical purposes. Administratively, the specific wage offers quoted also
would be useful in soliciting jo b orders and placing workers. Because of the
special im portance attached to the question on rates of pay offered for vacancies,
it was asked in six areas. In Los Angeles and New Orleans, it was requested
from only one-half of the area samples; for the former, by means of the pay-rate
offered schedule; for the latter, the line-item entry schedule. In the other four
areas in which this information was requested— Baltimore, Chicago, Miami, and
Minneapolis-St. Paul, the pay-rate offered schedule was used for the entire area
sample.
Testing so many variant approaches with only sixteen areas made it difficult
to reach conclusions about the relative effectiveness of each variant, because
em ployer response might also be affected by the industrial composition, size, or
location of each area, or by attitudes toward the job vacancy survey, surveys and
reports in general, or Em ploym ent Service local offices. It was therefore decided
to control for differential general characteristics of areas by splitting the reporting
sample within an area. This technique makes possible a more definitive assess
ment of the extent to which the degree and character of employer response would
be affected by requesting varying kinds of information, and by using schedules of
varying degrees of complexity. In five of the areas, Charleston, W. Va., Kansas
City, Los Angeles, New Orleans, and Philadelphia, the standard schedule was
used to collect inform ation from one-half of the employer sample and one of the
variant schedules was used to gether data from the other half. In each of the
other eleven areas, only one type of schedule was used to collect inform ation from
the entire sample. The distribution of areas according to type of schedule used
is shown in the following tabulation:
JOB VACANCY STATISTICS
Schedule number
Type of schedule
Number
of areas
105
Areas where schedule was used
___ Standard___ ________ ______
U
JV-2........... ................
Pay-rate offered......................
5
JV-3_______________
Omits vacancies existing 1
workweek or longer.
Total only_________________
Part-time jo b s .____________
Temporary jobs___________
Line-item entry____________
1
Birmingham, Charleston, S.C., Charles
ton, W . Va. (for ]/2 of area sample),
Kansas City (for y2 of area sample),
Los Angeles (for
of area sample),
Milwaukee, New Orleans (for of area
sample), Philadelphia (for >2 of area
sample), Portland, Oreg., Providence,
Richmond.
Baltimore, Chicago, Los Angeles (for K of
area sample), Miami, MinneapolisSt. Paul.
New York.
1
1
1
1
Philadelphia (for K of area sample).
Charleston, W . Va. (for H of area sample)Kansas City (for M of area sample).
New Orleans (for H of area sample).
JV-1_______
_
JV-4_______________
JV-5___ _________
JV-6_______________
JV-7_______________
CONCEPTS AND
D E F IN IT IO N S
The same jo b vacancy and reference period concepts and definitions were used
on all scheduies. The specified reference period was as of the close of business on a
given date, which in each instance was a Friday. The Friday of the calendar
week which included the 12th of the month was generally used, because that week
is the standard reference period used in the household labor force and establish
ment em ploym ent surveys. For the seven area surveys conducted in November,
however, the Friday of the following calendar week was used because additional
time was needed to prepare for those surveys. The definition of job vacancy, as
used in all surveys, is shown at the end of this paper. The most im portant feature
of this definition is that is spells out three conditions which must be met before a
job vacancy can be counted. First, the job must be unoccupied; second, the jo b
must be available for immediate occupancy b y a new worker from outside the
com pany; third, the job must be the object of management’s active search for a
new worker from outside the com pany. These requirements were stipulated to
secure conceptual com patibility with the unem ployment definition. Just as an
individual has to be actively looking for a job to be considered as unem ployed so a
“ bona fide” jo b vacancy must entail management’s positive effort to find a new
employee and not merely its intention or desire to hire one. Similarly, like an un
employed worker, a job vacancy must be unoccupied and available for im mediate
occupation.
For administrative purposes the definition used may be too restrictive in con
cept, since it excludes those situations in which new workers are currently being
sought through positive efforts to fill jobs which, although not immediately avail
able, are expected to becom e vacant or available in the future. The inclusion of
those openings might be particularly useful in connection w ith em ploym ent service
operations. T o accom m odate this use, it might be advantageous to collect data
on those openings in the future, but as a separately identified item. A separate
breakout would be essential because such expected openings would not be concep
tually complemental to current unem ployment.
Other questions may be raised in regard to the definition used for job vacancies.
Should it include those jobs that are expected to be filled in the future by recall
of employees on layoff, or by new workers, already hired, who have not yet
started working? Rather convincing arguments can be made both for and against
the inclusion of each of these.
S U M M A R IZ A T IO N
AND
E S T IM A T IO N
OF R E S U L T S
The editing, coding, tabulation, and inflation of the survey results was done
b y the state agencies in accordance with procedures prepared jointly by the Bureau
of Labor Statistics and the Bureau of Em ploym ent Security.
The job titles reported were classified and coded by occupation according to
the Dictionary of Occupational Titles {D O T ). This task was particularly trouble
some because no standard terminology is universally used in industry. Some
establishments, especially small ones, do not have formal job classification systems
and may tailor the description of their jobs to suit the capabilities of particular
applicants. Other establishments may use general job titles to cover broad
ranges of occupational duties and responsibilities. For these, as well as jother
106
JOB VACANCY
STATISTICS
reasons, some establishments might have been unable to use job titles which would
permit appropriate classification by D O T code. Possibly for many of the reported
vacancies the local personnel of the Em ploym ent Service were sufficiently well
acquainted with the occupational requirements of the employers, or were able
to obtain enough additional information from employers by telephone, to classify
them properly by D O T code, at least to the three-digit level of specificity for
which area estimates were required to be prepared. It is also quite probable that
an appreciable number of reported vacancies were not classified with that degree
of occupational precision. An evaluation of the precision of the occupational
classification of reported vacancies will be made, with particular attention directed
toward discovering possibilities for developing a sound system for reporting
occupational detail. One possibility may be the developm ent of a carefully
worked out precoded list of occupations with accom panying brief descriptions
that would be adapted to each industry but comparable among industries.
Estim ated area totals of jo b vacancies were prepared for each three-digit occu
pational classification for which job vacancies were reported. This was done by
inflating the area sample results to estimated area universe totals through the use
o f twenty-seven strata of industry categories and establishment size. Nine broad
industry categories and three establishment size groups were used to form these
strata. The inflation factor for each stratum was derived from the base period
total em ploym ent of the universe of establishments and of the responding estab
lishments falling into that stratum.
O P E R A T IO N A L A N A L Y S IS
An analysis of the results of the local Em ploym ent Service operational aspects
o f the jo b vacancy surveys will be prepared for each area. In the thirteen areas
in w hich follow -up jo b order development was undertaken, an examination and
evaluation will be made o f the effectiveness of the jo b order development and
placem ent efforts o f Em ploym ent Service local offices, by occupation, industry,
and size of establishment. This investigation will be carried out for both regularly
serviced establishments and establishments which have not placed jo b orders w ith
the local offices during the past twelve months. As another part of the opera
tional inquiry, applicants registered for work and unfilled jo b openings on hand in
Em ploym ent Service local offices in each of the sixteen areas will be analyzed in
com parison with the estimated total jo b vacancies for the area by occupation.
This study will reveal the occupational correspondence between all jo b openings
in the area and local office jo b openings and applicants. New training programs
suggested or developed as a result of the jo b vacancy surveys will also be reported
as part o f the operational analysis in each of the sixteen areas.
F O L L O W -U P
Q U A L IT Y
CHECK SU R V E Y
Follow-up response analysis surveys are being conducted by the Bureau of
Labor Statistics in six of the sixteen areas included in the experimental survey
program. The primary purpose of these follow-up surveys is to check and evalu
ate the quality of the inform ation reported by employers on the survey schedules
w ith respect to its accuracy, completeness, relevance, and reliability. In addi
tion, these response analysis surveys will probe into the feasibility of collecting
meaningful jo b vacancy information by mail with the voluntary cooperation of
employers. In this connection, inquiry will be directed toward the availability
o f records; the willingness and ability of business establishments to report with or
w ithout the benefit of records; the number of points or individuals in an establish
ment that must be contacted to account for all vacancies that may exist in the
establishment; the ability of respondents to report inform ation on certain addi
tional characteristics for vacancies which might be requested in the future; the
reliability and effectiveness with which respondents understood and interpreted
the reporting forms, definitions, and instructions; and the efficiency of the various
collection methods and instructions in obtaining the data desired.
Response analysis questionnaires are being completed by personal interview
in random ly selected subsamples of respondents to the survey im mediately after
the initial area survey ends. Trained and experienced BLS regional office per
sonnel are conducting the interviews.
Although rather comprehensive, the questionnaire is designed to facilitate the
interview by grouping the questions in general subject areas and incorporating a
skip pattern which avoids asking inappropriate and already answered questions.
A narrative evaluation report is prepared for each interview and at the conclusion
of the survey by each interviewer. The areas included in the response analysis
107
JOB VACANCY STATISTICS
surveys are: Providence, Charleston, S.C., Chicago, Los Angeles, New Orleans,
and Philadelphia.
A small random sample of nonrespondents to the job vacancy survey is also
being interviewed in the response analysis survey in the six areas covered. These
nonrespondents are being requested to complete a job vacancy schedule at the
tim e of the interview in order that an assessment can be made of the difference
or similarity in the level and nature of jo b vacancies as between nonrespondent
and respondent establishments. After a completed job vacancy schedule is
obtained, nonrespondents are asked the same questions as the respondents, plus
additional queries aimed at finding out the reason for their failure to respolin,
and the circumstances or conditions under which they might be able and wilngd
to cooperate in the future.
Response analysis surveys will be continued following the com pletion of the
second round of jo b vacancy surveys in the sixteen areas. It is expected that the
inform ation acquired through the response analysis surveys will be extremely
valuable in evaluating the feasibility of collecting accurate and meaningful
vacancy information, and in assessing the effectiveness of the various collection
schedules and methods being tested under the Departm ent of Labor’s experimental
job vacancy survey program.
JV-2
U.S. DEPARTMENT OF LABOR
Washington. D* C. 20210
Please enter data r e » #
quested.and return la *
accompanying envelope
by December-22.
Budget Bureau Nfr. 44-1M255
Approval Expires 6-30-65
JL 2 ,1. .4. 5 6 7 8
aEFoar on jo s vacancies
State o£ California
Department o f Emploient
Los Angeles Area O ffice
1S25 So* Broadway
‘ Los A n geles¿C alifornia 9001$
(Change Kame end Hailing Address I£ In correct)
Information reported w t K I s ' form
I s s t r ic t ly confidential» a n d .v ill
not be revealed to any unauttibrl*
*ed person nor published in such a
meaner that data relating to in di
vidual companies can be iden tified «
BEFORE flITESINS DATA PLEASE HEAP EXPLANATIONS ATTACHED
L
JOB VACANCIES. BY OCCUPATION* L ist below» by jo b t it le » a ll jo b vacancies
(as defined in attached explanations) in your establishment as o f the close
o f business on December 11» I f this is not possible» report vacancies fo r
the nearest possible day* A job vacancy i s a jo b opening that was u n fille d #
and immediately available to fu ll* » part-time» o r temporary workers which
your firm was actively seeking from outside your company* I f no vacancy»
wi a we
n u n it:
in
c u m y i e t e ’ jLwclnsi
w A u ;n n
-> a ii u
v*% o u u
ic w u r u
t il
Job Va<fancies
Occupational
Gode
(leave Blank)
Job T it le
(l)
(2 )
Number existinss
One work
One
Total
month or
week or
number loneer
longer
( 5)
(3 ;
(4)
Rate
«
o f oav offered
Indicate
whether
Amount per hour»
week» o r
month *
(7 )
'
(6 >
«
- .......................... ............................ ............................ ............................ ............................ .....................
«
o
<0
D
M
Continue on back o f fot•ffl i f mo
s needed#'
REFERENCE DATES Job vacancies reported above were as o f
( Date)«
c. «NUMBER OF’ EMPLOYEES z What was the to ta l number o f employees who worked durln£
/ o r received pay f o r any part o f the pay period which* Includes the 12th, ©£
December?
.
Signature ,_____________ :___________________T i t l e ________________
—
(Firm representative responsible fo r th is report#)
108
JOB VACANCY STATISTICS
A . JOB VACANCIES. BY OCCUPATION (Continued)
Job Vacancie*
Occupational
Code
(Leave Blank)
Job T it le
Cl)
(2 )
Total
number
(3 )
Kumber e x is tin g :
One
One work
month or
week or
lonser
longer
(4 )
(5 )
Rate
o f oav offered
Indicate
whether
Amount per hour»
veek, or
month
(7 )
(6 )
ARKS:
E x p la n a t io n s
fo r
R e p o r t on Job V a c a n c ie s
(Please read before entering data on report form)
A.
Job vacancies are defined as current, unfilled job openings in your estab
lishment which are immediately available for occupancy by workers from outside
your firm and for which your firm is actively seeking such workers. Included
are full-time, part-time, permanent, temporary, seasonal and short-term job
openings.
“ Actively seeking” is defined as current efforts to fill the job with a worker from
outside your firm through: (1) soliciting assistance of public or private em ploy
ment agencies, school or college placement offices, labor unions, em ployee groups,
business or professional organizations, business associates, friends and employees
in locating suitable candidates; (2) using “ help wanted” advertising (newspaper,
magazine, radio, television, direct mail, posted notice, etc.); (3) conducting re
cruitment programs or campaigns; (4) interviewing and selecting “ gate,” “ walk-in”
JOB VACANCY STATISTICS
109
or “ mail” applicants or workers searched out of applicant files; and (5) opening or
reopening the acceptance of applications from prospective candidates.
Do not include as vacancies 1) jobs held for employees who will be recalled;
2) jobs to be filled by transfer, prom otion, or dem otion; 3) jobs held for workers
on paid or unpaid leave; 4) jobs filled by overtime work which are not intended
to be filled by new workers; 5) job openings for which new workers were already
hired and scheduled to start work at a later date; and 6) those jobs unoccupied
because of labor-management disputes.
B. Reference date: Enter date for which vacancies are reported as Item B on
page 1. Enter date even though you report no vacancies.
C. Number of employees: Enter the total number of employees on all payrolls
of your establishment who worked full-time or part-tim e or received pay for any
part of the pay period which includes the 12 of December. Include persons on
vacation and sick leave who received pay directly from your firm for the pay
period reported, but exclude persons on leave without com pany pay the entire
period, and pensioners and members of the Armed Forces carried on the rolls
but not working during the pay period. Enter this figure even though you report
no vacancies.
Column 1: Leave blank. For office use only.
Column 2: List job titles for which job vacancies exist in your establishment.
Where possible, add modifier denoting material, product, process or subject
matter to make establishment title more specific, e.g., “ assembler, aircraft, wing
parts” or “ stenographer, legal.”
Use a single entry to report job vacancies with
identical job titles where more than one vacancy exists. If there are several
classes or grades for specific job titles, each class or grade should be listed
separately.
Column 3: For each jo b title listed in Column 2, report the total number of
jo b vacancies.
Column 4: For each jo b title enter the number o f jo b vacancies included in
Column 3 which have existed one work week or longer. I f none, enter “ N one.”
I f inform ation cannot be provided, enter “ X . ” D o not leave blank.
Column 5: Of the numbers shown in Column 4, enter the number of jo b va
cancies which have existed for 1 m onth or longer. If none, enter “ N on e.” If
inform ation cannot be provided, enter “ X .” D o not leave blank.
Column 6: For each job vacancy listed in Column 2, enter on the line opposite
that vacancy, the rate of pay offered for the jo b to which the jo b vacancy relates.
The entry of a single rate of pay is preferred; however, where a range of pay
rates is offered depending on the varying educational training and experience
qualifications of prospective applicants, the entry of the offered pay rate range
(that is, the low and high pay rates offered) wil] be acceptable. Wherever
possible, please enter hourly pay rates. If the wage offered for the opening
is on a piece work or commission incentive basis, please enter the estimated
average full-time weekly earnings the new worker is expected to receive.
Column 7: For each pay rate (or pay rate range) entered in Column 6, enter
the basis on which the offered pay rate is quoted in Column 6 (for example, in
dicate whether the pay rate given is per hour, per week, per m onth).
If you have any questions, please telephone 789-1124.
When form is completed, please return in the accom panying self-addressed
stamped envelope by D ecem ber 22. Please do so even though you report no
vacancy.
Thank you for your cooperation.
Chairman P r o x m i r e . This has all been very interesting. Since
Mr. Cassell had to leave and in the interest of not prolonging this
hearing unduly I am going to ask you to supply certain additional
material for the record.
(A copy of the letter sent by the chairman to Mr. Cassell follows:)
M a y 20, 1966.
Mr. F r a n k C a s s e l l ,
Director, U.S. Employment Service,
Department of Labor, Washington, D.C.
D e a r M r . C a s s e l l : Thank you for your very excellent presentations before
the Subcommittee on Econom ic Statistics on the feasibility and usefulness of job
vacancy data. We were impressed with the significant progress that has been
made in this area and with the operational uses that the data would serve.
63 - 947— 66 ------------ 8
110
JOB VACANCY
STATISTICS
The results of the pilot studies constitute very valuable pieces of information,
and the accounts of these studies were highly informative. It is m y feeling that
the more that is known about the experience of these studies, the better will be
the public understanding and the sooner the program can become operational
on a large scale.
Could you please supply for the record the following information and any
additional inform ation that you think would be helpful:
(1) Schedules used in the surveys of 1964, 1965, and 1966;
(2) The number of vacancies by occupation and by industry for each
area (and to the extent possible, cross-classified) ;
(3) The number of vacancies as a percent of the sum of em ploym ent plus
vacancies by broad industry and occupation for each area;
(4) Rates of entrance pay by occupation and percent of vacancies offering
compensation below this figure by area; and
(5) Examples attesting to the usefulness of the data, for example, from
businesses or local community organizations.
This inform ation would be of great interest and value both to the Subcommittee
and to others interested in an effective job market.
Sincerely yours,
W
il l ia m
P r o x m ir e .
(The material in response to the preceding questions, later supplied
by the Department of Labor, appears at p. 172.)
Chairman P r o x m i r e . The Committee will stand in recess until
10:00 o’clock tomorrow morning at which time we will hear from Mr.
Nathaniel Goldfinger, Director, Kesearch Department, AFL-CIO;
and Mr. Daniel Creamer, Manager, and John G. Myers, Senior
Economist, Special Projects Department, National Industrial Con
ference Board.
(Whereupon, at 12:33 p.m., the subcommittee recessed, to recon
vene at 10:00 o’clock, Wednesday, May 18, 1966.)
JOB VACANCY STATISTICS
W E D N E S D A Y , M A Y 18, 1966
C ongress of
S u b c o m m it t e e
of th e
U n it e d S t a t e s ,
E c o n o m ic S t a t is t ic s
J o in t E c o n o m ic C o m m it t e e ,
the
on
,
Washington D.C.
The subcommittee met, pursuant to recess, at 10 a.m., in room
S-407, the Capitol, Hon. William Proxmire (chairman of the sub
committee) , presiding.
Present: Senator Proxmire and Representative Curtis.
Also present: James W. Knowles, executive director; William H.
Moore, senior economist; George R. Iden, economist; Donald A.
Webster, minority economist; and, Hamilton D. Gewehr, adminis
trative clerk.
Chairman P r o x m i r e . The subcommittee will come to order.
Our first witness this morning is Mr. Nathaniel Goldfinger, director
of research, American Federation of Labor and Congress of Industrial
Organizations.
We are very pleased to have you, Mr. Goldfinger. You are an old
friend of the committee. You may proceed in your own manner.
STATEMENT OF NATHANIEL GOLDFINGER, DIRECTOR, RESEARCH
DEPARTMENT, AFL-CIO
Mr. G o l d f i n g e r . Thank you, Mr. Chairman.
I have a statement, which I might as well read.
At the outset, I should make it clear that while I obviously do not
object to research in the area of job vacancies, I do have serious doubts
concerning the state of our present knowledge about such data and their
underlying concepts. There are real questions about the validity of
job vacancy data—such as the basic issue of defining a vacancy.
And there is a danger that the U.S. Department of Labor may be
propelled into the collection and publication of a national series, long
before the questions of concept, validity, meaning, and usefulness are
adequately answered.
The fact that Western European countries and the United Kingdom
collect and publish job vacancy data is not particularly relevant to
the United States in 1966. It is my understanding that information
on job vacancies in those countries comes from the government em
ployment exchanges, with which the vast majority of employers list
their specific job openings, in their search for appropriate workers, with
the required skills, training, experience, and other qualifications. In
the United States, we do not have a Government employment ex
change similar to those in Western Europe and the United Kingdom.
I ll
112
JOB VACANCY STATISTICS
As a result, there is the attempt to establish a job vacancy series in
the United States, based upon surveys of a sample of employers. In
a sense, I suppose it is hoped that such data, derived from sample
surveys of employers, will somehow make up for the lack of an effective
public employment service in the United States.
But it seems to me that the attempt to rush headlong into a na
tional job vacancy index, as the obverse of the published unemploy
ment rate, involves an effort to substitute employers’ attitudes for
the actions of workers. The unemployment data, derived from the
monthly survey of a sample of households, is based on the reports of
specific actions of workers: an unemployed worker must report that
he is actively seeking work in order to be counted as unemployed.
However, a report that an employer believes he has a job vacancy
may be of questionable meaning—depending on such factors as
whether or not he is actively seeking to fill the reported vacancy and
what the wage rates and working conditions are in relation to those
for similar job classifications in the area.
As I view it, there is little, if anything, to be gained from the simple
addition of employer reports that x number of jobs are unfilled—
without adequate information on the specific occupations involved,
the skill requirements, and the wage rates and working conditions.
For example, I do not believe that the report of an unfilled job for a
50-cents-an-hour carpenter should be considered an actual job va
cancy. Moreover, the mere statistic—without information on occu
pation, skill, wages, et cetera—is hardly meaningful or useful.
In addition, I doubt that many employers keep records on job
vacancies. As a result, reports may tend to be at least partly im
pressionistic and the impressions may vary from one month to another.
They may also vary from one to another company executive. Fur
thermore, in the absence of records, information on the duration of
job vacancies will be difficult to achieve. And there are serious doubts
in my mind as to whether a report of a 1-day vacancy should be con
sidered an actual vacancy, at all.
Some economists have also pointed out that we do not know enough,
at present, about the durations of job vacancies and the movements in
job markets from unemployment to employment, in order to ascertain
some meaningful correspondence between unemployment statistics
and job vacancy data. Yet the publication of a national job vacancy
index will inevitably be compared with the unemployment statistics,
with inevitable interpretations and misinterpretations.
With all of these and similar difficulties, in terms of our present
knowledge about job vacancies, the construction and publication of a
national job vacancy series would be unwise.
It seems to me that a lot of w^ork is required, nowr, on concepts and
definitions and on economic meaning, as well as local job market feasi
bility studies and careful evaluation of such experimental work.
However, I believe that some meaningful use can be made of job
vacancy information on an occupational and local job market basis—
for operational purposes, to improve the operations of the public
employment service and the Government’s training programs.
The inherent difficulties in collecting any job vacancy information
would exist, even when developed as a working tool, on an occupational
and job market basis. But the shortcomings of such information
JOB VACANCY STATISTICS
113
would be of somewhat less significance than the great difficulties
involved in a national statistical series.
For operational purposes on an occupational and job market basis,
it should be possible to obtain and use meaningful details that could
be lost in a national series, whose emphasis would be on statistics,
rather than on improving the performance of the public employment
service and training programs for the benefit of workers, employers,
and the community. Details on occupations, skill and training re
quirements, and wage rate information for specific vacancies, in
specific firms and industries, would be essential to improve the effective
ness of the public employment service and the Government’s training
programs. Moreover, such detail can be checked on a job market
basis—particularly since detailed information would be essential for
operational purposes.
It seems to me that the development of such job vacancy informa
tion—as an operational tool for improving the efficiency and effective
ness of the public employment service and the Government’s training
programs—should be tested. But I believe that the development of a
national job vacancy statistical series—with its emphasis on some
kind of statistical average—would be an unwise step.
As the Federal Statistics Users’ Conference stated on March 24,
1966, before the Subcommittee on Labor, Health, Education, and
Welfare of the House Committee on Appropriations:
Further work on the resolution of conceptual and other problems should go
forward. The effectiveness of a regular jo b vacancy program as a tool for
promoting local labor market efficiency should be tested.
This approach is consistent with the BLS approach to im proving unem ployment
statistics, where proposed changes have been subject to experimentation on a
large scale over a period of years before introduction into the regular unem ploy
ment statistics program. It is consistent with the development of statistics of
em ploym ent by occupation which is proceeding over a period of several years.
Chairman P r o x m i r e . I think it would be desirable, Mr. Creamer,
in view of the fact that Mr. Goldfinger’s testimony is not on precisely
the same areas as yours—which I understand is largely a report on the
Rochester experience— that I might question Mr. Goldfinger at this
point and then proceed to hear your testimony, and that of Mr. Myers,
and then question you gentlemen.
Mr. C r e a m e r . Yes, sir.
Chairman P r o x m i r e . If Mr. Goldfinger would want to participate
in the discussion of Mr. Creamer and Mr. Myers, that would be very
helpful too.
Mr. Goldfinger, you are an extraordinary useful witness, not only
because you are very competent, but because you give a viewpoint
that we all too rarely have at this kind of hearing. So often at hear
ings we are scheduled to hear people who are unequivocally in favor
of whatever the committee is considering without any kind of criticism,
such as that which you so usefully and helpfully give us. I welcome
your thoughtful and helpful testimony.
We had some very strong testimony, in my judgment, yesterday,
from Mr. Cassell and Mr. Chavrid and Mr. Ross. Mr. Cassell who
is the Director of the U.S. Employment Service indicated in some de
tail the great pains—and I think they are quite extraordinary—that
the Department of Labor has gone into to make sure that this series
is sound, valid, accurate, and honest. Let me just read a page or so
from that testimony of yesterday.
114
JOB VACANCY
STATISTICS
Almost coincidentally with the release of the Gordon Com mittee report, the
report of the Illinois Governor’ s Com mittee on Unemployment which I [Mr.
Cassell] chaired was released in January, 1963. One of the recom mendations
of our report too, was for the initiation of a program to “ measure volum e and
com position of job vacancies and to report such information on a regular basis.”
I understand that this recommendation was one of the m ajor factors in the
selection of the Chicago area for the first feasibility study made by the USES
to test the possibility of collecting job vacancy data from employers.
This study was initiated in 1963, with the cooperation of the affiliated Illinois
Em ploym ent Service. Another feasibility study was conducted in the spring
of 1964 in Buffalo, sponsored locally by the New Y ork Em ploym ent Service and
the local Chamber of Commerce. The results of these preliminary studies were
highly encouraging. A t the same time, other research by the U.S. Em ploym ent
Service and the Bureau of Labor Statistics helped to define technical concepts
and m ethodology, and investigated the collection and uses of job vacancy in for
m ation in other countries.
A t the direction of Secretary of Labor W. Willard W irtz, and with the endorse
ment of the President, we began an experimental job vacancy program in the fall
of 1964 to test the possibility of collecting such data on a relatively wide scale,
and their usefulness for manpower operations. Primary emphasis was upon the
value of data for facilitating the operations of the jo b market, with a statistical
program as a possible by-product. The USES, in cooperation with its affiliated
State Em ploym ent Service agencies and the Bureau of Labor Statistics, conducted
tw o comprehensive jo b vacancy surveys.
It seems to me that this kind of practice testing of pilot program
operation represents a commendable effort on the part of the Depart
ment of Labor to determine whether or not this could be valid,
whether or not it could be practical, whether it could be the kind of
survey which would give you statistically meaningful results. And I
think it is somewhat unusual that the Department can go to these
pains and actually have pilot studies which can be examined and
considered over a period of years before they begin to gather the data.
* * * in each of 16 metropolitan areas representing a broad cross section o f
American industry in terms of size, industrial characteristics, geographical loca
tion, and nature of unem ployment. These areas include: Baltimore, Birmingham,
Charleston (S.C.), Charleston (W . Va.), Chicago, Kansas City, Los Angeles,
M iami, Milwaukee, Minneapolis-St. Paul, New Orleans, New York, Philadelphia,
Portland (Oreg.), Providence, and Richm ond. In the aggregate, these areas
account for approximately one-fourth of the N ation’s total nonfarm em ploym ent—
which indicates that it is not only a useful pilot study from the
standpoint of sampling but also that it is fully comprehensive.
The first round of surveys was conducted during the last quarter of calendar
year 1964 and the second survey took place In April, 1965.
And then Chavrid goes into a detailed discussion, and he made his
presentation yesterday, and it is a detailed, and, I think, highly
competent analysis of the problems that are very involved, and the
results. And he comes to the conclusion that these statistics would
be accurate, would be honest, would be valid.
I think it would be helpful to have your comment on Mr. Ross’
reasons why this job vacancy information would be so helpful. He
said, first, that it can be used to develop a picture of the size and
characteristics of unfilled demand for labor. Then information could
be analyzed in its own right.
Secondly, a trend in job vacancy especially if classified by occu
pation can be of considered value in throwing light on the ability of
our economy to adjust to changes in demand for labor, which may
serve as a lead indicator for changing economic conditions.
Third, job vacancy information used in conjunction with informa
tion on employment, unemployment, labor turnover, and hours of
JOB VACANCY STATISTICS
115
work, can enhance our ability to analyze the current economic situa
tion, for light on major policy decisions.
Fourth, in the present economic situation the question of labor
shortage has become sufficiently critical, especially in relation to
skilled manpower, that the President, as previously indicated, has
asked the Department of Labor to watch the situation closely and to
prepare regular reports. And, of course, the reports are much better
if they are not based on this kind of a vacuum determination.
Fifth, job vacancy information will throw additional light on
demand-supply conditions in the job market in relation to changing
wage levels. And analyses of the effect of employment changes upon
wage rates, although potentially very useful in appraising wage de
velopments and policy, has not exhibited highly precise results when
applied to data available for the United States. The additional
dimension of job vacancies in the measurement of labor demand would
contribute another powerful tool of analysis.
Also, job vacancy data can help us to sharpen the Bureau of Labor
Statistics projections of manpower requirements by occupation which
are so essential in developing estimates of training needs to guide in the
planning of the many education and training programs supported by
the Federal Government. These programs are the most immediate
pragmatic and practical results of this kind of thing that we can or
ganize our training programs in a way that is less wasteful and more
efficient than it could be without it. Congressman Curtis yesterday
estimated that the investment of $2.5 to $5 million in this program a
year would bring back returns many, many times that every year in a
more efficient training program based on a factual knowledge of what
jobs would be available.
And then job vacancy information would be used by business firms
to get an accurate, reliable picture of the area in which they are re
cruiting workers, and to help in developing more effective recruiting
policies.
And finally—and this is the area that directly touches on our own
area of interest— “ Such information could be of equal value to labor
organizations in evaluating the demand for the services of their mem
bers and in developing policies for training, apprenticeship, and col
lective bargaining.”
I have asked a long question. But I see you had a chance to take
some notes on it.
Mr. G o l d f i n g e r . I will attempt to comment on that, Senator
Proxmire.
There are two different approaches at the outset. One is the sta
tistical approach of moving immediately into some kind of national
statistical series on job vacancies. We in organized labor have
questioned the wisdom of moving now into such national statistical
series, for the various reasons that I indicated in this brief paper.
And there are many additional related reasons.
We simply do not know enough at this point about the basic con
cepts—-about the definition and other related matters.
For example, such statistics would immediately be used for inter
pretation of economic trends and economic development.
And yet w^e do not know at this point, as many economists have
pointed out, we do not know in terms of American society and the
American economy what the proper relationship is between job
116
JOB VACANCY STATISTICS
vacancy statistics and the unemployment rate. What is the meaning
of a comparison between a 4-percent job-vacancy statistic and a
4-percent unemployment rate? Moreover, what is the meaning of
a 4-percent job-vacancy statistic?
Chairman P r o x m i r e . We cannot find out if we don’t try it, so we
can make the comparison.
Mr. G o l d f i n g e r . I was going to get to the other side, and that is
to start on a job market basis, with the kind of detail that probably
w'ould get lost on a national statistical basis, where we could get oc
cupational information, industry information, skill requirements, wage
rate information, working conditions, and so forth, and test the job
market approach over a period of time, and begin to refine concepts,
definitions, and to begin to work out some of these numerous prob
lems. Everybody admits that there loads of such problems that
require a considerable amount of work.
Chairman P r o x m i r e . There are problems. But is it not true that
in the last 3 or 4 years there has been quite a bit of intensive work on
these problems, and that some of them have been resolved? For
example, you ask for a definition of job vacancy. And I think that
is certainly an understandable demand. Mr. Chavrid and Mr. Ross
and Mr. Cassell seem to feel that the definition was clear, and at
least as valid as the definition of an unemployed person. If the em
ployer indicates that he has a vacancy, that he wants to employ a
person, they feel that that would be as valid as going around, as we
do with our unemployment survey, and asking a person if he is
looking for work.
Now, they have tried to check this in various ways over a period,
as Mr. Cassell indicated, of a couple of years. And on the basis of
their findings, they feel that there was an underestimate of job
vacancies of 12 percent, which is a substantial error, perhaps. But
they have been able to doublecheck it, work with it, discuss the con
cept, and to consider, for example, the legitimate point that you
raised about a “ 50-cent-an-hour carpenter.” It was your fear, and
mine too, and I raised that question yesterday—if employers say they
would like to hire a certain number of workers, but they will pay them
far below the going wage, obviously the job vacancy would not be
legitimate under any fair basis. The Government witnesses say
that their study has convinced them that this is not a significant
problem, that they were concerned about it, too. They have checked
and found that vacancies are overwhelmingly in the going wage area,
and that there is not any attempt on the part of anybody—and it
would be very easy to determine if there were—to give phony sta
tistics by indicating that you would like several thousand workers if
you can get them for far less than the going wage.
So that these problems, I think, have been worked on. This area
that you say should be developed, that is what they have been doing,
as I understand it, last year.
Mr. G o l d f i n g e r . They have begun to do that, Senator. They
did engage in a number of pilot studies. To my knowledge there has
not yet been any careful outside evaluation of the results of those pilot
studies. What they told you yesterday may well be true. But I
have some doubts about that, because I am just not sure.
Furthermore, I wonder whether they have used the results of these
pilot studies for operational purposes. To what extent have they
JOB VACANCY STATISTICS
117
tested them? To what extent have they used them in connection
with the operations of the Employment Service? To what extent
have they used them in conjunction with the operations of the MDTA
training programs?
Furthermore, Senator-----Chairman P r o x m i r e . Let me say that I do think there have been
some independent studies, the National Industrial Conference Board
study, the Fisher and Ford study of 1963, independent studies.
Mr. G o l d f i n g e r . Yes. But these are all separate studies, in
dependent studies. What I am referring to is a careful-----Chairman P r o x m i r e . And also may I say that Mr. Creamer and
Mr. Myers have been involved in what has won national recognition
as almost a model study in Rochester on a foundation grant of job
vacancies.
Mr. G o l d f i n g e r . Yes; all of this is true. But I am not aware of
any careful outside evaluation, for example, of the Labor Depart
ment’s pilot studies.
Perhaps they have been made, but I am unaware of any such
evaluation.
Furthermore, they have had these pilot studies, I think you men
tioned, of 16 areas. To what extent have they used the results of
those pilot studies for operational purposes in connection with the
operations of the Employment Service and the operations of the
training programs?
Chairman P r o x m i r e . I understand— and perhaps Mr. Knowles can
correct me if I am wrong—as I understand it, job vacancy data could
be used and evaluated in preparing the reports on labor shortages,
which Commissioner Ross is asked to make regularly to the President
of the United States. And this is an area in which they have to make
estimates anyway. The pilot studies would permit them to refine
this. And one of the things they have been able to do, for example,
would determine that their USES estimates represented a fairly stable
and quite accurate representation, although limited, of the amount of
job vacancies, by making a more comprehensive study. They have
also been able to find that they receive very good cooperation from
employers, and a very high proportion of those who were requested to
reply did, especially in Milwaukee and Portland, Oreg. And it was
generally true.
Mr. G o l d f i n g e r . I am aware of the reports of the relative success
of these pilot studies. And I have glanced through the NICB report,
which Mr. Creamer kindly sent me some days ago. And I am aware
of the various bits and pieces of work that have been done within
the past number of years. However, I am suggesting that we are
a long way from resolving a lot of these very basic questions which
have been raised, including a lot of the basic questions which were
raised on this subject in the book that you just picked up by the
National Bureau of Economic Research Conference.
Chairman P r o x m i r e . I think this book itself— “ The Measurement
and Interpretation of Job Vacancies,” Conference of the National
Bureau of Economic Research—is a most impressive book, because
it has articles by a number of outstanding authorities, widely recog
nized economists and experts, and specialists in this area. And the
whole 560 pages is devoted to an analysis of this problem, and an
evaluation of it. And while there are not any joint collective recom
118
JOB VACANCY
STATISTICS
mendations, the general tone of this, as I understand, as well as the
conclusions of the three Government witnesses yesterday, is that we
are ready to move. In fact, they asked for an appropriation last year,
as you recall, and it was $2.5 million, and it was turned down by the
House Appropriations Committee. So they were ready last year to
move.
Mr. G o l d f i n g e r . That is a matter of judgment. It may be a
matter of their judgment and our judgment differing. I might also
indicate that the judgment of the Federal Statistics Users’ Conference
differs, to some extent, from the judgment of some of the people in
Government and in the universities about our ability, at this point,
to move rapidly into this area of job vacancy statistics.
Chairman P r o x m i r e . Just one more point I would like to make with
you.
I think you raised the question of the 1-day vacancy.
Mr. G o l d f i n g e r . Yes.
Chairman P r o x m i r e . It seems to me that this is not symmetrical
with unemployment, inasmuch as if a man has had a job within the
preceding week he is not considered unemployed, so he has to be
unemployed 7 days before he is counted as unemployed. And the
job vacancy is counted on the basis of these pilot studies, at least,
immediately. So you would not get a precise comparison. I think
that is a legitimate complaint. The witnesses yesterday indicated
that they would not make any difference in the statistics, because the
overwhelming majority of job vacancies have been available for more
than 7 days. I think if you are going to get symmetrical and com
parative statistics you should have them on the same basis. And
my own conclusion is that you ought to have a requirement of 7 days
for job vacancies just as you have for unemployment.
Mr. G o l d f i n g e r . On this one I would agree with you, sir. But
my interest here is not in the symmetry of the proposition. I think
that what we should be trying to do is to get at the meaning of a job
vacancy. If we are going to measure job vacancies, we should seek
real genuine vacancies, whatever that means. We first have to
define it. And there are difficulties in definition, as the NICB
report indicates. Moreover, there certainly are great difficulties in
getting information on the duration of job vacancies. Most com
panies do not have records on job vacancies. A lot of the information,
unfortunately, at this point is impressionistic. Perhaps over a period
of time it would be possible to get employers to keep some kind of
records on job vacancies. I am not sure that this is possible, but
perhaps it is.
But I think that there is some force that is pressing hard for the
immediate movement into a national statistical series. And it is this
that I believe is definitely unwise at this point in terms of the state of
our knowledge.
Now, if you are talking about local job market studies, where we test
out these various concepts and definitions and problems and attempt
to work them out, and particularly work them out for operational
purposes in conjunction with the MDTA and the public employment
service, that is one thing; and that I think is worth doing. On the
other hand, I do not think that we know enough at this point to move
ahead headlong into a national statistical series.
Chairman P r o x m i r e . Well, what concerns me is that unless we do
move I am afraid that we are going to deny ourselves some accurate
JOB VACANCY STATISTICS
119
and valid data which could help us in economic policy determination.
I think we are always better off when we have more rather than less
information. And it would seem to me that one of the ways that we
can proceed to improve this concept is to give it a chance to work.
And while you can argue that you have to improve it before you
really gather it, or before you report it, it would seem that these very
comprehensive studies that we have had have served that purpose.
Now, in addition, I think that we have had great improvement and
refinement of our unemployment statistics, since we have been report
ing the data. At the same time some legitimate questions have been
raised about the legitimacy of unemployment by labor and manage
ment and by others. But we are always trying to refine it and improve
it and get the additional reporting of unemployment so that we under
stand it better.
The same thing would be true in job vacancy. I would concur.
And as I said yesterday, I would be very fearful that some uninformed
people would argue that you can take the unemployment figure,
subtract the job vacancy figure from it, and come up with something
that would mean anything, or a net unemployment figure. It would
not mean anything at all, and I think the experts yesterday all agreed
that it would not mean anything, that you would be subtracting
job vacancies—a demand for several thousand nurses, for example from
unemployed—when perhaps none of the unemployed could ever
qualify as nurses. You just cannot on that basis get anything that
is valid.
But you can get something that would be of immense help in so
many areas of our operations. And I would say that there is no
single effort on the part of the Government which is more crucial
than education and manpower training, and womanpower training.
And these data would help us so greatly to refine it, improve it, and
make it more effective, that I would be very reluctant to see us
delay because of a fear of misinterpretation—no matter when you put
this into effect, it is going to take a few years before you develop a
degree of sophistication and a degree of understanding on the part of
the public. Admittedly, there would be some abuse of this, as there is
abuse of all statistics. But it would seem to me that the best way to
overcome that abuse is to try it, step in and use it, and then refine it
further as you go along, recognizing that we have gone as far as we
already have trying to come up with valid concepts and definitions.
Mr. G o l d f i n g e r . Why not do that, sir, over a period of time on a
job market basis, and work out a lot of these problems before you move
into the national statistical series? Furthermore, you mentioned that
this would be very valuable information. Now, I agree that it would
be valuable to get this kind of information if we had it. But there are
losts of other kinds of economic data and economic information which
are lacking.
For example, we in organized labor have been after the Bureau of
Labor Statistics for years now to update and refine the city worker’s
family budget. I believe that at long last they are finally doing
something on that.
We in organized labor have been after the Bureau of Labor Statistics
for years, and probably for decades, to obtain information on the
annual earnings of workers by industry and occupation. There is no
such information available in the United States.
120
JOB VACANCY STATISTICS
And there are all kinds of economic data which are needed. There
are all kinds of economic information in which there are gaps. The
lack of information on job vacancy statistics is not the only gfip.
Chairman P r o x m i r e . The Commissioner of Labor Statistics, Mr.
Ross, I think we would agree, is a competent expert in the field. And
he said yesterday:
The lack of job vacancy inform ation constitutes the most significant gap in our
knowledge of labor market conditions. Statistics on job vacancies would give us a
measure of unsatisfied demand for labor which, together with our data on em
ploym ent, would provide a more complete measure of the demand for labor—
something we have never had before.
Now this, coming from Mr. Ross, with the support of Mr. Wirtz,
two extraordinarily able and thoughtful men with no ax to grind on
any side that I know of, must carry a lot of weight.
Mr. G o l d f i n g e r . I agree that there is a gap in this area, Senator.
There certainly is a gap. And I think we should begin to move toward
filling it on a practical basis, of starting out in the development of this
kind of information as an operational tool. We in organized labor,
however, are much less interested in the statistics as such. We are
less interested in the numbers. We are interested in the development
of information that would be of assistance for the operation of the
employment service, and for the effective operation of the training
programs—information that could be used to help the unemployed.
Now, for example, if we wished to get the kind of information that
you are talking about, for which there is this gap, there are other ways
of doing it too. For example, we could develop an effective nation
wide and computerized public employment service, and probably
produce this kind of information as part of the operations of a compu
terized national public employment service. Why do we not do that?
Chairman P r o x m i r e . It would cost more than $2.5 million, would
it not?
Mr. C r e a m e r . Senator, would it be in order to enter the discussion
at this point?
Chairman P r o x m i r e . I think this would be a good time to have a
transition to Mr. Creamer. I see that the lead subtopic of your testi
mony is “ The Informational Gap.” Why don’t we move into your
testimony. Mr. Goldfinger is free to comment any time he wishes.
And you go ahead and answer any of the questions we have raised
or go into the testimony, any way you wish.
STATEMENT OF DANIEL
ECTS
DEPARTMENT,
CREAMER, MANAGER, SPECIAL PROJ
NATIONAL
INDUSTRIAL
CONFERENCE
BOARD, ACCOMPANIED BY JOHN G. MYERS, SENIOR ECONOMIST, SPECIAL PROJECTS DEPARTMENT, NATIONAL INDUS
TRIAL CONFERENCE BOARD
Mr. C r e a m e r . I would like to start out by offering one comment
on what I think is the main thrust of Mr. Goldfinger’s objection to
the Department of Labor’s proposed program. He says it should be
concerned—and I think Mr. Myers and I would concur in this—•
that it should deal with individual labor markets. But that is exactly
the proposal of the Department of Labor. They have asked for an
appropriation to extend their surveys from 16 to 75 or 80 labor areas.
JOB VACANCY STATISTICS
121
And they want this by particular labor areas, for the very reasons Mr.
Goldfinger enumerated.
Now, if it turns out that by adding up the figures for the 75 areas
you get something that is representative for all urban United States,
I am sure Mr. Goldfinger won’t object to that addition.
But their primary interest, as we heard yesterday, is in these pro
gram units.
Chairman P r o x m i r e . That is correct. But isn’t it also true that
the 75 labor market areas would be used as a national figure at least
on the basis of a projection or estimate or something to provide for-----Mr. C r e a m e r . It may or it may not. But this would be a spill
over. It is not the only use or the primary use.
Chairman P r o x m i r e . But at any rate, what you are telling us is
that this would not be reported as a national figure for job vacancies,
this would be reported simply as a figure for the job vacancies in the
75 markets which represent a certain proportion of the Nation’s popu
lation, and of the Nation’s employment?
Mr. C r e a m e r . I would hope so. I cannot speak for them.
Chairman P r o x m i r e . And if anybody wants to make a projection
on the basis of that, they are free to do it, but they could also try to
make a projection on the basis of the 16 areas from which they report
now, is that right?
Mr. C r e a m e r . Yes.
I think Mr. Goldfinger does have a legitimate complaint on the
point of not having seen the results of their surveys in 16 areas. To
my knowledge the first time this would be made public is in Mr.
Chavrid’s testimony that he presented yesterday. They have been
very cooperative with us. And they did make a summary available
to us. But it has not been made public.
I suspect one reason—and this is a supposition—why they have not
done this is that they have been a little fearful of what Mr. Gold
finger’ s organization would say about it, the doubts that he raises in
the very beginning of his statement, of rushing headlong into a na
tional series and not carrying out all this preliminary investigation
carefully.
I think they have been overly timid in making this available for
outside scrutiny.
Perhaps Mr. Goldfinger has now dispelled this fear.
Senator P r o x m i r e . Intimidate the Secretary of Labor.
Mr. G o l d f i n g e r . I am all together unaware of that.
Furthermore, if I may break in at this point, which I already have
done, I don’t know of one single article that has appeared in the
Monthly Labor Review, published by the Bureau of Labor Statistics
on the concepts of job vacancies, on the definition of job vacancies, or
on the possible meaning of such statistics. In fact, I do not know of
any continuing work by the Bureau of Labor Statistics on concepts,
definition, and evaluation of the pilot studies, in terms of concept and
definition. Frankly, I do not believe that they have defined for
themselves and explained to the public what they want in the area of
job vacancy statistics.
These are some of the problems. As I see it, this is not merely a
statistical issue. I have no interest—and organized labor does not
have any interest—in the mere creation of another statistical series,
another set of numbers. It is the meaning of the numbers which is
122
JOB VACANCY
STATISTICS
important, and the usefulness of the numbers. And unless we get a
meaningful, useful series of figures—of tangible usefulness to workers,
employers, and the community— then I think we are just wasting the
taxpayers’ money.
C h a i r m a n P r o x m i r e . I do not think you and Mr. Creamer are
very far apart.
Mr. C r e a m e r . N o; we are n o t.
Is a question in order to Mr. Goldfinger?
Chairman P r o x m i r e . Certainly, by all means.
Mr. C r e a m e r . In view of this understanding of what the Depart
ment of Labor is proposing, has organized labor given its endorsement
to the House Appropriations Committee for this appropriation?
Mr. G o l d f i n g e r . We have neither endorsed it nor opposed it.
Mr. C r e a m e r . But in view of this further elaboration, is there a
chance of endorsement?
Mr. G o l d f i n g e r . No, we neither endorse it nor oppose it. I have
tried to indicate the numerous problems in this area of job vacancy
statistics. In addition we have a real, vital interest and a vital stake
in other kinds of economic information, and in other kinds of eco
nomic data which have not been forthcoming from the Bureau of
Labor Statistics. Unfortunately we live in a world with limited
resources and there are limited appropriations.
As I said, we have been knocking on the door of the Labor Depart
ment for years to update and refine the city worker’s family budget,
which I believe is at long last being done. But we have been at the
door of the Labor Department for years, and perhaps decades, asking
for information on annual earnings of workers, which we do not have.
Job vacancy statistics are far from our top priority need in 1966.
Chairman P r o x m i r e . Y o u reject the argument that was made by
Mr. Ross yesterday that this would be very useful to organized
labor?
Mr. G o l d f i n g e r . It may be useful to some extent.
I can also see some bad results coming about as a result of mis
interpretation. The probability of misinterpretation is very great,
particularly in the absence of occupational, skill, educational, wage
rate, and similar detail, on a job market basis. And I think that-----Chairman P r o x m i r e . But labor has nothing to fear from the truth.
Mr. G
o l d f in g e r .
No, we have nothing to fear from the truth.
Chairman P r o x m i r e . To the extent that this is not an accurate
series, I think the case for labor’s objections is enormously strong.
But everything I have seen in my public or private career has per
suaded me that labor benefits from getting the fullest possible statis
tics and the fullest knowledge of what the job situation is, and em
ployment and unemployment, and so forth.
Mr. G o l d f i n g e r . I agree w ith y o u , sir.
Chairman P r o x m i r e . Why wouldn’t the development of 75 instead
of 16 markets, and the revealing of both these findings, so that there
would be greater public knowledge, why wouldn’t this-----Mr. G o l d f i n g e r . Well, for example, why haven’t the results of
the pilot studies to date been subject to a very careful evaluation be
fore moving ahead onto a much bigger scale, to a national statistical
index? It is this kind of rapid moving ahead into large-scale opera
tions that we question, with the very great danger of misinterpretation
JOB VACANCY STATISTICS
123
for the purpose of slowing down or cutting back Government expendi
ture programs.
Chairman P r o x m i r e . Mr. Creamer, do you want to go ahead?
Mr. C r e a m e r . Yes, thank you, Mr. Chairman.
A.
THE
IN F O R M A T IO N A L
GAP
Many of the economic problems of the first half of the current decade
that have claimed the attention of economists and statisticians had
their origin in the persistent high unemployment rate in a period of
sustained business expansion. Measuring job vacancies is one of these.
As the evidence of high level unemployment accumulated some were
prone to challenge the accuracy of the official count of the number un
employed. To evaluate these criticisms, President Kennedy appointed
a Committee of eminent economists and statisticians not in the
Government service to appraise employment and unemployment
statistics. The Committee was organized in November 1961 and one
year later submitted its report, “ Measuring Employment and Un
employment” (usually referred to as the “ Gordon Committee Re
port,” after its chairman, Prof. Robert A. Gordon). While the Com
mittee unanimously concluded that it “remains highly impressed by
the professional qualification and the scientific integrity and ob
jectivity of those responsible for the system of reporting the official
data on employment and unemployment,” 1 it also detailed ways of
improving the information on employment and unemployment.
High on the list of recommendations was one for further research
on the feasibility of initiating a program on the measurement of job
vacancies. The Committee reached this conclusion after noting
that—
It is doubtful that any suggestion for the im provem ent o f knowledge about the
nation’ s labor markets was more frequently voiced to this Com m ittee than that
calling for job vacancy statistics.2
Why had the Department of Labor not previously responded to
this interest? This is traceable in part to an experience of the Bureau
of Labor Statistics (BLS) in the mid-fifties. In response to the urging
of the then chairman of the Council of Economic Advisers, Prof.
Arthur F. Burns, the BLS actually made a survey in 1956 to determine
whether employers maintain records that would be adequate for
reporting job vacancies. Of the 102 plants surveyed, 29 had formal
records of the number of job vacancies and another 54, without formal
records, could give an estimate based on personal knowledge. Al
though 80 percent of the respondents on the first inquiry could provide
the information, the BLS staff members concluded that:
. ..
it would be impractical to initiate a regular mail collection of statistical
data on job vacancies. D ata resulting from such an attem pt would certainly not
be comparable in quality to that obtained in our related statistical programs in
the manpower and employm ent field * * *. The principal difficulty is that
employers do not keep records of vacancies comparable in accuracy or detail to
their payroll records.3
With the benefit of hindsight, the conclusion seems based on inap
propriate criteria and therefore unwarranted. That is, a statistical
1 Op. cit., p. 3.
Ibid., pp. 26 and 201-2.
s Ibid., p. 279.
2
124
JOB VACANCY STATISTICS
series at the outset cannot be expected to meet the quality standards
of a long-established series. Moreover, it seems to discount the
lessons of previous survey experiences that employers do develop
adequate records once they become aware of the usefulness of the
information. Skepticism about the usefulness of job vacancy data
on the part of some sections of organized labor probably also con
tributed to the Department's reluctance in recent years to take the
lead in some exploratory fieldwork.4
At any rate, in fiscal year 1963-64 the BLS was committed only to
study foreign experience in the collection and use of job vacancy
statistics and to analyze the conceptual problems in defining a job
vacancy. I might interpolate here that BLS had kindly made avail
able to us a very fine and extended discussion of the conceptual and
definitional problem. But here again—for what reason I do not
know, Mr. Goldfinger is right, they have not published it—but they
do have this material, and have gone through the analysis in a very
competent and diligent way. The Department's Bureau of Employ
ment Security (BES) had no concrete plans for survey work in this
field for that year.
In these circumstances, the Conference Board took the view that
more progress could be made by a direct attempt to measure job
vacancies in a significant labor market.
And this proposal was first made by Mr. Gainsbrugh, the chief
economist and senior vice president of the Conference Board, in
testimony given to this subcommittee in the spring of 1963, when it
was holding hearings on the Gordon Committee report. And it was
as the result of Senator Proxmire's sympathetic interest to that pro
posal that the Conference Board did go ahead late in 1963 and sub
mit an application to the Ford Foundation to support an effort of
actual measurement in the labor area of Rochester, N.Y. The
Ford Foundation acted favorably on the Board's request in the
spring of 1964 and the exploratory project was begun in the second
half of June 1964. At that time no other full sample survey was
contemplated.5
This solitary position was soon overwhelmed by the mounting pres
sures on and in the Department of Labor for job vacancy data. In
addition to the persistence of high level unemployment was the grow
ing realization that the implementation of many parts of a fully articu
lated manpower program was severely handicapped by the lack of job
vacancy data by occupation. Superimposed on the operational needs
was the expectation that a continuing series on job vacancies would
contribute to a resolution of a policy debate. This was between the
“ expansionists” and the “ structuralists” — that is, between those who
argued that excessive unemployment should be attacked by raising
aggregate demand and those who argued that it should be reduced
by correcting structural imbalances.
4 For evidence of an essentially negative attitude, see discussions by Marvin Friedman, A F L-C IO De
partment of Research and Nat Weinberg, United Automobile, Aerospace & Agricultural Implement
Workers of America, in “ The Measurement and Interpretation of Job Vacancies,” a conference of the Na
tional Bureau of Economic Research (Columbia University Press, 1966), pp. 132-137 and pp. 463-476
respectively.
5 Late in 1963, Profs. Robert Ferber and Neil Ford of the University of Illinois initiated a pioneering
survey of pilot proportions, in Champaign-Urbana, 111. Job vacancy and related turnover data were col
lected monthly between October 1963 and May 1964 from 17 employers. The results have been presented
in two articles: “ The Collection of Job Vacancy Data Within a Labor Turnover Framework,” in Employ
ment Policy and the Labor Market, Arthur M . Ross, ed., Berkeley, 1965, pp. 162-190 and “ The Time D i
mension in the Collection of Job Vacancy Data” in the Measurement and Interpretation of Job Vacancies,
op. cit., pp. 447-461.
JOB VACANCY STATISTICS
125
Presumably these were some of the major considerations that per
suaded the Secretary of Labor in August 1964 to instruct the appro
priate agencies in his Department to make a start in measuring job
vacancies. The Bureau of Employment Security and the Bureau of
Labor Statistics jointly designed a survey and selected 16 labor areas
for the experimental program. Thus the Conference Board's survey,
which appeared to be a solo effort when it was launched, soon became
1 of 17 surveys. This much more solid experimental base, is, of
course, most welcome.
B.
P O T E N T IA L U S E S F O R E C O N O M IC P O L IC Y A N D M A N P O W E R P R O G R A M S
What are the potential uses of job vacancy statistics that persuade
analysts to embark on the rather arduous task of exploring the
feasibility of measurement?
The uses that should be considered are those that would follow
from a continuing collection of job vacancy statistics. A single
survey or one that is repeated only once or twice can do little more
than serve to demonstrate whether a collection program is or is not
feasible. Still another distinction is required. The uses of the sta
tistical output of an on-going program will be of one sort if the sta
tistical program is designed to provide only national totals with
occupational detail; there will be other uses if the program yields job
vacancies by occupations for each of the larger labor market areas
as well as for the Nation. The former is examined first.
Analytical and policy uses of job vacancy statistics
The principal analytic problem for which job vacancy data are
needed is the determination of the most appropriate policy, or balance
between policies, to reduce unemployment without at the same time
causing inflation. The balance between the demand for and the
supply of labor, and particularly movements in this balance, will help
to indicate whether aggregate expenditure can be increased to reduce
unemployment without causing inflationary pressure from rising
wage rates.
The demand for labor at a given level of wages and at a point in
time can be conveniently divided into two parts, the “ satisfied de
mand," represented by the number employed, and the “ unsatisfied
demand,” represented by the number of persons employers wish to
add to their payrolls but have not succeeded in doing so. Similarly,
the supply of labor at the same wage level and instant is composed of
the “satisfied supply,” those employed, and the “ unsatisfied supply,”
those seeking work. The “satisfied” portions of demand and supply
are identical, by definition, since they are the number employed. We
can therefore concentrate on the other components.
The number of job vacancies, appropriately defined, can furnish a
measure of the unsatisfied, or excess demand for labor, while the
number of unemployed can represent the unsatisfied, or excess supply.
The extent to which the two measures offset one another indicates
how smoothly the labor market is functioning. The greater the
number of unfilled jobs matched by unemployed persons, the poorer
the adjustment achieved by the labor market. The simultaneous
existence of unfilled jobs and unemployed workers arises from a
variety of reasons, such as the migration of industry and of persons,
63 - 947 — 66 ------------ 9
126
JOB VACANCY
STATISTICS
technological changes, demographic factors, et cetera. A convenient
term to cover the problem is “ maladjustment in the labor market.”
Without job vacancy data, neither the amount of maladjustment pre
vailing in the economy nor movements in this amount are easily
determined.
While maladjustment is measured by the smaller of the number of
vacancies and the number of unemployed, at a given level of pros
perity,6 the difference between the two indicates "»which factor is
predominant, demand or supply. For clarity of discussion, let the
difference be consistently measured by the number of vacancies less
the number of unemployed, which can be positive or negative. If
this measure is negative, that is, there are more unemployed than
vacancies, and particularly if it is falling (i.e., toward larger negative
values), this suggests that an increase in aggregate expenditure will
reduce unemployment without undue inflationary pressure coming
from wage rates. If the measure is positive and particularly if it is
rising, the implication is that manpower policies, designed to increase
mobility and to provide training, and increased efforts to improve the
placement of unemployed persons are the more appropriate policies
to reduce unemployment.
Chairman P r o x m i r e . I hate to interrupt you, because there are a lot
of things to discuss, but at this point I cannot resist. For my own
clarification: you are not saying that you get anything if you subtract
from the number of people out of work— the number of vacancies?
Because if you say that, as I understand it, you are not making any
discrimination as to the kind of vacancies. You might have an enor
mous number of vacancies for engineers, draftsmen, people in that par
ticular category, and you might have a relatively modest unemploy
ment, but none of the unemployment might be in any position to
qualify as engineers unless they had more years of education than they
are ever going to get—or as draftsmen. Or you might have a situa
tion in which you have in one part of the country—if you are talking
about a national survey' here—in one part of the country you might
have a vacancy, and in another part of the country you might have a
surplus of labor. And you do not really get a very good picture of the
situation if you are going to try to get a national figure, because you
are just not going to get people to go from here to Oregon to work, or
vice versa, or it will be very difficult.
Mr. C r e a m e r . But, Senator, isn't it important to know the size of
this maladjustment? Now, if there is perfect mobility between re
gions and between occupations, and so on, presumably there would be
this complete matching. But we know this does not happen.
Now, it seems to us that it is important to know the size of the
maladjustment, and more particularly, is this maladjustment growing
or diminishing? If it involves a small number of persons, well, then,
perhaps in due course, the usual operations of the labor market will
take care of it. If it involves large numbers, and it keeps growing
rather than declining, then it seems to me that this may very well
o With the same degree of maladjustment, a rise in aggregate demand will draw persons into employment
who were unemployed owing to friction, structural imbalances, etc. Thus the number of vacancies offset
by unemployed will vary inversely with the level of aggregate demand. While the statement is correct
that the degree of maladjustment is shown at the same level of prosperity (or aggregate demand) by the
smaller of the number of vacancies and the number of unemployed, a practical measure requires some as
sumption about the substitution of unemployment for vacancies as aggregate demand is increased. Dis
cussion of the concepts of maladjustment and excess demand may be found in J.C.R. Dow and L.A. DicksMireaux, “ The Excess Demand for Labour: A Study of Conditions in Great Britain, 1946-56,” Oxford
Economic Papers (N.S.), February 1958, pp. 1-33.
JOB VACANCY
STATISTICS
127
signify, or the legitimate inference is, that perhaps we need some new
policies to deal with it. And this is, then, in our view a critical indi
cator of whether it is time to do something about it, and in a very
broad way, whether you should act through raising aggregate demand,
or through manpower policies, or both, which probably is frequently
called for. And that is why we talk about balance between policies,
there is a need for indication as to where the major emphasis needs
to be.
Chairman P r o x m i r e . Maybe you are right on this, I have not had
a chance to think this through. But it seems to me the aggregations
are not going to be useful on the national basis and the overall basis
without having a very careful determination of precisely the kind of
occupations you are talking about, and doing it in an area where you
have at least a reasonable degree of labor mobility, an area where you
have at least a reasonable correction between the supply of labor and
the capacity of that labor to be trained, or to move into a particular
job. In other words, the overall aggregate figure, without careful
qualification, isn't going to be very useful in any way that I can im
mediately see that we do not already have now. And maybe I am
being unfair.
Mr. C r e a m e r . Again, we do not have it in the sense that it is not
quantified. We do not know with any approximate precision the
magnitude of the maladjustment.
Chairman P r o x m i r e . For example, we had a dispute before this
committee a couple of years ago, and Mr. Martin of the Federal
Reserve Board told us that in his judgment we could not get below
5-percent unemployment without inflation, because of the structural
problems. And you feel that this kind of information would give us a
clearer picture of whether or not this would be possible?
Mr. C r e a m e r . Well, it certainly could make a contribution to the
debate. If at the time that Mr. Martin made this statement there
were very few jobs going begging, that is, the employers were able to
fill all the jobs they had, then it seems clear that something has to be
done to increase aggregate demand, so the employers would want to
put more people to work.
But if the other situation prevailed, that there were many more jobs
that could not be filled, and employers were actively seeking to fill
these jobs, and they could not be filled, despite a 5-percent unemploy
ment rate, then this suggests, we would think, that you do need man
power programs to enable these unemployed to acquire the training
that is needed to fill these jobs-—or there could be quite a variety
of programs, they may have to define or redefine the jobs, break them
down, and so on. But this could serve as an alerting signal.
Now, I think you are completely right, Senator, when you say you
need this other detail. Certainly when you move to particular man
power programs it seems to me you would be completely frustrated in
shaping programs without this additional detail. But this detail, or
some of this detail, can come out of the job vacancy surveys, plus
matching it with information on unemployment.
Chairman P r o x m i r e . Mr. Knowles. Mr. Goldfinger wanted to
comment on it; Mr. Knowles and then Mr. Goldfinger.
Mr. K n o w l e s . Let's go back in time to some actual situation
that this Congress and this country faced, when we had to make
some decisions, and ask ourselves what we would have done in that
128
JOB VACANCY STATISTICS
situation if we had the information. Take, for example, 1961, when
the Kennedy administration came in. I think if you had the data
available then—and I am guessing, but I think this is a good scien
tific guess—you would have shown that unemployment was far in
excess of vacancies for unskilled workers. For the most highly skilled
occupation the reverse would have been true, you had more jobs
than you had unemployment. And somewhere in between you would
have spread all the other occupations. And some of them would
have been about in balance. We don't know at the moment which
ones of those would have been in which category, but in general this
kind of a distribution up and down the skill category would prob
ably have prevailed; if the labor market analysis of the last 50 years
is worth anything, it ought to tell us that, though in broad terms
you cannot say which would have been where. If you had the data,
you would have known what occupations were where, and what
skill was needed, and what the quantities were. My question is,
would this additional information about winch occupations were where
on that scale, from one end to the other, plus a knowledge of the
magnitude in each one of these steps in the schedule, have told you
more for policy purposes than you knew at the time, and that you
could do anything with? What could have been done in designing
policy that you could not do without it?
Chairman P r o x m i r e . Y ou see, this is the kind of information, it
seems to me, that is much more valuable in determining Government
policy than the aggregate.
Mr. C r e a m e r . I agree, particularly when you come to programs.
We are suggesting this only—why throw away information? You
would be getting this information. And it is relevant to judging the
magnitude of the maladjustment, and more particularly whether the
maladjustment is growing or diminishing.
Mr. K n o w l e s . But if you had all of this detail as to quantities and
the identification of what occupations were where in that scale in 1961,
would you have decided, recommended a different policy than what
was recommended to the administration? Would there have been any
difference in the recommendation of policy? This is a test of whether
it is worth spending public funds.
Mr. C r e a m e r . Let me say one thing, that this is only one, and I
think in our judgment a minor, use—as the statement indicates— o f
the job vacancies.
Chairman P r o x m i r e , Why can't you give an emphatic yes to Mr.
Knowles? It seems to me that you would certainly recommend
sharply different policies, because you would know the precise area
where you would move into manpower training, and the area where
manpower training would be useful, and the amount of money you
would want to spend on manpower training in a particular area, to a
much more refined and useful degree than you would without it, isn't
that true?
Mr. C r e a m e r . I think I have indicated that before, that when you
move into any sort of manpower program this is what you need. But
to indicate where the emphasis needs to be at a particular point, it
seems to me you have to know something about the magnitude and the
direction of change in the maladjustment.
Chairman P r o x m i r e . Mr. Goldfinger?
Mr. G o l d f i n g e r . It is precisely the kind o f excellent statement by
Mr. Creamer that leads me to a negative attitude—much more nega
JOB VACANCY STATISTICS
129
tive than my prepared statement, by the way—toward the entire
approach of aggregating job vacancy averages.
For example, Mr. Creamer says, the difference between the two
aggregates, the job vacancies aggregate and the unemployment
aggregate, indicates which factor is predominant, demand or supply.
And then he goes on: This suggests that an increase in aggregate ex
penditure will reduce unemployment without undue inflationary
pressure. In other words, the mere matching of these two aggregates
should be used for determining our Nation's fiscal and monetary
policies. Well, I seriously question this approach. And I am seri
ously bothered by this aggregate approach.
For example, as you stated or implied, Senator, labor is not substi
tutable. We are not dealing with numbers, here, of x and y and z
and a lot of abstract numbers. We are dealing with real people,
human beings with occupations and skill levels and educational
levels, with people who may be unemployed and who live in one part
of the country rather than in another part of the country. Mr.
Creamer may wind up with an aggregate job vacancy statistic which
is very large. Yet if we were to have the detail— the occupational,
skill, educational, wage, and area detail we may find that, over
whelmingly, the job vacancies are for schoolteachers and engineers
and technicians, whereas large numbers of people are unemployed
and those unemployed are unskilled and semiskilled industrial and
service workers.
However, if we determine our fiscal and monetary policies on the
basis of the aggregate statistics, we may prolong the unemployment
period and aggravate the unemplyment situation for the unskilled
and semiskilled workers, while we do nothing about the shortages
that may well exist for schoolteachers, engineers, and technicians.
This is one of the problems that bothers me a good deal in this whole
discussion of job vacancies.
Chairman P r o x m i r e . It bothers me. And I think it is a very
good point.
Mr. C r e a m e r . I think our statement is more careful than has been
suggested here.
Whenever we talk about a balance of policies—and in the example
that Mr. Goldfinger poses, it seems to me you need both, if you have
people who cannot fit into the jobs, you have to have manpower
programs that will create this supply. At the same time you may
very well have to have fiscal and monetary policies to increase the
demand for workers.
So we speak of a balance. And certainly we don't say that this
would be the sole determinant of fiscal and monetary policy, we just
say it could make a contribution to the sort of balance that seems to
be appropriate. And I think that is all we claim, And I don't-----Chairman P r o x m i r e . You think that this aggregate point is, as
you say, a relatively minor point, in your judgment?
Mr. C r e a m e r . Yes.
Chairman P r o x m i r e . And it is one that we have to guard against.
And I am particularly concerned because a man as expert and as
competent as you are in giving this weight to this, it seems to me,
puts us on guard as to what very influential people of less competence
might construe this to mean. If they use aggregates in a way which
governs their use of fiscal and monetary policy, it could be quite
disastrous.
130
JOB VACANCY
STATISTICS
Mr. M y e r s . May I interject a word on the question of policy in
1961, raised by Mr. Knowles?
A decision on the most appropriate policies to reduce unemploy
ment should depend not only on the numbers of vacancies and un
employed in individual occupations but also on the total numbers of
vacancies and unemployed. A different combination of policies is
appropriate, for example, when total unemployment is much greater
than total vacancies than when the two are nearly equal. A tax cut
may be the most desirable policy in one situation, an expansion of
training and placement activities in another.
(Supplementary material appearing below was later supplied for the
record by Mr. Myers:)
T
he
U ses
of
A
ggregate
St a t is t ic s
on
Job V
a c a n c ie s
This is an expansion of a portion of the statement “ Feasibility and Uses of Job
Vacancy Statistics” presented by Mr. Creamer and me. The discussion during
the Com m ittee hearings indicated that some clarification of the uses of aggregate
jo b vacancy statistics may be helpful.
W e shall begin by indicating some uses we are not proposing. W e do not sug
gest that the total number of jo b vacancies in the United States be subtracted
from the total num ber of unem ployed to get a “ corrected” unem ploym ent total.
W e do n ot think that such a measure would be meaningful, and such an interpre
tation of our statement would be a distortion. Secondly, we are not proposing
that measures derived from aggregate statistics on jo b vacancies should be used
in autom atic rules for policies of the government, such as “ if ‘X ’ falls tw o points,
apply policy ‘C V ’ Rather, we are suggesting that aggregate statistics on jo b
vacancies be used as analytic tools to aid in obtaining a better understanding of
the operation of the econom y. They may serve, among many other measures,
as indicators of the condition of the econom y and of the direction of its movem ent,
as reflected in the labor market. Finally, we do not suggest that the aggregates
should be considered to the exclusion of jo b vacancy inform ation on specific occu
pations, geographic areas, etc. W e do suggest that the aggregates provide addi
tional and potentially highly useful information to that which would be gained
from detailed data alone.
W hat will aggregate statistics on job vacancies show? Properly defined, such
aggregates will show the num ber of workers sought by employers at a poin t of
time, the unsatisfied or unfilled demand for labor. Further, if an appropriate
definition is used in collecting job vacancy statistics, one that is symm etrical to a
definition of employment, then inform ative measures can be obtained by the joint
use of the total number of unem ployed and the total num ber of jo b vacancies.
Such measures will be more reliable, the more symmetrical are the definitions
of vacancies and unem ployment. The following considerations would go a long
way toward obtainining symmetry.
(a) The definition should be comparable with respect to jo b occupancy and
lack of em ploym ent. For example, a jo b might be required to be vacant, or
unoccupied, to be counted as a vacancy. This would correspond to the requirement
that a worker not be gainfully employed.
(b) Supplementing the first point, a jo b vacancy could be defined so as to
include only those openings that had been vacant at least one week. This would
be comparable to the requirement that a person had not been gainfully em ployed
for the week preceding the survey in order to be counted as unem ployed.
(c) The starting date of the vacancy should be comparable to the definition
used for unem ploym ent. For example, immediate occupancy might be a require
ment corresponding to the definition of unem ployment which includes only
persons immediately available for work.
(d) Considerations of wages and working conditions might be included. F or
example, a jo b opening could be defined as a vacancy only if it offered at least
the normal pay at the entrance level of the occupation in question. The corre
sponding requirement for unem ployment would be that the person w ould demand
a wage that would at most be the normal pay for the occupation sought.
As first approximations to measures using vacancies and unemployment, to be
refined and qualified later in the discussion, two may be set forth. The first is
“ E ,” representing excess demand, measured as the total number of vacancies
minus the total number o f unemployed. The second is “ M ,” representing mal-
JOB
VACANCY
STATISTICS
131
adjustment, measured by the smaller of the total number of vacancies and the
total number of unemployed, at the same level of aggregate demand. The first
measure, E, is proposed as an indicator of cyclical movements in the econom y, as
reflected in the labor market. The second measure, M, is proposed as an indica
tion o f the am ount of “ friction” in the labor market, or the mismatching o f
persons and jobs. This mismatching can arise for a number o f reasons, o f course,
some of which will be mentioned later.
Before proceeding to discuss these measures in detail, including qualifications
and refinements, it is im portant to examine the following question: W hat meaning
can be attached to comparisons of total vacancies with total unem ployment when
the content o f the two totals differ with respect to skills and geographic location?
W ould it be meaningful, for example, to compare an aggregate number of job s
that represented mainly vacancies for engineers, nurses, and school teachers to
an aggregate number of persons composed mainly of unemployed manual factory
and construction workers, farm laborers, and unskilled service workers? The
answer to both versions of the question is yes, when an appropriate interpretation
is made.
A comparison between total vacancies and total em ploym ent would not imply,
for example, that unskilled workers would don white caps and take over the
duties of nurses. Rather, it is justified by clearly observable and closely related
processes in the labor market: the mobility of workers between occupations, the
flexibility of the content o f jobs set up b y employers, and the m obility of labor and
capital between geographic regions.
The processes of labor m obility and flexibility of job content operate constantly
in the labor market. The speeds at which they operate depend on the balance
between supply and demand, in total and in detail. As workers in an occupation
becom e scarce, the resulting rise in wages and in costs of hiring lead employers to
reorganize their operations, changing the content of jobs, making it possible to
em ploy workers with skills different from those associated with the occupation in
short supply.
Similarly, workers who find that jobs are scarce in their usual occupation will
look for jobs in other fields for which they are qualified or can becom e qualified.
The two processes can be illustrated by the example of a “ shortage” of professional
nurses and a “ surplus” of unskilled workers.
In response to a shortage of professional nurses, they are relieved of more and
more unskilled or semi-skilled tasks— making beds, carrying bed pans, taking
temperatures, cleaning, etc. Practical nurses, nurse’s aides, and porters are
hired, and if necessary trained, for these tasks. In this way (a) the care of patients
is reorganized so as to need fewer professional nurses and (b) unskilled persons are
brought into new occupations (nurse’s aides, cleaning help, etc.) and trained to
the extent necessary to make them productive.1
When job vacancies exist primarily in one geographic region while unem ploy
ment is concentrated in another, both workers and jobs migrate in response. The
migration is not necessarily done by the unem ployed workers nor the unfilled
jobs. Some m ovem ent of workers from labor surplus to labor shortage areas
usually takes place, however, as does some m ovem ent of factories, offices, and so
forth, in the opposite direction.
The m obility of workers between occupations, the flexibility of job content and
requirements, and the geographic m obility of both labor and capital bring meaning
to a comparison of total vacancies to total unemployment.
Furthermore, an aid to the interpretation of detailed data on job vacancies by
occupation and area, in conjunction with unem ployment data, is suggested by a
consideration of these processes. There is a danger of interpreting data on occu
pational structure and geographic location too rigidly, ignoring the m obility of
labor, the flexibility of job content, and the like. Specific “ shortages” or “ sur
pluses” should always be studied within the larger context of the area or national
labor market. Examination of the situation in the larger market may lead to a
better solution of specific problems. It may also help to avoid mistakes such as a
costly program undertaken to ameliorate a problem that “ solves itself” b y m o
bility, reorganization, or a similar process.
The original statement outlining the uses of aggregate measures of vacancies and
unem ploym ent stated that these measures could aid in determining the proper
balance between two types of policies. The first type is fiscal and m onetary
policies, designed to bring about more or less rapid expansion of the Gross Na
1 See also “ Reply" by Charles C. Holt and Martin H. David, in The Measurement and Interpretation of
Job Vacancies, pp. 137-141.
132
JOB VACANCY
STATISTICS
tional Product, indicating the aggregate demand for goods and services in the
econom y. The second type is labor market policies designed to make that market
operate more sm oothly, by im proving placem ent machinery, setting up training
programs, and aiding workers to relocate.
It may be helpful at this point to review how a general rise in aggregate dem and
for goods and services affects the labor market. Employers respond to increased
dem and for their products and services at first by reducing inventories, to the
extent that this is possible. Quickly, they respond by increasing production and
thereby increasing their demand for labor input. The demand for labor is, of
course, derived from the demand for the goods and services labor helps to produce.
Labor em ployed can be increased in three ways, by increasing hours w orked per
employee, by hiring unem ployed persons, and by hiring persons who were not in
the labor force but enter it and take jobs in repsonse to attractive opportunities
for employm ent. As unem ployment falls (but less than the rise in total em ploy
ment), prices, then wages, rise. The extent to which they rise depends upon the
extent to which previously em ployed persons were working short weeks, i.e., were
under-employed, the size of the pool of previously unem ployed persons, and the
number of persons readily drawn into the labor force. It also depends upon the
ease with which the last tw o groups, newly hired, can be trained and put to w ork
in newly-created jobs. This may require migration of labor and capital between
geographic areas, as well. When a rise in aggregate demand affects an industry
where there were already shortages in specific occupations, the last tw o considera
tions becom e critical.
The primary function of aggregate statistics on jo b vacancies is, again to pro
vide helpful guides, along with other econom ic knowledge, to the solution of the
question: H ow can unem ployment be reduced and output of goods and services
in the econom y correspondingly increased in the most efficient manner? It is
observably true that unem ployment can be eliminated almost completely if the
total demand for goods and services rises rapidly enough. In the United States
during war time and today in many countries in Western Europe, unem ploym ent
has been practically eliminated. This did not occur w ithout cost, however.
In every case, rapid price and wage rises have accom panied the elimination of
unem ploym ent when this came about primarily through the expansion of aggre
gate demand. Programs to retrain workers, im prove and speed up placem ent
and otherwise increase labor m obility can sm ooth the reduction of unem ploy
ment, helping to achieve the desired goal without inflation.
A hypothetical example may help to clarify the problem. Let us consider tw o
alternatives. In Case A, there are, simultaneously, four million jo b vacancies
in the econom y and four million unem ployed persons. In B, there are one million
jo b vacancies and four million unemployed. If fiscal and monetary policies were
used to reduce unem ployment through expansion of aggregate demand, these
would be much more likely to lead to rapid price and wage increases in Case A
than in Case B. While the number of unem ployed in both situations is the same,
the degree of maladjustment is much greater in Case A than in Case B. Further,
the aggregate demand for and supply of labor are apparently in approximate
balance in situation A, while there is a “ labor surplus” in B, in the sense that
supply exceeds demand by a large margin.
Comparisons of total vacancies and total unem ployment lead to the conclusion
that it would be better to rely more heavily on labor market policies in A and more
heavily on general expansionary policies, fiscal and monetary, in B. A precise,
analytic statement of the interpretation of total of unem ployment and vacancies
has been given by Jacob Mincer. 2 He states that the mix of policies should be
that which yields the greatest excess of social returns over social cost. That is,
apply labor market policies to the extent that a better allocation of resources is
obtained through their use and similarly for fiscal and monetary policies. The
knowledge required to translate that statement into a practical aid to the form a
tion of public policy can only, of course, be obtained after much study and the
collection of a large and continuing body of vacancy data.
The com putation of the two aggregate measures suggested above requires a
solution of the practical problem of measuring the degree of m aladjustm ent in
the econom y, represented by “ M .” W ith the same degree of maladjustment, a
rise in aggregate demand will draw persons into em ploym ent who were unem ployed
owing primarily to friction, structural imbalances, etc. Thus the number of
vacancies offset by unem ployed will vary inversely with the level of aggregate
demand. W hile the statement is correct that the degree of maladjustm ent is
* “ Comment,*' The Measurement and Interpretation of Job Vacancies, pages 120-127, especially page 125.
JOB VACANCY STATISTICS
133
shown at the same level of prosperity (or aggregate demand) by the smaller of
the number of vacancies and the number of unemployed, a practical measure
requires some assumption about the substitution of unem ploym ent for vacancies
as aggregate demand is increased. Once an appropriate assumption has been
adopted, we can estimate the number of employed that would exist if total va
cancies and unem ploym ent were equal, and use this number as our measure of
the am ount of maladjustment, M .3 Different values of M will indicate m ore or
less “ mismatching” of jobs and persons, owing to structural causes (both skills
and location), simple turnover, etc.4 The adoption of a method o f measuring
maladjustm ent permits the com putation of the excess demand for labor, E. A
meaningful m ethod of com putation is as follows: When the total number of
vacancies is greater than the total number of unemployed, E is the excess of
total vacancies over maladjustm ent (a positive q u an tity ); when the total number
o f vacancies is smaller than total unem ployment, E is the excess of maladjustment
over unem ployment (a negative qu an tity). In this way we will have two measures
that will show the following, at any given time: (a) the number of persons that
are unem ployed owing to the failure of the labor market to function perfectly
when the supply of and demand for labor are in approximate balance, shown by
M ; (b) the degree to which the demand for labor exceeds or falls short of the
am ount needed to achieve an approximate equality of the demand for and supply
of labor.
T o return to the hypothetical example given above, we could deduce that mal
adjustm ent was 4 million in Case A but only 2 million in Case B, while the excess
demand for labor was zero in Case A and minus 2 million in Case B.
M any of the ideas contained in this memorandum and in other writing on the
subject, both in this country and abroad, were inspired by an article that ap
peared in 1956 by the British economists J. C. R. D ow and L. A. Dicks-M ireaux.5
In their study a measure similar to E was proposed and com puted for the British
econom y. It is regularly published in the National Institute Economic Review.6
Definitional problems in jo b vacancy data that are difficult to resolve arise
when it is wished to obtain a high degree of symmetry with unem ployment.
Perhaps the most troublesome are the determination of whether or not a job is
“ unfilled” and the related question of job openings arising from search for and
training of workers in anticipation of growth and turnover. These considera
tions lead to the suggestion that movements in E be given more attention than
the level of E.7
The relationship between the variables and the implied policies is summarized
in the following diagram.
Changes in vacancy and unemployment measures and public policies
“ E ” rising (positive or negative)
Policy of constraint plus man
power policies.
Policy of constraint_______ ______
____________________
“ M ” falling
“ M ” rising_____________________
“ E ” falling (positive or negative)
Policy of expansion plus man
power policies.
Policy of expansion.
There are problem s in the interpretation of aggregate jo b vacancy statistics,
as well as in the collection of appropriate data. These problem s will not be
solved in casual discussion nor in newspaper debate, but only by careful study.
3 For a discussion of this measurement problem see the Measurement and Interpretation of Job Vacancies,
p. 406, fh. 2.
4 Robert Ferber has proposed a measure of maladjustment that is more restrictive than M . His measure
excludes the frictional element within occupations, areas, or other classifications. It would thus more
closely resemble a measure of structural imbalance. See “ Introduction and Summary” in the Measurement
and Interpretation of Job Vacancies, pp. 15 to 18.
5 “ The Excess Demand for Labour: A Study of Conditions in Great Britain, 1946-1956/' Oxford Economic
Papers (N.S.), February 1958, p. 1-33.
6 The British index is derived from a rather elaborate computation of maladjustment and excess demand,
industry by industry. This is necessitated by the British data. Vacancy statistics in Britain are collected
through the public employment service and the coverage varies by industry and over time. Dow and
Dicks-Mireaux corrected the data by industry, computed an index of excess demand for each industry
group, and then combined these to obtain a national total. A published index is available for the period
1948 to date for study in conjunction with other measures of the operation of the British economy and should
be of some aid in resolving questions of interpretation of U.S. job vacancy data.
7 Compare the suggestion of Joan Robinson: “ A growing or falling excess of vacancies over unemployed is
quite a useful indication, over the short run of the movements of demand, and a drop in both would pre
sumably indicate an improvement in the general conditions of mobility of labor or versatility of management.
Economic Philosophy, p. 92.
134
JOB VACANCY
STATISTICS
Once a comprehensive b od y o f data becomes available, covering a significant
fraction o f the U.S. labor market as well as a reasonable time span, we can
determine the validity of job vacancy statistics in the uses mentioned above.
The promise they hold for obtaining new, valuable inform ation makes such study
and the prerequisite collection of data appear to be worthwhile ventures.
Chairman P r o x m i r e . Would you say that the suggestion of Mr.
Goldfinger that we have a completely comprehensive employment
service survey with computers, so that we hava a more accurate and
perhaps a more reliable basis for judging excess demand, or however
you want to put it, would be preferable, disregarding for the moment
the cost?
M r. M yers . I could offer one piece of evidence which suggests to
me that it might not. In some Western European countries with
comprehensive employment services, employer surveys are used to get
better information on the demand for labor. I mention two countries,
Sweden and the Netherlands, where this is done.
Chairman P r o x m i r e . What is the relationship between the em
ployer survey and the employment service? Does the employment
service bring in 20 or 30 or 40 or 50 percent or is it simply a fringe,
5, 10 percent area? Or is it one checking the other?
M r. M yers . The employer surveys do bring in more vacancies
than the employment services.
Chairman P r o x m i r e . That is very interesting.
Mr. C r e a m e r . I would also like to suggest that you can bring the
horse to the watering trough, but you cannot make him drink. It is
all very well to envision a comprehensive employment service in this
country. But at this point, and in the foreseeable future, there are
many employers who—some for legitimate reasons and some for
illegitimate reasons—won't use the employment service.
Chairman P r o x m i r e . So in order to have one that is comprehensive
you have to have a degree of compulsion?
Mr. C r e a m e r . Yes; and enforcement.
Chairman P r o x m i r e . And how about the cost? Would the cost
be very great or not?
Mr. C r e a m e r . Y ou m e a n budgetary cost?
Chairman P r o x m i r e . Yes.
Mr. C r e a m e r . I really do not know. But I think, judging by our
experience in Rochester, if that is at all typical, there would be some
serious social costs.
Chairman P r o x m i r e . Your experience in Rochester would be that
you feel you could proceed—I do not want to anticipate—you could
proceed on the basis of voluntary responses which you think would be
adequate and accurate?
Mr. C r e a m e r . Yes, I think for the next decade or so, if you want
this information, that is how you have to get it.
Chairman P r o x m i r e . Go ahead.
Mr. C r e a m e r . Another aspect of the same use is that the two
series—job vacancies and unemployment—could serve as an early
warning system for impending changes in wage rates as well as in
prices because of the ramifying effect of the former. If prices of output
and of labor input continue to be determined by the play of forces in
unregulated markets, wage rates will tend to rise faster than the rate
of productivity increase when the vacancy rate exceeds the un
employment rate and especially if the excess continues to expand.
As the data accumulate, it should be possible to determine empirically
JOB VACANCY
STATISTICS
135
at what point the discrepancy between vacancy and unemployment
rates signals extra upward pressure on wage rates and output prices.
An early signal creates the possibility of invoking fiscal and monetary
measures to counter the incipient inflationary pressures. This
possibility takes on added importance whenever the U.S. economy
has a deficit on its international account.
If the national sample is large enough to provide occupational
detail, the series should be helpful to those in the Federal Government
who have the responsibility for the formulation of a national manpower
program which must include Federal support and general directives
for vocational training, retraining, counseling and assisted geographic
transfer. The greater the excess of job vacancies over unemployment,
the greater is the need for these programs to assist the labor market
to perform its function of allocation, so as to minimize inflationary
pressures. If certain occupations are important in the roster of job
vacancies over a period of time, in particular if they show high vacancy
rates (relative to employment in those occupations), this would give
some guidance for the selection of training programs. The guidance
would be more sure for those programs that can be completed in 24
or fewer months.
Uses at the local level
Equally important in an understanding of what a national series
would not accomplish. The Department of Labor, and especially
the Bureau of Employment Security, rate placement use of job
vacancy data equal to the policy uses. However, job vacancy data
at the national level can contribute virtually nothing directly to the
placement of the unemployed in the job openings. This required
knowledge of a particular opening in a specific location together with
knowledge of a qualified applicant. This suggests that implementation
of the placement use requires job vacancy data for each important
labor market. The national total required for policy uses would
then be obtained by aggregating the job vacancies for all the labor
markets surveyed.
Even at the level of the local labor market the usefulness of job
vacancy information for placement will depend on the frequency of
the reports, the speed with which the information is made available,
and the efficiency of the placement agencies, both private and public.
The first two considerations are organizational matters which surely
can be mastered. While a more effective matching of jobs and
applicants is more difficult to accomplish, certainly all parties to the
matching process—employers, job seekers, and placement agencies—
will gain from being better informed about the number and type of
unfilled vacancies. This possibility of assisting the placement process
constitutes one of the principal justifications for developing job
vacancy statistics by local areas. It is, of course, a more costly
program than one designed merely to provide national statistics.
Another important justification of local area statistics is the con
tribution these data could make to the development of guidelines for
vocational guidance and counseling at all educational levels in the
local area and for training and retraining programs. Since no one
labor market is a replica of the national market, national statistics on
job vacancies would provide little or no guidance to local authorities
charged with these responsibilities. This can come only from the
statistics that relate to a local area.
136
JOB VACANCY
STATISTICS
It is only local information moreover that will enable the employers
in an area to appraise the outlook for wage rate changes and the
outlook for the success of their own recruitment programs. Being
forewarned at least creates the possibility of being forearmed. But
with the current absence of local job vacancy statistics, few employers
have comprehensive information on the number of workers that other
employers are seeking for the same occupations he trying to fill.
With this information each employer should be in a better position
to devise recruiting strategies to accomplish his hiring objectives.
This would also be true in the matter of pressure on wage rates in a
particular labor market. The same information, of course, would
also benefit job seekers.
If the availability of job vacancy statistics helps the formulator
of economic policy, administrators of various manpower programs,
placement agencies, and last but not least, individual employers
and job seekers perform their respective tasks more effectively, the
collection of these statistics on a continuing basis certainly deserves
serious consideration and exploration. Moreover, experience in the
past few decades with newly developed statistical series suggests
that many significant but unanticipated uses emerge as the analysts
make use of the new data. There is a high probability this would also
happen with job vacancy statistics.
Let me say that in all relevant respects that we could determine—•
the Rochester area seems typical of other substantial labor market
areas in the country. I say in summary, we feel that the Rochester
area is not atypical of other metropolitan areas in the United States,
except -possibly in the degree of prosperity which existed there during
1965. The changes in the national employment situation in the suc
ceeding 12 months have rendered Rochester more representative of
other metropolitan areas today than it was when we carried out our
survey—that is, the rest of the country has caught up with Rochester.
C.
P R IN C IP A L
F IN D IN G S
Representativeness of the Rochester area
Monroe County, N.Y., was chosen for the NICB survey. The
special characteristics or features of the Rochester area are important
insofar as they affect the generality of the conclusions that may be
drawn from the NICB study. The question is whether or not Roches
ter is sufficiently similar to other metropolitan areas to permit a judg
ment on the feasibility of nationwide job vacancy surveys, based on
our experience. A search for a “ typical” or “ average” area is, of
course, fruitless. Further, the structure of the labor market and
specific vacancy data that may be collected are not pertinent to the
question; only the conclusions on survey feasibility.
The county is heavily industrialized and urbanized. Manufactur
ing of durable goods, particularly in the “ photographic, optical, and
instruments” industry, is especially important and skilled workers
represent a large fraction of employed workers. The area may be
further characterized as very prosperous with low unemployment
and high average incomes and educational attainment. A detailed
comparison of these characteristics for Rochester with those of other
metropolitan areas in New York State does not reveal any sharp con
trasts, however. Rochester has the highest average family income in
JOB VACANCY STATISTICS
137
the State and, currently the lowest unemployment rate, but the dif
ferences between Rochester and other metropolitan areas are not
great and do not indicate that the area is unrepresentative. The
occupational structure and the proportion of nonwhites in the popu
lation are reasonably typical of the State. The proportion of foreign
born in the population is rather high but is the result, in large part,
of migration during a period many years removed.
One special characteristic worthy of mention is the presence of a
highly developed and effective employers’ organization in the area,
which aids in placement and other aspects of employment. In sum
mary, we feel that the Rochester area is not atypical of other metro
politan areas in the United States, except possibly in the degree of
prosperity that existed there during the first half of 1965. The
changes in the national employment situation in the succeeding 12
months have produced a situation where Rochester is more representa
tive of other metropolitan areas today.
H IG H L IG H T S O F S T A T IS T IC A L F IN D IN G S
In each of three surveys in Monroe County, about 400 employers
were interviewed. The surveys collected information as of Feb
ruary 12, May 14, and August 13, 1965. Almost all data were collected
by personal visits to the employers’ offices; all interviews were com
pleted in the 2 weeks following the reference date. This is one differ
ence from the BES survey techniques that relied primarily on mail
canvass.
The following findings are the ones most relevant, in our judgment,
to the committee’s discussion:
1. The definition of a job vacancy—an unfilled job that an em
ployer is actively seeking to fill by hiring a person from outside his
organization—is operational, is understood by employers, and elicits
reasonably accurate responses. Equally important, the Rochester
employers have clearly demonstrated their willingness to respond to
a voluntary statistical reporting program on job vacancies—the
response rate was 99 percent.
2. Total number of vacancies in the county was estimated at
about 8,000—and there is little variation for the three dates—approx
imately 3 percent of all jobs, filled and unfilled. In the mid-May
and mid-August survey this number and rate exceeded the unem
ployment rate.
3. A sample of 400 employers, selected from predetermined ratios
of employers of different sizes, is adequate to provide reliable estimates
of the total number of vacancies.
4. The vacancy rate, or proportion of jobs unfilled on the survey
date, did not differ greatly between firms of different size.
5. The classification of the occupational titles supplied by employ
ers into the standardized classifications of the Dictionary of Occupa
tional Titles is practicable.
6. I might mention here—this is another difference in the survey
schedule—the Conference Board schedule did ask for each employer
to indicate on each unfilled job which he was seeking to fill the mini
mum acceptable education, and the minimum acceptable related job
experience. This was not attempted on the BES schedules.
From 36 to 45 percent of all vacancies, depending upon the
survey month, were open to those without a high school diploma.
138
JOB VACANCY
STATISTICS
7.
From 14 to 25 percent of vacancies were available only to per
sons who were college graduates. (The rather wide range results
from seasonal variation in vacancies for schoolteachers.)
Now, here is another difference in our schedule. We asked not
only for job vacancies with immediate starting dates, but also job
vacancies with future starting dates. The BES restricted theirs only
to immediate starting dates.
And I am pleased to note in Mr. Ross’ statement that in their
proposal for new work they are planning to include future starting
dates as well.
Chairman P r o x m i r e . How much difference does this make?
Mr. C r e a m e r . Well, of the 8,000 I think roughly 1,000 were for
future starting dates.
Is that about right, John?
Mr. M y e r s . It varies widely. It was quite high in February,
and low in August, because of schoolteachers, primarily.
Chairman P r o x m i r e . But overall it is about 12 or 15 percent?
Mr. C r e a m e r . In Rochester for that year the percentages were
35, 16, and 14 for the 3 survey months.
What it catches in large part are those seeking to hire college
graduates, university graduates, and high school seniors. And
these would not be caught if the survey is restricted solely to immediate
starting dates.
And of course, by tabulating these openings separately by immediate
and future starting dates you can retain symmetry with the unem
ployment definition. If you do want symmetry for total vacancies
(immediate and future starting dates combined), this calls for rede
fining the labor force. But this is aside-----Chairman P r o x m i r e . I think it destroys your symmetry, doesn’t
it? If a college student expects to go to work after he graduates,
he is not considered as part of the labor force or unemployed if he is
in college or still in high school. On the other hand, if you are going
to have future starting dates you exaggerate your vacancies, inas
much as you do not expect to fill them, and they are not going to be
available until the student has finished school.
Mr. C r e a m e r . That is correct. And therefore, we tabulate the
two separately, so that you have job vacancies with immediate starting
dates. Now, that is symmetrical at that point with the unemployment.
In addition, it seems to me you are needlessly throwing away
information if you exclude job vacancies with a future starting date.
Chairman P r o x m i r e . They are segregated?
Mr. C r e a m e r . Yes.
My only point is that for this total number of job vacancies, if you
wanted symmetry, you would have to define the labor force concept.
Now, this hiring for future starting dates is a form of pressure on the
labor market, and it seems to me some count should be made of it.
Chairman P r o x m i r e . Mr. Goldfinger.
Mr. G o l d f i n g e r . I have a question.
To what extent are these future job vacancies related to seasonal
employment kinds of occupations and casual labor type of occupations?
Mr. C r e a m e r . Some of it is seasonal. We found in our February
survey that there were a large number of workers that were needed
in the construction industry. It seemed to us that the construction
contractors were anticipating their labor needs for the spring pickup
in construction.
JOB VACANCY STATISTICS
139
Mr. G o l d f i n g e r . T o what extent are those actual job vacancies?
For example, if a building contractor says that he is going to have
openings, let’s say, for 10 pipefitters and 12 sheetmetal workers
on July 1, is that a job vacancy, today, or isn’t it?
Mr. C r e a m e r . If he is willing to enter into a commitment. He
may know that he gets his contract with a certain starting date, and
he has to prepare in advance to staff up. And if he is ready to enter
into an employment commitment, we counted it as a vacancy, but
with a future starting date.
Chairman P r o x m i r e . If he is willing to, or he did enter into a
contract?
Mr. C r e a m e r . If he did, it is not a vacancy. It is if he is willing
to. He has already filled it if he has already entered into a contract.
Mr. G o l d f i n g e r . I would suggest that we have come upon an
additional kind of problem, and that without some real depth and
knowledge about the hiring and employment practices in a particular
industry, we are in a kind of never-never land. In the case of the
construction industry, is that a job vacancy or isn't it? This con
tractor who is going to have a need for 10 pipefitters on July 1 may
have informed the union, or whatever other source he uses, that on
July 1 he will need 10 pipefitters. He will probably get them on
July 1. This is not an unfilled job today. It undoubtedly will not
be a job vacancy on July 1. Or is it? How can you be so sure?
Representative C u r t i s . Sure it is, people are unemployed, and
these jobs are vacant, and will be available.
Mr. G o l d f i n g e r . But it is not vacant today, sir. And it probably
will not be vacant on July 1. This is in the nature of the construc
tion industry. What I am suggesting is that there is a problem here.
Unless we know, in some kind of depth, the practices of the industry
and of the occupation—particularly in seasonal and casual labor oc
cupations—we may be getting numbers that are of questionable
meaning. I have no interest in numbers for the mere purpose of
having another statistical series.
Representative C u r t i s . I agree that you have to go into depth in
many of these things. And you are rightly pointing out some of the
difficulties. But you have got a great deal that lies within the center
here, which this paper is pointing up to. I think your remarks are
quite appropriate.
Mr. C r e a m e r . Certainly more is needed. And this in a sense
brings you to the seasonal correction factor. But to know how to
develop a seasonal correction here you have to have surveys over many
years. And in order to have surveys over many years you need a
congressional appropriation that is adequate to the job.
8. About 50 percent required no related work experience. (How
ever, a large proportion of these required the completion of at least
12 years of schooling.)
9. A large proportion of the openings with comparatively low ed
ucational requirements required related work experience.
10. From 16 to 22 percent of all vacancies required neither high
school graduation nor related work experience. However, a large
proportion— 59 to 63 percent— of these were in unskilled and service
occupations.
11. Of those job vacancies reported as of mid-August 1965, 27
percent had been open at least since mid-May, that is 3 months or
140
JOB VACANCY STATISTICS
more, and 12 percent had been open at least since February, or 6
months or more. This is one possible measure of the magnitude^of
hard-to-fill jobs.
That is, from our preliminary conversations with several dozen
employers in the Rochester area before we undertook our formal
survey we tried to determine how accurately they could answer this
question on the duration for which the vacancy had been unfilled.
And we found that they could not give you very good answers without
their doing a good deal of work on their own files. And we thought
this might handicap the course of the survey if we asked for this
information. So we did not.
But if you have what is known as a shuttle schedule, that is, the
information for the preceding quarter for each particular job is given
along with the information for the current schedule, you can determine,
then, which jobs had remained open. And that is what we did in our
August survey. After we got the information of the number of job
vacancies on August 13 we then asked how many of these jobs had
been open in May when we called on him, and in February. And the
interviewer had the schedule with him. And he could check it out.
Representative C u r t i s . You are now going on to the next topic?
Mr. C r e a m e r . Yes.
Representative C u r t i s . Could I ask just this one point.
You didn’t make any— this is not said in the way of criticism, I am
just trying to find out—you didn’t make any attempt to find out what
had created the job vacancy?
Mr. C r e a m e r . No, we did not. But I might say that with the
help of a research contract from what was called the Office of Man
power, Automation and Training a couple of weeks ago—I understand
now it has another name—we are looking into that aspect of it this
coming year. And Dr. Myers is engaged in trying to determine what
analytical value job vacancy data have in analyzing the operations of a
local labor market.
Representative C u r t i s . These data that you have presented here are
very intriguing, and suggest something that I have concluded. Let us
say that company X — and this relates really to some of the problems
that I think Mr. Goldfinger was directing our attention to— company
X has a job vacancy created as a result of retirement, let’s say.
Mr. C r e a m e r . Yes.
Representative C u r t i s . Well, this is a higher skilled job, and
probably what the company will do is promote from within. There
will be a shift upward, so that the job vacancy is at the low^er echelon.
The statistics seem to indicate that there is this feeding in process.
I do not know.
Mr. C r e a m e r . Very definitely.
Chairman P r o x m i r e . As long as we are on this, did you have any
chance to compare your findings with those of the study that was
reported yesterday by Mr. Chavrid?
Mr. C r e a m e r . Yes, we have. And it does form part of chapter 4
of the full report that the committee has.
Chairman P r o x m i r e . The reason I asked is because he has such a
terrific disparity in some of these areas. For example, in Minneapolis-St. Paul, as I recall, only 30 percent of the jobs were hard to
fill, and in other areas as high as 90 percent were, indicating to me that
there was a great difference in the way the questions were asked, or
JOB VACANCY STATISTICS
141
the way the responses were secured. With such a discrepancy you
certainly could not take the Rochester findings and project anything
on a national basis.
Mr. C r e a m e r . No.
Chairman P r o x m i r e . All you could do was to say there is what the
situation in Rochester was at the particular time.
Would you agree that this would apply to each of your findings here,
each of your 10 points, that you cannot make a projection that 20
percent required new work experience?
Mr. C r e a m e r . N o, this varies with the industry structure, which
again determines the occupational structure, which again determines
the relative difficulty of filling jobs.
Chairman P r o x m i r e . Do you think your mail questionnaire ap
proach instead of direct interview has resulted in any difference in
your findings?
Mr. C r e a m e r . Yes; I think so, at the exploratory stage.
Chairman P r o x m i r e . A direct interview would be more satis
factory?
Mr. C r e a m e r . Yes; and on this very point that you mentioned, one
reason why they got this large variation in the percent—in the dura
tion of jobs—I am sorry, did you mention the duration of jobs?
Chairman P r o x m i r e . Yes; as hard to fill.
Mr. C r e a m e r . Yes; and they based that on the duration.
There is ambiguity in the way they put the question. Many
employers didn’t know whether they were referring to how long the
job they were offering would be in being, that is, whether it was a
temporary job, or a regular job from year to year.
Chairman P r o x m i r e . If they make that kind of a mistake, it is
really a distorted picture.
Mr. C r e a m e r . This certainly happened in the Buffalo survey that
you mentioned. And for some reason or other they didn’t improve
the wording of that question in the 16-area survey. So I am very
sure that this explains a good deal of that variation.
Well, my point is, with the interview it is easier to catch these mis
interpretations of the questions and correct them, and therefore,
eliminate that source of error, or come closer to eliminating it.
Representative C u r t i s . Could I come back just a minute to the
reasons for job vacancies? I was just jotting down for my own
information the kind of things that would create a vacancy. Retire
ment is one. And I would probably include escalation, where you
would not be filling the actual job from which the person retired.
Quits, of course, would be the other. And we have national statistics
on quits, which is very interesting.
Mr. C r e a m e r . Yes.
Representative C u r t i s . And expansion of the particular industry
could be a local expansion, or it could be in the nature of the national
trend in that industry.
I included automation. Jobs are destroyed by automation all
right, but new jobs are created. They take on more data processors,
for example.
Am I missing some big categories of the cause of job vacancies?
This covers normal turnover. I do think that industries vary in
the rate of turnover. There seems to be a turnover from normal
63-947— 66,------ 10
142
JOB VACANCY STATISTICS
circumstances, and probably a rate to be reached at. But have I
missed some major areas that cause job vacancies?
Mr. C r e a m e r . I think Mr. Myers ought to answer that, since
that is what he is working on.
But I might just say that we try to subdivide the expansion into
two parts, one which you might consider a cyclical pickup or recession,
and sort of a longer term growth.
Representative C u r t i s . Yes; during the period of a cyclical pickup
you would have the same industry filling more job vacancies resulting
from that kind of expansion. I have put down, of course, the ex
pansion more in the nature of technological advancement, and then I
have broken it down further on automation.
Mr. Myers, would you comment?
Mr. M y e r s . I believe you covered all the categories we have, Mr.
Curtis.
Representative C u r t i s . If it is limited to that maybe it would not
be too difficult to get information on what has created the vacancy,
which I would think would be quite revealing.
Mr. C r e a m e r . Yes; that is why we applied for this assistance
to carry out this particular project.
Representative C u r t i s . Thank y o u , Mr. Chairman.
Mr. C r e a m e r . We have estimated the cost of a continuing, quar
terly job vacancy survey from our experience in the Rochester area.
The estimates are for the collection of data in the 146 major metropoli
tan areas of the United States with a labor force of at least 50,000
persons; the estimates assume that the survey has passed the initial
stage of organization, planning, and training of personnel. For a mail
survey we estimate an annual cost of $7.1 million.
Chairman P r o x m i r e . I believe at one point I asked either Mr.
Chavrid or Mr. Ross what would be their estimate of a comprehensive
and adequate survey. And I think they said around $5 million.
Mr. C r e a m e r . I see.
Chairman P r o x m i r e . That does not mean that it would have to
be 146 major metropolitan areas, maybe this can be less than that.
And as I understand, the two and a half million would be for 75 to
80 labor markets.
Mr. C r e a m e r . Yes.
For an interview survey, $9.5 million. As stated in the section on
recommendations for future surveys, we believe that it will always be
necessary to visit some employers personally in order to obtain ac
curate data. An approximate estimate for the continuing collection
of accurate data is thus $8 to $8.5 million per year. To get a basing
point here, I might indicate that I am told that the budget for the
current population survey on the unrevised basis, that is, the 35,000
household basis, is $3 million-plus a year for all the surveys, tabulating,
and analysis.
Now, this, of course, provides a monthly reading of current em
ployment status. But it yields only national totals. You have no
regional or particular labor market totals. So this 7 or 8 million that
we are estimating has to be read against the fact that you will be
getting a fairly detailed count of unfilled jobs by occupations for 146
specific labor areas in addition to national and regional totals.
The uses of job vacancy statistics for analysis and policy decisions
at the national level have already been mentioned above, as well as
JOB VACANCY STATISTICS
143
the value of such data for indicating the national need for manpower
and accelerated placement programs. Less publicized uses in local
areas are for placement, manpower planning (both by community
organizations and employers), and generally to increase the knowledge
available in local communities on the state of the labor market. In
an attempt to learn some of the dimensions of such uses, we asked all
employers responding in our surveys, as well as about 60 schools and
other community organizations in Rochester which had received
summaries of each of the 3 surveys, about (a) the value of our
surveys at the local level and (b) the specific uses to which the data
might be put.
Slightly less than one-fifth of all employers (but nearly one-third
of those employing 250 or more) stated that job vacancy data would
be of direct use in their own operations. However, as many as 72
percent believe that it would be valuable to community organizations.
That is, while the information was of direct use to a modest minority
of employers, most thought that training, guidance, and other labor
market functions could be accomplished more effectively with the aid
of job vacancy data than without. The community organizations,
including schools, concerned with manpower planning and training
stated emphatically that job vacancy data were useful in carrying out
some of their program objectives.
Chairman P r o x m i r e . If it is useful to the community, is there any
consideration given, or do you think Congress should give any con
sideration to sharing the cost, requiring the community to pay a third
of the cost, or part of the cost?
Mr. C r e a m e r . I had not thought about it, Senator.
Chairman P r o x m i r e . Are you recommending any particular
amount? I take it that you support the $2.5 million appropriation
that the Department is asking?
Mr. C r e a m e r . Definitely; yes.
Chairman P r o x m i r e . And you feel that on the basis of what we
discover here we should decide whether or not to go ahead with the
more expensive program, the 146 market areas; is that right?
Mr. C r e a m e r . Yes, I would think so.
Chairman P r o x m i r e . Y o u are not recommending that at the
moment, you are just giving us an estimate?
Mr. C r e a m e r . No. But I think a jump to 75 would certainly be
a big step forward, and probably no more is called for at this time.
Chairman P r o x m i r e . The 75-80 would be sufficient?
Mr. C r e a m e r . Yes.
Senator P r o x m i r e . That would give us what proportion total
employment? A big proportion, much more than half, wouldn’t it?
Mr. C r e a m e r . Yes, I would think so. I have forgotten what
proportion of the total the 16 areas represented.
Chairman P r o x m i r e . Then we would be discriminating against the
smaller markets, I suppose?
Mr. C r e a m e r . Not necessarily. The small markets could be
represented in the selection of the 75 to 80 areas.
I would like to elaborate one point here, the fact that the percentage
of employers finding it directly useful to them came to somewhat less
than one-fifth. In this respect, I am sure Rochester is not typical.
It has an organization called the Industrial Management Council to
144
JOB VACANCY
STATISTICS
which most of the large companies belong, and mostly the manufac
turing companies.
Now, the personnel managers of all associated companies meet
weekly to exchange information on layoffs and hiring needs. So that
they have— they already have a good deal of this information from
these weekly meetings. However, it seems that the Rochester em
ployers have had some second thoughts on the usefulness of job
vacancy data. And if it is agreeable, I would like to read into the
record excerpts from two letters received from the manager of the
Industrial Management Council of Rochester. He had written to us
after receiving our preliminary draft of the report, saying—this is
under date of May 3:
Y on will be interested to know that we in the Industrial Management Council
plan to continue jo b vacancy surveys on a quarterly basis am ong our member
firms. This means, of course, that our surveys will be limited to manufacturing
firms.
Chairman P r o x m i r e . What area would this encompass? Do you
think that while it might be of less use to Rochester to have this kind
of study in Rochester, as long as they have this kind or organization
it might be very useful if they would get a region-wide notion, so that
the communities that are maybe 30 or 40 or 50 miles away, in comput
ing distance, might provide information also, so that labor which
might be available in one area could move into another, and vice
versa?
Mr. C r e a m e r . Yes. Of course, this probably calls for interven
tion or coordination at the Federal level to achieve this.
Chairman P r o x m i r e . One other question here. Is this one-fifth
all employers? Employing what proportion of the labor force? You
say one-third of those employing 250 or more. Would it be half or
more of the labor force? In other words, the big employers in many
towns—for instance, in Kenosha, Wis., one employer, American
Motors, employs a very big proportion of all who work.
Mr. C r e a m e r . It is not that concentrated in Rochester, but East
man Kodak is located there.
Chairman P r o x m i r e . The very large employers, have they indicated
that they would find it useful, or not?
Mr. C r e a m e r . The highest percentage of affirmative answers on
this question came from the medium-sized companies. But as I say,
this second thought seems to suggest that the large companies now
think it worthwhile, and they want to continue it on their own.
Representative C u r t i s . Could you give us an estimate of the
Rochester labor market? What percent of the jobs are in manufac
turing and how many are in, say, service and distribution, including
education and Government employment?
Mr. C r e a m e r . We have that in our final report. And Mr. Myers
is looking it up.
Representative C u r t i s . It would be interesting to have that
figure, because nationally the service and distributive area has been
increasing much more rapidly than employment in manufacturing.
And in manufacturing itself the national trend shows a shift from blue
collar to white collar, which is another important trend.
Mr. C r e a m e r . Yes. Well, these trends have appeared in Rochester.
If you compare the 1950 census of occupation with the 1960, you do
find this.
145
JOB VACANCY STATISTICS
Mr. M y e r s . In August of last year 53 percent of the vacancies
were in manufacturing.
Representative C u r t i s . And what is manufacturing as a percentage
of the employment in Rochester?
That would give us a pretty good indication.
Mr. M y e r s . It is approximately the same.
Representative C u r t i s . In other words, this conforms to the full
labor market?
M r . M y e r s . Yes.
Representative C u r t i s . That is very interesting.
Mr. C r e a m e r . On the point of usage, I would like to read from
another letter from the manager of the industrial management council.
On the receipt of its first letter, I asked his permission to present it to
the committee if the occasion arose. He writes then on May 11:
I was interested in the fact that John and you are to appear before the Sub
com m ittee on E conom ic Statistics of the Joint Econom ic Com m ittee relative to
the feasibility and usefulness of jo b vacancy information. Certainly there should
be no doubt as to the usefulness of such information.
Since you com pleted your survey of the Rochester area we have had many
occasions to use the statistical data that your survey provided. Such informa
tion has been particularly useful with our area educational institutions and with
^agencies and employers interested in manpower developm ent programs. W e
w ould have no objection to your introducing our letter indicating that we intend
rto continue jo b vacancies surveys among our member firms.
Chairman P r o x m i r e . It would be helpful if we could get some
iurther and more specific information from the schools, and so forth.
You raise a helpful point here, that it appears to be not the employer
alone who would benefit, but the worker and the school which would
train him.
Mr. M y e r s . And then indirectly the employers.
Chairman P r o x m i r e . However, we do not have the kind of evi
dence before us right now that supports that as much as I think the
committee should have. So that if you can get a little more informa
tion, that might be helpful too—if the schools, in other words, could
indicate why they say that, if they have instituted programs, and so
forth.
Mr. C r e a m e r . Senator, we submitted a questionnaire to 60 com
munity organizations. About 40 of them I think were schools. This
is the same group to whom we sent a summary of each of our three
surveys. So they had some idea as to what was in it, and presumably
had found out what was useful to them in the 5 or 6 months over
which this operation took place.
Now, as I said, we sent this opinion questionnaire to them.
This is analyzed in our final report. And we would be happy to
submit for the record that part of the report that analyzes these
answers.
(Material referred to follows:)
S
um m ary of
O p in io n s
of
E m p l o y e r s a n d C o m m u n it y O r g a n iz a t io n s
U ses o f Job V a c a n c y D a t a
on
L
ocal
A t the outset some of the potential uses o f a continuing survey of job vacancies
-were mentioned. H ow much of these potentials can be realized must wait upon
the experience with a regular reporting program. In the meantime, however, it
has been possible to proceed a few paces beyond the speculations of the research
analysts on certain of the hoped-for benefits. This is accomplished by reporting
the opinions of those in Rochester who received the survey results.
146
JOB VACANCY STATISTICS
These are several facets, at least, to the idea of usefulness of job vacancy
statistics. One facet is its usefulness for manpower programs in specific labor
areas— job placement, vocational guidance, counseling and training, and, for
employers, short-run manpower planning. Still at the local level is the analytical
usefulness of these statistics in furthering the understanding of how a specific
labor area functions. A t the national level its usefulness turns largely on what
these data can contribute to the formulation of econom ic policy for prom oting
econom ic growth with minimal short-term fluctuations. Here the contribution
might well be in matters relating to timing and direction of specific changes in
monetary and fiscal policies as well as in the more precise articulation of a federal
manpower program. Our own opinion survey was directed exclusively to the
views on the usefulness o f job vacancy data for local manpower programs— those
of particular employers and community organizations. The opinions of employers
and com m unity organizations, mostly schools, are presented separately.
The results of each of the three surveys, as previously noted, were summarized
and copies were sent to each of the respondents and to about 60 schools and c o m
m unity organizations in Rochester but not in the survey sample. Late in October
1965, after the results of the August survey had been circulated, this group of
nearly 500 was sent a questionnaire entitled “ Opinion Survey on Usefulness o f
Job V acancy Statistics.” The substantive paragraphs of the covering letter
n oted:
“ B y this date, you have received summary reports on three surveys of job
vacancies (February, M ay, and August) in M onroe County. These surveys are
part of an exploratory effort to determine the feasibility of a permanent reporting
program. Because of its experimental character and national importance, our
study has had the financial support of the Ford Foundation.
“ The next stage is to prepare a report on our conclusions and recommendations
on the feasibility of a permanent program. In this connection it would be m ost
helpful to have your opinions, based on your reading of the three summary reports.
W e would be obliged therefore if you would be good enough to fill out the brief
questionnaire that is enclosed. A stamped, addressed envelope is also included
for your convenience.
“ Needless to say, your frank views are being solicited. Please be assured we do
not take offense at criticism.”
O P IN IO N S O P E M P L O Y E R S IN
JO B
VACANCY
SAM PLE
The questionnaire was mailed to 405 respondents and 183, or 45 per cent
returned usable schedules with one follow-up effort. In general, the larger the
com pany the higher was the percent returning the questionnaire. W hen the
usable schedules are weighted b y August 1965 employment, the replies cover tw othirds of total em ploym ent of sampled employers.
Number of questionnaires
Percent
returned
Number of employees
Mailed
Oto 9______________________________________
10 to 19_______________________________________
20 to 49________________________________________
50 to 99________________________________________
100 to 249______________________________________
250 to 999______________________________________
1,000 to 2,499__________________________________
2,500 or more________________________ _____ __
Total________________________ _
_
1 Based on employment in mid-August 1965.
-
Returned
Respondents
employment 1
as a percent
of total
employment
112
39
51
35
52
82
18
14
32
11
18
20
30
55
11
6
28.6
28.2
35.3
57.1
57.7
67.1
61.1
42.9
27.4
28.0
35.9
57.2
58.3
67.9
59.2
68.4
405
183
45.2
65.7
147
JOB VACANCY STATISTICS
The first item attem pted to determine how many read the reports and with
what care. It was phrased as follows:
“ 1. Y ou have participated in our surveys in M onroe County and received our
reports summarizing the results of each survey. Please indicate, b y checking the
appropriate boxes, whether or not you have read the reports.
1st report on
February
survey
H ave not read the report__________________
Have read the report—
Carefully______________________________
Casually_______________________________
2d report on
May
survey
Sd report on
August
turvey
□
□
□
□
□
□
□
□
□
I f the four respondents who failed to answer question No. 1 are excluded, 179
returned an answer. Since each employer received three reports, the maximum
number of reports to be read was 537. Of this number, 53, or 10 per cent, were
not read. Fifteen respondents, accounting for 45 of the 53 reports, said they
read none of the three reports. About one-quarter of the reports were read care
fully and nearly two-thirds casually. The answers classified b y the size of the
respondent (size measured b y number em ployed in August 1965) is revealing.
The per cent of respondents in each size-class that read at least one report care
fully rises sharply as size of firm increases— from 10 per cent for those employing
less than ten persons to 77 per cent for those employing 1,000 or more. Of the
15 who read none of the reports, 8 were firms employing less than 10 persons and
no firm em ployed as many as 250.
Number of reports read
Number of employees
At least 1
but none
carefully
None
At least 1
carefully
Number of
respondents
Percent
reporting
at least 1
report read
carefully
0 to 9___________________________
10 to 99_________________________
100 to 249_______________________
250 to 999_______________________
1,000 or more---------- -------------------
8
5
2
0
0
20
32
17
24
4
3
11
10
30
13
31
48
29
54
17
10
23
34
56
77
-----------------
15
97
67
179
31
Total-----------
This suggests that the small employers, though they are characterized b y the
highest jo b vacancy rate, have little need for manpower planning because of the
smallness of the absolute numbers of unfilled jobs. H aving no felt need, they
have little interest in, and probably still less time for, reading the results of a
jo b vacancy survey. On the other hand, the larger employer— those with a labor
force of 250 or more— are more apt to have a personnel office with a professionally
trained staff with the need, interest, and time to acquire jo b vacancy inform ation.
The em ployer’s opinion on the usefulness of the survey data was given in re
sponse to the following set of three questions:
“ 2. If jo b vacancy inform ation were collected each quarter on a continuing
basis and statistics on job vacancies by occupation made public within 30 days
following the survey date, do you think the inform ation w ould be useful—
(a) to you in the day-to-day operations and planning of your organization?
Yes □
No □
If yes, please specify how':
(b) to com m unity organizations concerned with manpower planning or
training?
Yes □
No □
If yes, please specify how :
(c) for another purpose?
Yes □
If yes, please specify:
No □
148
JOB VACANCY
STATISTICS
(A) DIRECTLY USEFUL TO EMPLOYES
There is a clear and positive association between the per cent answering yes
and the degree of care in reading the reports. For example, all fifteen who
read none of the reports consistently replied that the survey would not be useful
in their own operations or were undecided.1 One-eighth of those who read at
least one report, but none carefully, responded in the affirmative com pared with
slightly more than one-quarter of those who read at least one report with care.2
Whether the reading persuaded some to answer with a yes or whether they read
the reports because they had already reached this conclusion cannot be deter
mined from these data. Of the 170 responding to this question, 18 per cent
answered “ yes,” and 82 per cent “ no.” There is virtually no difference in the
proportions when the yes and no answers are weighted by August 1965 employment.
Reading of reports
Useful to employer
Yes................................................ ...................
No......... .............................................................
None
At least 1 but none
carefully
At least 1 carefully
Number
Percent
Number
Percent
Number
0
12
0
100
21
81
13
87
18
46
Percent
28
72
Nonetheless there is some association between size o f firm and an affirmative
answer. This was implicit from the other relationships between firm-size and
care in reading and between care in reading and affirmative answer. N o firm
em ploying less than 20 persons reported a “ yes;” 11 per cent o f those em ploying
20 to 249 replied in the affirmative; and o f those with 250 or more nearly one-third
reported that the jo b vacancy data were directly useful in com pany operations.
However, am ong the latter a relatively higher proportion were in the size-class
250 to 999 employees.
The distribution o f affirmative replies by industry largely reflects the correlation
between industry and size of firm. Thus nearly 30 per cent o f the manufacturing
firms returned a yes answer compared with 10 per cent o f the trade firms, 8 per cent
o f the service employers, and none in construction.
Of the 30 employers answering yes, all except two responded to the request to
specify how jo b vacancy data would be useful in the operations of their organiza
tion. Aside from a handful of miscellaneous uses, the specific uses may be grouped
under tw o general headings:
1. Better informed about labor market when seeking employees.— Alternate
phrasing of same general idea:
Indicate supply shortages
Show employm ent conditions
Indicate expected com petition for particular jobs
Provide a “ feel” of the labor market
Show trend
Estimate competition to be met in securing labor
2. Planning and recruitment.— Alternate phrasing of same general idea:
Adjust recruitment to conditions
Show need for developing alternative plans for filling vacancies
W ould help decide how to fill vacancies
M ay show need to revise jo b qualifications
Provide a better idea of area needs
Training programs
Planning and timing of reductions and additions to staff
3. Miscellaneous.—
Indicate whether vacancies in one firm are com m on to others
Indicate occupations where automation should be considered
Union negotiations
Could assist in bringing workers in shortage occupations to Rochester
1 Those returning a schedule but not answering this question are entered as n.a. and interpreted to be
undecided.
2 The undecided are excluded from the totals in computing these percentages.
149*
JOB VACANCY STATISTICS
W ith one or two exceptions, these appear to reflect genuine uses and can be
taken to be fully responsive. W hy then did not more employers, at least among;
the larger ones, report an affirmative answer? For some the exposure to this in
form ation may have been too brief for them to realize its possible uses in com pany
operations. Others may be obtaining much the same inform ation in another way.
In this connection, it is relevant to recall the role of the Industrial Management:
Council (IM C ) in the functioning of the Rochester job market.
In chapter 2, it was suggested that the weekly meeting of personnel managers*
under its auspices m ight render the jo b vacancy survey data of little use to these
member companies. For this reason the opinions of members of the IM C and
nonmembers are shown separately for the manufacturing companies who comprise
the vast m ajority of IM C membership.
Percent answering
Total
Yes
Manufacturing firms with membership in IMC— ____________
Other manufacturing firms ____________________ _____________
27
35
No
37
23
63
77
Actually a higher percentage of the IM C members returned an affirmative
answer than did the non-members. However, the latter tend to be the much
smaller firms. Even so, it is quite possible that most of the 17 IM C firms answer
ing with a “ no” may well have responded “ yes” if the weekly meeting of personnel,
managers were not held.
(b)
u s e f u l n e s s to c o m m u n it y o r g a n iz a t io n s
The preponderance of em ployer opinion swings sharply in the other direction
when the usefulness of job vacancy data to com m unity organizations is the
question.3 Of the 179 respondents, 34, or 19 per cent, did not reply to this partic
ular query but of the remainder of those that did reply, 72 per cent expressed
the view that job vacancy data would be useful to com m unity organizations for
a variety of purposes.
The specific uses enumerated conform closely to the expected uses as set down
in Chapter I. Seventeen respondents, for example, specified the use of these
data “ to indicate areas of continuing shortages as guides to program s.” Another
15 expressed much the same view by saying this inform ation would help “ to adjust
training programs to com m unity needs,” including programs for “ unem ployable”
and dropouts. Some particularly stressed the use of these data in vocational
planning in schools. Others emphasized the assist given to job placem ent:
“ helping high school students;” “ show the individual where he may be needed;”
“ unem ploym ent boards would know there are jobs available;” “ indicate jobs
requiring little education and experience” and this inform ation would “ help
em ploym ent agencies, public and private.”
Several noted administrative uses. The number of unfilled jobs and per
sistence of particular shortage occupations could serve to evaluate the effective
ness of current manpower programs. The same sort of data could also be used
to support the need for expanded manpower programs. And one, obviously
with training in economics, mentioned that such inform ation could “ show the
need for higher pay in ‘tight’ occupations.”
OPINIONS OF SCHOOLS AND COMMUNITY ORGANIZATIONS NOT IN JOB VACANCY SAMPLE
For the most part the opinions just summarized are those of employers in private
business and reflect their views on how other organizations— schools and other
com m unity organizations— could use job vacancy inform ation reported on a
regular basis. Even more pertinent, however, are the opinions of the schools and
organizations themselves. The articles that summarized each of the three surveys
were mailed as each was published to 62 organizations in the Rochester area
deemed to have an operational interest in this inform ation. The same roster was
asked to respond to the opinion questionnaire. Exactly half returned schedules
and 25 of the 31 were from schools. As m any as three-fourths read all three reports
* The responses to 2b and 2e are combined since the specified uses in both parts are similar. Apparently
our distinction on the questionnaire between community organizations concerned with manpower planning
or training and other purposes was not meaningful.
150
JOB VACANCY STATISTICS
and all read at least one report, although two gave no answer to this particular
question. Better than half reported reading at least one article with care and
nearly 55 per cent of the 87 articles distributed were reported as being read with
care. This suggests a greater interest in the material than the employers in the
sample have reported.
The relevant question for this group is the usefulness of survey data on jo b
vacancies for com m unity organizations concerned with manpower planning and
training, i.e., questions 2b and 2c of the questionnaire. Since three failed to
answer this question, the number of usable replies is reduced to 28. All bu t one
expressed the view that this inform ation is useful in carrying out some part of
their program objectives. The specific uses that underlay this virtual unanimity
of opinion are those that had been anticipated. A id in the placem ent of students,
both graduates and dropouts, ranked high among the uses. Also frequently men
tioned was the usefulness of these data fror determining the type of technical
and vocational programs that are currently needed as well as providing guidelines
for the long-term planning of high school curriculum, counseling and technical
training.
WILLINGNESS TO PARTICIPATE IN A REGULAR REPORTING PROGRAM
Another test o f employer
tinuing reporting program.
“ 4. W ould you be willing
at quarterly intervals?
Y es______
N o ______
“ 5. W ould you be willing
to a mail questionnaire?
Yes---------N o______
interest is the willingness to participate in a con
T o gauge this interest tw o questions were asked:
to participate in a permanent jo b vacancy survey
to furnish the same basic inform ation in response
W ith respect to the first of these tw o questions, all but 14 of the 183 respondents
gave their answer. Of those replying, 104, or 62 per cent, answered “ yes” — that
is, as of the survey date, this number was willing to participate in a continuing
reporting program on jo b vacancies. Perhaps even more impressive, em ploym ent
in these 104 organizations accounted for about 85 per cent of all em ploym ent
engaged by the 183 organizations.
The com position of the 104 employers who answered yes is worth noting.
Percent willing
to participate
in continuing
survey
Number
Useful in own operations:
Yes_________________________________________
N o_________________________________________________________________
Useful for schools and community organizations:
Yes_________ ________ _______________________ _______________ ____
N o____________________ _________
Not useful to either__________ _________ _
_ __
30
141
100
49
107
36
35
75
28
40
Of those who consider the survey results useful, a large percentage, 80 per cent,
are also interested in being a part of a continuing survey. This may be taken to
mean they judge the benefits to themselves or to the com m unity at least equal to
the costs to themselves in supplying the data. As for the substantial fraction
willing to participate in a continuing survey despite their reported opinion that
the inform ation is not useful, this may be interpreted as their being not sure of
their own position and believe a longer experience is required.
Once again, the lack of interest of the small employers is evident:
Number employed
Oto 19............................... ...................... ..............................................................
20 to 249............... .............. .......... ...........................................
250 or more.
_
_ _______________
Total__
Number
responding
Percent willing
to participate
in continuing
survey
43
68
72
21
60
76
183
60
JOB VACANCY STATISTICS
151
Perhaps more of the small organizations would be willing to participate if they
were aware that their com m itm ent would be limited in time since they would be
members of a rotating sub-sample. In any case, there are sufficient numbers of
small employers to provide a representative sub-sample. Of the larger organiza
tions, those employing 250 or more who were included with certainty in the
Rochester survey, about one-quarter, 10 b y number, reported an unwillingness to
participate. If this proved to be typical in most labor areas, it might pose some
sampling problems but certainly abstentions of this proportion would not be
crippling. Indeed, as experience with jo b vacancy data accumulates, m any of
these m ay reverse their opinion on participation.
The purpose of question 5 was to ascertain whether the employers, once the
survey had been introduced, are willing to use a mail questionnaire in reporting
the jo b vacancy inform ation. This is of some im portance since data collection
b y mail costs less than collection by personal call.4 Of the 104 employers indi
cating a willingness to participate in a continuing reporting program, 101 were
agreeable to using a mail reporting form. This is the evidence for assuming data
collection b y mail in the cost estimate presented earlier in this chapter.
T o summarize, the results of the opinion survey make it clear beyond any dou bt
that the regular reporting of job vacancies with occupational detail for a particular
labor area would be useful and valuable in the operations of schools and com m unity
organizations concerned with manpower planning, training, and placement.
This is the view expressed by almost all the schools and organizations and by a
large m ajority of private employers. However, when the issue is the usefulness
of the same information to employers in the day-to-day operations and manpower
planning of their own organizations only about 20 per cent answered in the affirma
tive. The affirmative view was most widespread among the medium-sized em
ployers. It was suggested that many of the larger employers found the inform a
tion superfluous because much the same inform ation was obtained through the
weekly meetings of personnel managers under the auspices of the Industrial
Management Council. While this arrangement may not be unique to Rochester,
neither is it widespread. The Rochester findings on this point therefore probably
are not representative of how the larger employers think about this matter in most
other labor areas.
Interest in a continuing survey is not necessarily equated to views on its use
fulness, at least in the early stages of a new reporting program. Thus, while 45
per cent of 183 employers reporting considered the survey results either useful to
themselves or to schools and community organizations as many as 60 per cent
expressed their willingness to participate in a continuing survey. M oreover, since
the negative attitude was most prevalent among the small employers we conclude,
if the Rochester survey results are typical, that there is sufficient interest to
support a statistically valid reporting program on job vacancies.
Before one passes judgm ent on whether the uses justify the estimated cost, it is
im portant to remember that the uses that have been discussed— those in a local
labor area and restricted there to employers, schools, and com m unity organizations
— do not cover the whole gamut of uses. Even at the local level mention must be
made of the possibility that job vacancy data could significantly im prove the
effectiveness of the placement. A t the national level one should remember the
possible contribution these data can make to policy formulation. Finally, there
is the area of labor market analysis. These data together with other data already
available may well assist us to extend our understanding of how particular labor
markets operate. The N IC B has already embarked on one test of this possibility.
In our view the docum ented uses and the additional expected uses am ply
justify the estimated costs. Whether or not this sum w ould yield more benefits
if it were expended on filling other informational gaps on the supply of and dem and
for labor requires a judgm ent which we are ill qualified to make. W e therefore
leave the decision to others with longer and deeper experience in this area of
knowledge.
Chairman P r o x m i r e . D o you think that would give us the kind o f
information that would document the desirability of this for educa
tional institutions?
Mr. C r e a m e r . It would go some distance toward that.
Chairman P r o x m i r e . Give u s what y o u have.
<We note again our earlier reservation about possible deterioration in the quality of the survey results
obtained by mail.
152
JOB VACANCY
STATISTICS
Mr. C r e a m e r . We did ask them how, in what way was it helpful
to them. And we jotted down some of the phrases that they used,
that sort of thing.
Chairman P r o x m i r e . Mr. Goldfinger?
Mr. G o l d f i n g e r . A question, sir.
In that connection I wonder whether the Labor Department has
used to any degree its own pilot studies in connection with vocational
training, MDTA training programs, and so on.
Mr. K n o w l e s . My understanding of this is yes, though not in
every area they surveyed, Mr. Goldfinger. I think that this varied.
I am not aware of the reason why it varied. But my impression from
the testimony and the conversation was that in at least two that I can
recall being mentioned—Minneapolis-St. Paul, or Milwaukee, and
Portland—they actually used the information. They experimented
with actually using the data that came in in the office itself, since the
survey was done by the local employment service office.
Chairman P r o x m i r e . I do not know if this is responsive to your
question, but witnesses yesterday said in Milwaukee there are 3,000
young people who are being trained in 12 different occupations as a
result of the specific survey that they made in Milwaukee.
Representative C u r t i s . There was testimony along that line.
I raised the question, too, whether they used the data in implement
ing the new immigration laws, which are based on whether or not there
is a job vacancy, or rather whether there is a need for these skills
within our society. And the answer was “ to some degree.”
Mr. C r e a m e r . A s part of the 16-area survey program each State
agency was asked to report on what uses they had made of this infor
mation. And in the summary that I saw a number of State agencies
said it was useful for the sort of programs you are referring to. But I
do not think——
Chairman P r o x m i r e . In connection with the Rochester study,
were there any specific instances of the use, such as stepped up training
programs in certain areas, or anything of that kind?
Mr. C r e a m e r . None has come to my attention as being that
specific. But we did have letters from the vocational guidance
counselors in the schools saying that it was very helpful to them.
Chairman P r o x m i r e . Y o u see, it is so much more useful if they can
be specific and indicate how they used it, because people always like to
be nice, and if you ask them, “ Is our work helpful to you,” they will
say, “ Sure.”
Mr. C r e a m e r . This question was on our schedule, “ Specify how.”
But we received rather general phrases in reply.
Representative C u r t i s . The best test would be to ask if they would
be willing to pay something for it.
Chairman P r o x m i r e . Or share in it.
Feasibility of a continuing survey
Mr. C r e a m e r . The principal conclusion of this report is that
accurate data on job vacancies can be obtained from sample surveys
of employers. This is the consensus of the staff, both interviewers
and analysts, concerned with the NICB surveys. The costs of data
collection, outlined briefly in the preceding section, do not seem ex
orbitant compared to the possible benefits that may be derived,
particularly in comparison to other data collection programs on the
JOB VACANCY STATISTICS
153
current status of the labor force. Another dimension of feasibility
is the willingness of employers to cooperate in a continuing data
collection program. We have asked employers in the Rochester
area about their willingness to furnish such data quarterly on a long
term basis. Of those responding to the question, 62 percent replied
“ yes” ; these accounted for 85 percent of the total employment of
those responding to the question.
There is more interest in job vacancy data among large than among
small employers, as evidenced by attention paid to published reports
and willingness to cooperate in the collection of the data. This is,
in part at least, the result of the view more common among large
than among small employers that the data are valuable either directly
or to the community. Very few small employers said this was of use
to them or to the community. Almost all of those who are willing
to cooperate indicated that they would reply by mail, although this
often takes more of an employer’s time than does a personal interview.
In summary, a nationwide data collection program appears to be a
feasible operation—perhaps in view of Mr. Goldfinger’s comments
that is not a careful wording. When we say nationwide data collec
tion we mean by the local areas. This appears to be a feasible
operation—the costs seem to be within acceptable limits, and the
benefits in terms of supporting placement activities, a variety of man
power programs, and formulation of economic policy are at least
commensurate with the estimated cost.
Moreover, there seems to be no adequate substitute for job vacancy
statistics to judge by our limited analysis of two sets of statistics
considered by some as possible proxies. One is an index of help
wanted advertising in newspapers, a continuing series now compiled
by the NICB for 52 cities, including Rochester. This series is based
on a count of the number of advertisements appearing in particular
newspapers.
The Conference Board has a help wanted index for 12 of the 16
cities surveyed by the BES for job vacanacies. For these 12 cities
we compared the percent change in job vacancies between survey
rounds to the percent change in the help wanted index for the same
period. In one-third of the cities there were increases in the help
wanted index but decreases in job vacancies. That would have given
you a completely wrong reading even as to direction of change. In
four of the remaining eight cities the help wanted index would have
substantially overstated and in one substantially understated the rise.
Thus in only 3, or one-fourth of the 12 cities, would the relative
change in the help wanted index have approximated that of job vacan
cies. Even in these three cities the index could not provide the
number of job vacancies unless there were repeated job vacancy
surveys to establish benchmark numbers.
In Rochester one would not expect a close relationship between
the index and job vacancies because the larger employers, mostly
engaged in manufacturing, are affiliated with the Industrial Manage
ment Council which requests its members not to advertise for em
ployees in the Rochester newspapers. Accordingly, we tested the
Rochester index as a predictor for job vacancies in the nonmanufac
turing sector for two periods—mid-February to mid-May and midMay to mid-August. In the first period it projected a 50-percent
increase compared with an actual decrease of 7.6 percent in job vacan-
154
JOB VACANCY STATISTICS
cies in the nonmanufacturing sector; in the second period the index
was within acceptable range of the target—a projected decrease of
10.8 percent compared with an actual decrease of 12.3. However, a
50-percent probability of success in this matter is not acceptable.
Moreover, should it ever prove to be an adequate predictor of the
number of vacancies, essential information on occupations would be
still lacking.
(The following additional material relating to the preceding testi
mony was later submitted:)
H
elp- w anted
A d v e r t is in g *
Placing a help-wanted advertisement in a newspaper entails an outlay of m oney
by an employer. This action indicates active recruiting for a worker. An index
o f the total number of help-wanted advertisements for a given city or for certain
occupations appears at first glance to be a possible proxy measure of job vacancies.
This, indeed, will be the case if the ads placed in a newspaper for workers are
representative of all efforts to secure workers. From our investigation, however,
the representativeness of such an index as a proxy measure of job vacancies is poor.
The Conference Board maintains a series on help-wanted advertisting, hereafter
referred to as the N IC B index. This series is a monthly tally of the number o f
help-wanted ads placed in newspapers located in 52 cities in the United States.1
As part of the job vacancy study, we undertook a detailed counting of help-wanted
ads in the Sunday Rochester Dem ocrat and Chronicle by occupation group, here
after referred to as the Sunday index, to see if there was any promise in using this
additional information to project job vacancies in the aggregate or by occupation.
A tally was made of only the Sunday papers for two reasons, the am ount of time
needed to prepare a tally (3 hours a paper) and the apparent representativeness of
Sundays to the daily count.
T
able
A .l — N IC B and Sunday indexes, Rochester Democrat and Chronicle
Index (September 1964=100)
NICB index (daily).- __ ________________
Sunday index (Sunday only)_____________
September
100
100
October
90
90
November
December
76
76
56
58
January
80
84
Table A .l shows the N IC B index and the index for an average of the Sunday ads.
While the period of comparison is short, the tw o indexes appear to m ove closely
together.
W e divide the ads into male and female, agency and non-agency. Our occu
pational breakdown conforms to the pattern of the Dictionary of Occupational
Titles, Second Edition. W ithin each major group, detailed occupations are
listed for those titles frequently mentioned in the Rochester newspaper (e.g.
Drivers). Counting want-ads by occupation leads one to make many arbitrary
decisions, owing to the nature of the ads. Agency ads are handled in a different
manner than non-agency ads. M ost agency ads do not mention a num ber
sought, only a plural indication of job, for example, engineers, typists, etc. T o be
consistent, therefore, only one entry was made for each occupation group men
tioned, regardless o f the number of persons desired in the ad. For example, an
ad stating that an agency wanted 34 typists, 2 stenographers, and 20 salesladies
would be tallied as one clerical worker and one salesperson, since these are the
only two major groups represented. For non-agency ads stating that more than
one person was sought in the same occupation, the exact number was tallied and
the excess over one entered in a “ multiple job ads” column. Ads written in the
plural not specifying the exact number sought are counted as ads for one person.
While this system surely yields an incorrect measure, we could not devise a better
substitute. All ads for part-time and temporary help are counted. Some ads
clearly state that accepting a position will require a locational change; these are
tallied as “ O ut-of-tow n” ads.
♦Prepared by Richard Towber of the staff of The Conference Board.
* For a detailed account of the help-wanted index, see “ The Conference*Board’s New Index of Helpwanted Advertising.” Technical Paper Number Sixteen. National .^Industrial ^Conference Board, Inc..
New York.
155
JOB VACANCY STATISTICS
The Sunday index tally differs from the N IC B tally in several respects. For
the Sunday index we counted the actual number of persons sought in a non-agency
ad. In the N IC B index, each ad placed is counted once. T o enable us to recon
cile one tallying m ethod with the other we have a “ multiple jo b ads” colum n,
which indicates the number of jobs we had, in excess of one per ad. This more
elaborate scheme did not affect the changes in the number o f ads placed over time,
since both series m oved closely together. The purpose of “ multiple counting”
of help-wanted ads is to obtain a better occupational distribution.
Help-wanted ads by occupational detail
Total
Male
Female
Total Agency
Total Agency
Total Agency
Total..
ProfessionalProfessional___________
Semiprofessional_______
Managerial and official-.
Clerical and sales.
Clerical___
Secretary..
Sales--------Technical..
Service _
Domestic____________ _____
Babysitters..----------------------Personal________________ ___
W aiters (resses)_____________
Protective__________________
Building service and porters.
Agricultural and horticultural.
Skilled________ - ____________ Painters____________________________ _____ ______
Carpenters..----------------- -------------------- ----------- ----Masons______ ____ ____________ _____ ___________
Plumbers______________ _________________ ____
Tool and die makers and machinists_____________
Mechanics, repairmen, maintenance and installa
tion workers____ , ______ :----- __________________
Welders____ ______________________ ____ ____ _—
Auto mechanics------ ---------------- --------------------------Semiskilled.
Auto doll-up men and metal polishers___
Gas station attendant___________________
Screw machine operators--------------------Sewing, stitching, mending, seamstress..
Drivers (truck, auto, etc.)------------- --------Unskilled___________
Unclassified_________
Out-of-town ads^-----Multiple job ads..
HELP-WANTED ADVERTISING AS A MEASURE OF AGGREGATE JOB VACANCIES
Help-wanted advertising as a measure of aggregate vacancies is restricted in
Rochester because of an agreement among members of the Industrial M anage
ment Council of Rochester (an employer organization com posed mainly of manu
facturing companies) not to advertise for vacancies in the local newspapers. For
that reason we would not expect newspaper advertising to be a good proxy for
job vacancy data in all industries in this area.
156
JOB
VACANCY STATISTICS
A measure of the use of newspaper advertising to hire workers was collected in
ou r preliminary job vacancy survey (Septem ber-O ctober 1964).
T a b le
A . 2 . — Use of newspaper advertisements by occupation, 27 firms, SeptemberOctober 1964
Percent of total vacancies
O ccupation group:
advertised in newspapers
Professional, semiprofessional, and managerial workers_______________
30. 1
Clerical and sales workers_____________________________________________
9. 3
Service w orkers________________________________________________________ 84. 6
Skilled workers________________________________________________________
25. 8
Semiskilled workers------------------------------------------------------------------------------10. 6
Unskilled workers_____________________________________________________
8. 6
T o ta l________________________________________________________________
25. 1
Source: Conceptual and Measurement Problems in Job Vacancies: A Progress Report on the NICB
Study, table 3, op. cit.
Newspapers were used as one method of hiring for 25.1 per cent of vacancies
existing at that time. The only occupation widely advertised in newspapers was
service workers (84.6 per cent of total vacancies).
The N IC B index, when it was used to project job vacancies from mid-February
to m id-M ay, yielded unsatisfactory projections, judging by the estimated number
o f job vacancies based on the surveys. The predicted M ay vacancies were
about 12,000 using the index, compared to a survey count of 8,700 for all em
ployers.
T a b le
A . 3 . — Survey vacancies projected by N IC B help-wanted advertising index
Employers
All
Actual February vacancies_________________________________________________
Predicted May vacancies___________________________________________________
Actual May vacancies____________ _________________________________________
Predicted August vacancies________________________________________________
Actual August vacancies__ ________________________________________________
7,947
11,935
8,776
7,828
8,568
Nonmanufac
turing
4,926
7,398
4,554
4,062
3,995
The predicted number of vacancies for August was 7,800; the survey result was
8,600 vacancies. While the index predicted the correct direction of change in
vacancies, the magnitude was overstated by the index. For nonmanufacturing em
ployers, survey vacancies decreased from February to M ay, while the index pre
dicted a rise in vacancies. The M ay to August prediction is however, correct in
both direction and magnitude. It may be of some value to look at these compari
sons when more data are available. Our results suggest, however, that helpwanted advertising is a poor proxy for the aggregate number of job vacancies.
Further negative evidence on the use of a help-wanted index as a proxy measure
is obtained by looking at similar sets of data for the cities surveyed by the Depart
ment of Labor. The Department of Labor conducted job vacancy surveys in
16 areas during the last quarter of 1964 and again in April 1965.
157
JOB VACANCY STATISTICS
Chart A -l
Per Cent Change in Vacancies and Help-W anted Index (U nadjusted)
12 Selected Cities— All Employers
-1.0
-30
-2 0
-1 0
0
10
2Q
30
i£>
$0
Per Cent Change in Help-Wanted Index
Percent change in vacancies and help-wanted index, all employers— 12 selected cities
Percent change
Area
Vacancies
Providence.. . _________ ________ __ __________ ___ __ ________
________ __ ________ ________ _ _
.
Kansas City___ __________
Minneapolis— _ ___________________ _ ________ _______ _ ____________ __
New York— ________ ________ __ _
_ ______ _______
____ _______ _ .
Philadelphia_______________________________________________________________
Miami_____ _____________ ___ ___ ______________ _______________________
_ _
____
Richmond_____ ____________________ _ _ _________
Chicago___ __________ _______ __ __________ __ _ _____ _____________
M ilw aukee.____ _________ ______________
_ _________ _ _ __________
Birmingham
________ ______________
________________ _ _____________
Los Angeles______ __ _______________
_ __
____ ___ ___________ ___ _
Baltimore_________ _______ _________________________ ____
_
__
63-947 0 — 66 ------ 11
337
-444
177
77
237
-324
385
163
-4 2
-395
31
136
Help-wanted
advertising
205
269
414
307
192
84
320
359
338
475
131
187
158
JOB VACANCY STATISTICS
Ch art A -2
Per Cent Change in Vacancies and Help-Wanted Index
Non-Manufacturing Employers— 12 Selected Cities
-ijd
-30
-20
-10
0
10
20
ko
30
5o
Per Cent Change in Help-Wanted Index
Percent change in nonmanufacturing job vacancies and help-wanted advertising
12 selected cities
—
Percent change
Area
Vacancies
__ - - - _________ _____ __
Providence
Kansas City___ ____
____ ____
____ _________________ _______ - - ____
Minneapolis
______
___ _____ _____
_______ ________ _____ ___
New York______ ____
___ __ _ __
__
_ ___________ . . _______
Philadelphia
_
_ - ____ . . . ____________ _________ - __
Miami
_ __ ____ ____ _______ _ _____________________
_______
Richmond. _________
__ ____ ______ _ _______________ _____________ -Chicago
. . ___ ___ _____________ - . ____________ ___________
Milwaukee
-- ____ _______ ________ ______ ______
_______ _______________
___________ ____- -Birmingham
_ _ __
Los Angeles
-- .
. . __________ ___ __________ - Baltimore _ _
____________________ -- ___________ _____ ____________
413
-506
121
183
287
-275
382
99
282
-446
76
153
Help-wanted
advertising
205
269
414
307
192
84
302
359
338
475
131
187
JOB VACANCY STATISTICS
159
Tw elve o f the areas are covered by the N IC B index. For these twelve cities
we have plotted the per cent change in vacancies, and the per cent change in the
N IC B index. The optim um points will lie along a 45° line, representing points
where the per cent change in vacancies and the index are equal. There are four
cities for which there was a positive change in the N IC B index and a negative
change in the number of vacancies (Chart A .l). For eight cities the change for
the index and number of vacancies were both positive. Of these, the N IC B
index understated the changes in vacancies for five cities, and overstated the
change for the remaining three.
On the assumption that manufacturers generally make little use of newspaper
help-wanted ads we compared relative changes in nonmanufacturing jo b vacancies
and the N IC B index for the same twelve cities (Chart A .2). Three cities showed
a positive change in the N IC B index and a negative change in the number of jo b
vacancies. For nine cities, both the N IC B index and the number of vacancies
showed positive changes between the tw o vacancy surveys. The N IC B index
overstated the change for six cities and understated it for three. However, the
magnitudes were closer for nonmanufacturing employers than for all employers.
While at present our conclusion must be negative, some further testing may be
called for when more data are available.
USE OF HELP-WANTED ADVERTISING FOR PREDICTING VACANCIES BY OCCUPATION
Although the use of help-wanted advertising as a proxy measure of total vacan
cies does not seem satisfactory, it is possible that it can be used to estimate the
occupational distribution of the total. Y et another possibility is that relative
changes in the number of jo b vacancies in a particular group can be projected by
the per cent change in the want ads for that occupation group.
T
able
,
A .4.— Occupational distribution job vacancy surveys and N IC B index
February and M ay 1965
February
Occupation group
Sunday
index
Professional___________________________________
Semiprofessional______________________________
Managerial-------------------------------------------------------Clerical__________________ __________________
Sales____ _____________________________________
Service________________________________________
Skilled_____ __________________________________
Semiskilled___________ _______________________
Unskilled_____________________________________
Total___________________________________
,
May
Job vacancy
survey
Sunday
index
Job vacancy
survey
5.1
4.1
5.1
24.0
19.4
18.0
13.0
7.9
3.1
26.6
4.1
2.3
9.4
5.0
7.0
17.4
21.6
6.3
4.8
3.5
3.7
17.0
13.6
22.0
17.9
11.5
6.1
17.0
4.9
1.8
13.7
6.9
9.2
17.0
18.7
10.9
100.0
100.0
100.0
100.0
The differences in occupational distribution are substantial for professional,
clerical, sales, service, and semiskilled workers, between want ads and survey
findings. (Table A.4.) Professional and semiskilled workers are greatly under
represented while clerical, sales, and service workers are overrepresented in
newspaper advertising for both February and M ay. Substantial differences
were also found between the occupational distributions o f want ads and the
survey findings for nonmanufacturing employers.
The Rochester evidence
suggests that estimates of the occupational distribution of total vacancies by
want ads will be poor.
W e have also projected the numbers of vacancies in each of the nine occupation
groups by the per cent changes in want ads for the corresponding groups from
February to M ay. In every case, the projected number differed considerably
from the number found in the M ay survey.
CONCLUSION
The preceding analysis indicates that the use of a help-wanted advertising
index as a proxy measure for job vacancies, in total or by occupation, will not
prove satisfactory. While the data used for comparison purposes are limited,
they show no relationship between help-wan ted indexes (the N IC B index, or
160
JOB VACANCY STATISTICS
the Sunday index) and number o f job vacancies, whether considered in the
aggregate, for individual occupations, or for nonmanufacturing firms, except in
Rochester between m id-M ay and mid-August.
While further exploratories
may be worthwhile when additional job vacancy data becom e available, the
prospects are not bright for developing a useful proxy measure from help-wanted
ads.
An interesting fact is that over several years the help-wanted advertising
series has m oved inversely but in close conform ity to the unem ployment rate.2
Thus it appears that the index is more related to unem ploym ent rates than to
vacancies.
The second measure has the important advantage of providing
occupational detail. It consists of the unfilled job orders on file
with the public employment service on a given date. By means of a
special tabulation provided by the offices of the employment service
in Monroe County, comparisons of the number of unfilled job orders
and the number of job vacancies as of mid-May and mid-August
were made. On the first date, unfilled job orders amounted to 53
percent of estimated job vacancies and on the second date, only 36
percent. The comparable fraction for 14 of the areas surveyed by
the BES was one-third. There is little reason for believing that this
performance will appreciably improve in the foreseeable future. As
to occupational composition, the record is no better. In the Rochester
area, there were substantial differences in the occupational composi
tion of unfilled job orders and job vacancies, even at the level of broad
occupational classifications.
Now, this differs from the finding reported by Mr. Chavrid yester
day for—either 14 or 12 areas, it is not clear to me. His percentage
distribution by occupation is an aggregate for all the areas combined.
And I suspect that if we had available to us the distribution for each
of the areas, you would find that there was much less conformity in
the occupational distribution of the unfilled job orders on hand on a
particular date and the number of job vacancies obtained from the
survey. So I think it would be much closer to the Rochester finding
than are the findings for all areas combined. There probably have
been offsetting changes within the 14 areas that give you the rather
close fit that Mr. Chavrid reported.
(Additional documentation of the preceding, later submitted,
follows:)
J ob
O r d e r s O u t s t a n d in g
W
it h
the
E
m ploym ent
S e r v ic e
The recently awakened interest in job vacancy statistics has led to a quickening
of interest in some of the operating statistics of U.S. Em ploym ent Service and its
State affiliated offices. Of particular interest is the possibility that the statistics
on unfilled job orders on file with the local Em ploym ent Service offices would
prove to be a good proxy for job vacancy data collected from a sample of em
ployers. Wherever this proved to be the case, there would be no need to collect
the data by survey— the resulting econom y to the employer and the governm ent
would be considerable.
In an institutional setting that required that all unfilled jobs be registered with
the public em ploym ent services and all placements required the intercession of
the same offices, the number of unfilled jobs on a given date would constitute the
universe of unfilled jobs, provided the requirements were effectively enforced.
In these circumstances a sample survey of employers would be entirely redun
dant. This is an extreme case which probably does not exist in any nontotalitarian country even where compulsory registration of jobs is a statutory require
ment as in France— usually the enforcement is lax.3 The relevant question then
2 The Conference Board; New Index of Help Wanted Advertising, op, cit., chart 5, p. 17.
3 For evidence on this see Jacques Chazelle, “ The Collection and Utilization of Job Vacancy Data in
France,” in The Measurement and Interpretation of Job Vacancies, A Conference Report of the National
Bureau of Economic Research (1965), pp. 237-264.
JOB VACANCY
STATISTICS
161
becomes how wide a departure from this extreme can there be and still have the
list of unfilled job orders be as complete and accurate as those derived from sample
surveys of employers?
M ore specifically, does the system of voluntary notification of jobs with the
E m ploym ent Service— the arrangement that prevails in the United States—
result in a complete or even representative coverage of unfilled jobs. Obviously,
this will depend on the employers’ opinion on the effectiveness with which the
Em ploym ent Service in a particular area fulfills the recruitment, screening and
referral functions. T o the extent that employers base their opinions on strictly
business considerations, their attitude towards dealing with the local office of the
Em ploym ent Service will depend on the competence and morale of the E m ploy
ment Service personnel. Since the latter, of course, are bound to vary from area
to area, adequacy of unfilled job orders as a proxy for jo b vacancy statistics can
best be determined area by area. The following describes our effort to determine
this for the Rochester Area.
It was part of the original design of the Rochester study to obtain information
on unfilled job orders on file with the Em ploym ent Service offices in M onroe
County for dates that coincided with the survey dates. Our request to the New
Y ork State Em ploym ent Service for these data was readily and graciously granted
and the appropriate arrangements were made at the local level where the co
operation was equally gracious.
The offices of the Em ploym ent Service serving M onroe County provided a
tabulation of unfilled job orders on file in mid-February to coincide with our
survey of job vacancies on the same date. The comparison of tw o series was
presented in the article summarizing the February survey findings which appeared
in the M ay issue of The Conference Board Record. Subsequently we learned that
the statistics on unfilled job orders owing to certain administrative practices were
not the appropriate ones for comparison with the job vacancy statistics. The
February comparison in the Record article, therefore, should be ignored. Un
fortunately the appropriate figures for February could not be derived at a later
date.
The source of the difficulty was the administrative practice of recording a
specific job order calling for more than one person as 1 + . In the February count
these instances were recorded as one unfilled order. Other omissions resulted
from the exclusion of job categories of job orders kept in separate files.
The realization of these omissions for purposes of statistical reporting led to a
revision of recording and tabulating procedures. Technical guidance for the
revision was provided by the Statistical Unit of the Rochester Bureau of M unicipal
Research. As a result of these efforts, the data on unfilled jobs on file with the
Em ploym ent Service in M onroe County in m id-M ay and mid-August 1966 seem
to be accurate. They probably are characterized, however, by some overstate
ment. This arises because a m onth may have elapsed between the tabulating
date and date of the receipt of the order or a telephone confirmation that the
order still remains unfilled. Thus a job order counted as unfilled may have been
filled by an employer through another hiring channel. Another source of over
statement is the practice of counting as unfilled job orders for teachers with a
future starting date who have not yet reported for duty but have signed a con
tract to do so. These would not be counted as a jo b vacancy according to the
definitions used in the job vacancy surveys. T o achieve com parability, therefore,
it is necessary to eliminate these from the count of jo b orders outstanding in the
files of the Em ploym ent Service.
Still another adjustment is required in the interest of industrial com parability—
jo b orders filed by households and by employers engaged in agricultural, forestry,
fishing and mining. These industry classifications were excluded from the jo b
vacancy surveys.
When the official count of unfilled jo b orders as of m id-M ay 1965 are adjusted
for the above differences the following comparison can be made:
All
Unfilled, job orders
___________________ _______
Job vacancies
- __ _________- ____ _____________
Unfilled job orders as percent of job vacancies------------------------
4,660
8, 776
53.1
Present
starting
4,315
7,342
58.8
Future
starting
347
1,434
24.2
Thus, in m id-M ay 1965 the number of unfilled jo b orders in the files of the
Em ploym ent Service in the Rochester Area amounted to only slightly more than
JOB VACANCY STATISTICS
162
one half of the estimated number of jo b vacancies in our survey of a sample of
employers. Clearly, at that particular point in time the number of unfilled jo b
orders on file with the Em ploym ent Service would have given a under-count of the
number of jo b vacancies employers were seeking to fill.
The record, however, worsens at the next point of comparison, mid-August 1965.
On that date unfilled jo b orders numbered 3,113, compared with 8,568 jo b vacan
cies, or 36 per cent of the survey total. The Em ploym ent Service series recorded
a decrease of 33 per cent between m id-M ay and mid-August while the survey series
on jo b vacancies showed a decrease of only 2.5 per cent.
Once again, the Rochester experience appears to be typical of other labor areas.
Thus, the Departm ent of Labor in its unpublished summary of its own jo b va
cancy surveys reports “ On an overall basis, unfilled Em ploym ent Service jo b
openings were equivalent to about one-third of the aggregate total of jo b vacan
cies.” 4 Nothing is indicated about the extent of agreement with respect to direc
tion and degree of change between the first and second surveys.
The conclusion seems inescapable that at the present time and for the foreseeable
future the number of unfilled jo b orders in the files of the Em ploym ent Service
will substantially understate the number of job vacancies. M oreover, as an index,
unfilled jo b orders may overstate the relative change in jo b vacancies over time—
at least this is the indication from the Rochester data.
There remains the possibility that the occupational com position of the unfilled
job orders nevertheless may be representative of the com position of job vacancies.
Wherever this might prove to be the case, it would not be necessary to ask the
employer to report occupational details on his job vacancy schedule. The relative
com position of the tw o series in terms of broad occupational groups for m id-M ay
is shown in Table A -5 . The broader the occupational groupings, the greater is the
possibility that the two distributions will be similar. Despite this, there is con
siderable difference in the tw o distributions. The average difference amounts to
5.8 percentage points. The most substantial differences occur in the professional
group, which was significantly understated by the job orders, and in the clerical
sales group, which was appreciably over-represented.
The Rochester results on this point differ significantly from the findings of the
Departm ent of Labor in 12 labor areas. On a combined basis, the occupational
com position by broad categories was similar for the ES unfilled job openings and
the job vacancy totals.5 The average difference amounted only to about 2.2
percentage points in both survey rounds. The distributions for individual labor
areas are not shown. Undoubtedly, the similarity comes from the possibility of
offsetting differences in combining the 12 to 14 areas into a single series. This
evidence suggests that the unfilled job orders of the Em ploym ent Service may
serve as an adequate proxy for the occupational com position of job vacancies at a
regional or national level. There is a high probability, however, that it is not a
satisfactory proxy for specific labor areas.
T
able
A -5 .— Unfilled job orders with the Employment Service and job vacancies,
mid-May 1965, by broad occupation groups
Number
Percent of total
Occupation groups
Unfilled
job orders
Professional _
___ _____ ____
__ _______
Clerical-sales
__ __ _ _ _ _ _____
Service
________ ___________________
Skilled
_____________________________________
Semiskilled _ . __________________________ Unskilled__ _________ _ __ __ - __ ______
All occupations
__
_ __
__ - -
Job
vacancies
Unfilled
job orders
Job
vacancies
603
1,613
455
1,058
561
370
2,102
1,826
823
1,731
1,261
1,033
12.9
34.6
9.8
22.7
12.0
7.9
24.0
20.8
9.4
19.7
14.4
11.8
4,660
8,776
100.0
100.0
Source: Special tabulations of Rochester Employment Service for unfilled job orders and table E-10 of
the final report on job vacancies.
Finally, the conclusion seems inescapable that if job vacancy data
are worth having, they must be obtained by sample survey of em
ployers.
4 Op. cit., Findings p. 51. This statistic presumably is an average for 14 areas. The percentages for in
dividual areas are not given. Nor is it stated whether this average is based on the first survey round, the
second, or both.
s Op. cit. Findings, pp. 51 & 53.
JOB VACANCY STATISTICS
163
Thank you.
Chairman P r o x m i r e . Y ou raise a very fascinating question to me
by this last part of your presentation. I think it is a terrific indict
ment of the present methods of determining the demand for labor,
and so serious that I would like your comment on the statement
made yesterday by Mr. Ross. He said:
In the present economic situation, the question of labor shortages has becom e
sufficiently critical, especially in relation to skilled manpower, that the President,
as previously indicated, has asked the Departm ent of Labor to w atch the situation
closely and to prepare regular reports. I have already pointed out that much of
our evidence on labor shortages is indirect and circumstantial. We could do a
much better job if we had direct evidence on labor shortages through measures of
job vacancies classified by occupation, industry and area.
Now, Mr. Ross indicated that he had to rely on what was available
now, which is very spotty and erratic. And from what you have
just told us, it might be somewhat inaccurate.
Am I wrong?
The National Industrial Conference Board materials, I suspect, are
also spotty and incomplete so far as they depend upon classified
advertising as a source.
Mr. C r e a m e r . The conference board index has no occupational
breakdown at all.
Chairman P r o x m i r e . I know that. It relies on classified ad
vertising?
Mr. C r e a m e r . It just counts the number.
Chairman P r o x m i r e . Y o u say that goes in reverse directions,
though, to Rochester?
Mr. C r e a m e r . In job vacancies in one-third of these 12 cities; yes.
But I do not think that Dr. Ross was relying on that index to indicate
which were the critical skill shortages.
Chairman P r o x m i r e . What was he relying on?
Mr. C r e a m e r . I have not discussed this with him, and I have not
seen any of the documents. But— again coming back to the Rochester
area—I know that the Employment Service in each area has a labor
market analyst who has this responsibility of keeping in touch with
at least the large employers, and speaking with them about what sort
of changes they expect in their-----Chairman P r o x m i r e . Rochester has this, but is this common in
the 146 labor market areas in the country?
Mr. C r e a m e r . I think so—certainly the larger ones; it must be.
Chairman P r o x m i r e . And then they systematically report to the
Bureau of Labor Statistics or the Secretary of Labor?
Mr. C r e a m e r . I guess to the Bureau of Employment Security,
which in turn—in fact this is how they derive their estimates—or this
person, let's say, is responsible for the estimates of local area un
employed.
Chairman P r o x m i r e . Do you have any opportunity to check that
estimate to determine the extent to which that is accurate?
M r . C r e a m e r . The unemployment estimate?
Chairman P r o x m i r e . No, the report of the personnel people.
Mr. C r e a m e r . No, we had no opportunity and my guess would be
that such a person on that particular matter cannot be quantitative,
but he could only indicate the area in which the employers are com
plaining about shortages.
164
JOB VACANCY
STATISTICS
Chairman P r o x m i r e . Let me ask the fundamental question that
Mr. Goldfinger has raised earlier.
The statistics users, as Mr. Goldfinger has properly said, raise a
very serious question here. And I would like to read to you, Mr.
Creamer and Mr. Myers, what they said and get your comments.
The Congress has been reluctant—
These are statistical users—
to support any proposal for initiation of the program on a large scale until results
of the exploratory work are made public and proper study and evaluation of the
work can be made. There is no practical evidence yet available to the public as
to the feasibility of collecting reliable information or whether or not such inform a
tion would be of material assistance in the formulation of public policy.
Now, these are competent people. They have appeared often before
this committee. And they do not have any particular interest that
would be adverse here, in fact they should have a very strong interest
in getting this information. What is your comment?
Mr. C r e a m e r . I quite agree particularly with that part of it which
says the evidence should be made public. And I am delighted-----Chairman P r o x m i r e . H o w long would it take to get an evaluation
from a number of experts who have had a chance to analyze it and
discuss it, write about it, comment on it, so that Congress could act
on the basis of knowledge, adequate knowledge of the situation?
Mr. C r e a m e r . Say 3 to 6 months, I should think.
Representative C u r t i s . Mr. Chairman, may I say that if they will
read the hearings of this committee, which I recommend, I think they
will change this position.
Mr. C r e a m e r . Much of the evidence that was made available to
us has now been made available to this committee. And I think that
is among the many great functions-----Chairman P r o x m i r e . Mr. Goldfinger, how long a period do you
think it would take?
Mr. G o l d f i n g e r . I would agree with Mr. Creamer that it would
take at least 3 to 6 months.
But I might add that while I was not here yesterday, and I have
not read the testimony of yesterday, everything that I have heard
today onJy strengthens my feeling that we need much better answers
to various questions concerning concept and definition, and that there
is great need for getting detail by occupation, skill, industry, educa
tion, wage rates, and working conditions, on a meaningful job-market
basis.
Chairman P r o x m i r e . So you think they must study it 3 to 6
months and then turn it down?
Mr. G o l d f i n g e r . I would think that they may; that is true. But
what I am saying is that as far as I am concerned, on the basis of
listening to Mr. Creamer today in a well-reasoned and excellent
presentation of the NICB findings and point of view, it only
strengthens my conviction that there are grave difficulties in this
area, particularly in moving rapidly to a national statistical index.
I am more wary of job vacancy statistics—more negative— than I
was when the hearings opened this morning.
Chairman P r o x m i r e . Mr. Myers.
Mr. M y e r s . I have nothing particular to add to Mr. Creamer’s
statement.
JOB VACANCY
STATISTICS
165
Chairman P r o x m i r e . At any rate this would mean, considering
the way that Congress moves, that there would not be any action
on this until next year? This is May, and if we took 3 to 6 months
to make an analysis that would be adequate for Congress to be sure
they knew what they were doing, and had the best and most com
petent and expert advice, this session would be through?
Mr. C r e a m e r . I would hate to——
Representative C u r t i s . I would sure hate to think that.
Chairman P r o x m i r e . That is just an opinion, I am not saying w e
should.
Representative C u r t i s . I would sure hate to think it.
Mr. Ross said something that you emphasized. I was trying to
remember where it was. He commented about what he felt was the
great need for this. He said, if I remember right, something about
the most important—and you emphasized it:
The lack of jo b vacancy inform ation constitutes the most significant gap in our
knowledge of labor market conditions. Statistics on job vacancies would give us
a measure of unsatisfied demand for labor which together with data on em ploy
ment would provide a more efficient measure of the demand for labor, something
we have never had before.
And, Mr. Goldfinger, he pointed to a caveat that you have rightly
pointed to. He said he thought that some people would misuse these
statistics. But he said he felt that that is what you are always up
against. I myself can understand how people might misuse it. But
the Ways and Means Committee has just been through some lengthy
hearings extending over many months on unemployment insurance.
And the lack of data that we have on the profile of the unemployed,
let alone the other side of the coin, of job creation, and what jobs
there are, indicates to me the great need to move forward on this.
I had something to do with the development of the Manpower Training
Act and feel that without job vacancy data we might just as well not
have it. You do not train people unless you have a job in sight.
That discipline is in the Manpower Training Act. It was a require
ment of getting up to date the Dictionary of Occupational Titles.
I was distressed, and expressed myself forcefully in those hearings
in January, that the 1949 Dictionary of Occupational Titles had still
not been updated. I am happy to say that as of January of this year
the new edition is out. And I am not trying to evaluate it. But at
the same time the Manpower Training Act had the requirement that
they develop job vacancy statistics. And here we are in 1966, and
the Department of Labor has still not complied with the law. I
would have preferred, if they did not want to comply with the law,
to have them report and say why they were not doing it. Instead,
I got an answer from the Secretary of Labor, which was that the reason
we have not done it is because Congress—blaming us—had not given
the administration the $2.5 million. My answer to that was that if
the administration had really felt that this was such an essential area
to develop in the field of statistics, then certainly Congress, composed
of leaders of its own party, would not have denied this. This is
particularly true when the Republicans like myself, at any rate, were
willing to get behind this, and were arguing for it.
So I really would be distressed, Mr. Chairman, if I felt that the
administration was going to back away again, and not ask your
leaders in the Congress to move forward with this program, which I
think, rightly or wrongly, is according to law and is required.
166
JOB VACANCY
STATISTICS
Now, Congress can cut off the appropriations.
Chairman P r o x m i r e . They asked for two and a half million dollars.
They may not ask for it again.
Representative C u r t i s . But there are ways of asking for it. You
can ask for it in a footnote and then forget about it. But President
Johnson has demonstrated that he can ask for things in a certain way
and get them. I would like that kind of “ ask.”
Mr. C r e a m e r . If I may add to my last comment, I fully agree that
these data, the results of the BES survey, should be made public.
We were criticized in some quarters for releasing our findings on
each of the surveys. But we took the position that this was the only
way you get the validity of the findings tested. You have to make
them available to others to work with.
So it is on that basis primarily which led me to say that certainly
these findings of the BES survey should be made available.
Chairman P r o x m i r e . By publishing the findings do you mean the
kind of data that were released to this committee and reported to this
committee yesterday, is that sufficient, will that do it, or is more
required? What additional information is required to be released?
Mr. C r e a m e r . Probably what would be helpful would be more
data by a particular area, rather than combined data. But also they
have been unwilling to present any material on what is called the job
vacancy rate, the reason being that they say this requires more
analysis, more investigation to see how reliable it is. However, they
do make available the number of jobs. And they do indicate the
employment by area. And so they are in the position of saying the
numerator and the denominator passed muster, they are going to make
it public, but they won’t divide one by the other and get a job vacancy
rate.
Chairman P r o x m i r e . Y o u see, we do have a very real, practical
problem here. Mr. Knowles tells me it will take a month or so for the
report to be printed and published and made available by areas, which
you say is necessary. That takes us to mid-June, then a minimum of
3 months, a maximum of 6 months, which would be between midSeptember and mid-December, before objective outsider experts can
analyze it and be prepared to report on it. That kills it for this year.
Mr. C r e a m e r . I did want to add also that I do not think that you
need wait for the experts’ recommendation here, at least I myself have
enough faith in the competence of the technicians of the Department of
Labor to go along with their recommendations in this matter.
Chairman P r o x m i r e . Let me ask you one other question on this.
Mr. Goldfinger raised a question of definition and concepts. Let me
just ask you: Do you feel that there is any problem in determining what
is a job vacancy? Do you think that they are sufficiently advanced so
that they can determine this without much question, and without a
great controversy?
Mr. C r e a m e r . Well, we found that this was understood by the
employers in Rochester, that it is operational, and this need not— this
does not invalidate the program.
Chairman P r o x m i r e . What concepts do we have to clarify, then,
in your judgment?
Mr. C r e a m e r . It is not perhaps so much concept as how much
detail you think you need and can get on occupations, for example, on
wage rates and working conditions. And this, I think, is a matter of
JOB VACANCY
STATISTICS
167
experimentation. We have found, for example, that you can get
usable information on what we call the minimum acceptable educa
tional requirements and related job experience. I would like to see,
or hope that BES would try and get that in some of their areas.
But this all comes with further work in the field. And as I men
tioned earlier, you have to have larger appropriations to extend the
scope of the surveys—perhaps 75 to 80. If that is not manageable,
I am sure they would settle for a smaller number. But I think it is
important that there be an enlargement, and that there be scope for
further experimentation, which the Bureau of Labor Statistics has in
mind, of using the labor turnover approach.
Chairman P r o x m i r e . Y ou are all set to go ahead with this. You
would, on the basis of your study in Rochester, on the basis of your
competence in the field, your knowledge in the field, you would ap
prove of Congress going ahead this session promptly. You would
also urge, I understand, that the Labor Department do publish and
make available just as soon as they possibly can the results of these
area studies?
Mr. C r e a m e r . Yes, that would reflect our view, yes.
Chairman P r o x m i r e . Mr. Goldfinger?
Mr. G o l d f i n g e r . Well, I still have these numerous questions
about concept and definition. For example, the NICB study in
Rochester used some definitions that I questioned, such as future
openings. I think that there are real difficulties about that.
We discussed about half an hour ago or an hour ago whether a
1-day opening is a job vacancy, or whether a real job vacancy is
something which has been unfilled for at least a week and in which
the employer has followed through with an attempt to fill it.
I think that there are all of those kinds of questions which have to
be answered which would require further experimentation and further
work.
I think that there are loads of questions concerning concept,
particularly in connection with a part of Mr. Creamer’s paper that
dealt with the aggregate area, of aggregating the data into some kind
of aggregate averages and dropping the meaningful detail.
For all of those reasons, I see the need for a good deal of work in
terms of concept and definition. And as I said, I do not remember
seeing in the Bureau of Labor Statistics’ publication, “ The Monthly
Labor Review,” any articles dealing with the concept and definition.
It may well be that they have done some preliminary work on this
in BLS. But I am not aware of published materials along these
lines. Why do they seek to rush ahead, without such careful and
tested work?
I see the need for going ahead slowly in terms of experimentation
and working out a lot of these problems on an area basis.
Moreover if there is any significant usefulness for this kind of data
in the near future, it seems to me that it is in the sense of an opera
tional tool, with available detail on occupation, industry, wage rate
information, working conditions, and so on, for the purpose of im
proving the effectiveness of the public employment service, the train
ing programs, the guidance and counselling programs, and vocational
education. Why not move ahead on an area basis, developing an
operational tool? I see no need to rush ahead into some kind of
large-scale national program when we have all of these questions before
168
JOB VACANCY
STATISTICS
us that are unresolved. In addition, I question the meaningfulness of
a national index, a statistical average, with no detail on occupation,
skill, and so forth.
Chairman P r o x m i r e . Mr. Knowles makes this point. He says
we have not settled similar questions in regard to GNP and unemploy
ment even after three decades of successfully running these programs.
Mr. G o l d f i n g e r . That is not quite true, because we have resolved
some kind of consensus on basic concepts. I don’t think-----Chairman P r o x m i r e . But they are controversial, we disagree on
them.
Mr. G o l d f i n g e r . There are disagreements. But the disagree
ments are more toward the periphery than toward the core. And I
think that there are questions concerning the core in this area. Fur
thermore, changes in the unemployment statistics are made carefully
and slowly, after considerable work, testing and evaluation.
I indicated very serious doubts particularly about the part of Mr.
Creamer’s paper which is entitled “Analytical and Policy Uses of
Job Vacancy Statistics.” In fact, as I mentioned before, after reading
this section of Mr. Creamer’s paper, I have graver doubts about the
wisdom of moving in this direction than I stated in my prepared paper.
Chairman P r o x m i r e . Mr. Curtis?
Representative C u r t i s . I would like to say that many of these
definitions that you are talking about have received considerable
attention in the unemployment insurance program. I was just jotting
down a few of these. This is empirical evidence based on the statutorial interpretation over 30 years. What is “ attachment to labor
market to qualify for benefits” ? What is “ available for work” ?
What are seasonal workers? The definition of an employer of one
or more. What does that mean? Does he have a job of one man
with him 1 day, 1 week, or one man for 20 days? The problem we
get into—this has not been statutorily defined, but it has certainly
been in the concept of the economists—is when we relate benefits to
50 percent of the person’s wage. What is the wage base? Does it
mean 1 hour a week or 26 weeks, or what does it mean?
In other words, I think there is a lot of work that has been done in
this area that can easily be translated into some proper definitions
when we move over to the positive side of the ledger, when we are
talking about jobs to be filled as opposed to men who are unemployed.
Mr. G o l d f i n g e r . What I am suggesting, sir, is that this has not
been done in the area of job vacancies as yet. We have just begun.
And although I have not read the BES and BLS presentations of
yesterday, I would suspect that a lot of the data, and a lot of their
conclusions are broadly tentative rather than firm conclusions.
Representative C u r t i s . When you deal with cardinal numbers
and you put a minus sign in front of them, it is pretty easy to just
put a plus sign, then you have got your definition. We have been
defining in the negative area of unemployed people. But in making
our definitions there we have created definitions that with very little
adjustment are available when we turn over to the plus sign. That
is my argument.
Mr. G o l d f i n g e r . For example, a question occurred to me during
Mr. Creamer’s presentation. I do not have the answer to it. But
much of what Mr. Creamer presented is for the purpose of policy, in
terms of interpretation of data for broad economic policy purposes.
Here’s what occurred to me.
JOB VACANCY STATISTICS
169
In February, May, and August of 1965, the NICB study apparently
found something that could be interpreted as full employment in the
Rochester area, on the basis of a comparison of the aggregate job
vacancy and unemployment figures. Yet something like 6 months
before that first survey, there was a serious riot in the city of Rochester,
involving unemployed, underemployed, low-wage Negro workers in
the city of Rochester, which 6 months later is apparently reported to
be a city of full employment.
Now, the question in my mind is, perhaps if the NICB study, on
the basis of their concepts and definitions, had been done 1 year
before, they may well have found on the basis of these statistics— and
they are simply numbers and estimates that are based upon the NICB
definitions and NICB concepts— they may well have found something
very close to full employment. Such statistical averages would then
have implied to policy makers— and some people would have shouted
loudly—let’s go easy on Government expenditure programs, on fiscal
policy and monetary policy, whereas there were some very serious
problems in that community that required Government job-creating
programs for unskilled workers and training programs. I don’t see
that such problems and needs are pointed up here, in the absence of
detailed occupational information, skill requirements, educational
levels-----Representative C u r t i s . Mr. Goldfinger, I think you are making the
case for the need of this. I will tell you right now, where did this
crazy—I call it “ crazy” —concept that 4-percent unemployment is
full employment come from?
Mr. G o l d f i n g e r . Well, I never accepted that.
Representative C u r t i s . Well, I did not either. But this has come
from the misuse of the unemployment statistics and the desire of some
people to put forward the policy that you can solve structural and
frictional unemployment through increasing aggregate demand or
by heating up the economy. As I have observed, you can get roast
pork by burning the barn down, but it is a poor way to get it.
Now, your theory— and I could not agree with it more—is that in
Rochester the fact that you have your unemployed among Negroes
and young Negroes clearly revealed that this was structural, par
ticularly in the light of the fact that you had jobs going begging. Why
is it that you can have unemployed in a society at the same time
that you have a high rate of job vacancies? And I suggest that the
data clearly reveal that we have structural problems. The Negro
unemployment is structural.
Mr. G o l d f i n g e r . I agree that there are structural problems in
the economy, and I felt that there were structural problems back in
1961 and 1962. But in the summer of 1964, in Rochester, there was an
urgent need for jobs for unskilled workers, as well as for training pro
grams.
It seems to me, sir, that when you rely on aggregates, and you get
away from detail, you are losing the usefulness and importance of
whatever usefulness and importance there may be in job vacancy
data. To the extent that you have detail on occupations, on skill
requirements, on wage rates and working conditions, on educational
requirements, and all of that kind of detailed information on a job
market basis, then I think that we are moving into a useful area of
information that can and should be used by the local community in
170
JOB VACANCY STATISTICS
terms of their education programs, vocational training, MDTA, and
so forth. But the emphasis on aggregates for determining fiscal and
monetary policy, I think, is moving into a dangerous area. And this
is what bothers me. The NICB study shows nothing about the
unemployment and underemployment of Negroes in Rochester.
Yet it could be and probably would have been used to kill off the
urgently needed Government programs to create jobs for unskilled
workers.
Representative C u r t i s . I have certainly hit the time that you are
hitting right now. But to me it is ironic that people who are using
the limited aggregate statistics we have now to make policy appar
ently don’t want other statistics to be developed on an aggregate
basis, even though this would make the aggregates we have a great
deal more meaningful. They now use the argument that you do need
the composition of that which makes the aggregates.
I happen to think you need both.
Mr. G o l d f i n g e r . I think we need both too.
Representative C u r t i s . I certainly do agree that it is a very dan
gerous thing to start using aggregate statistics without regard to their
limitations. And I have argued this for many years in regard to
unemployment statistics. I know Senator Douglas has pointed out,
and with good reason, that when we use unemployment statistics we
tend to lose sight of the underemployed, those are in and out. And
so I argue too, that without any concept of what jobs are going
begging, the aggregate unemployment statistics are leading us astray.
And I would argue that for a number of years we have been following
very erroneous fiscal policies, because the aggregate economists have
been able to use this unemployment statistic to support their case for
liberal, as they call it, monetary and fiscal policy. For many years
it has seemed clear to me that we have a problem of very serious
structural and frictional unemployment.
That is why I helped create a group that got behind the develop
ment of the Manpower Training Act. The labor leaders were in there
trying to say, oh, just extend unemployment benefits. But I said,
let’s look and see why these people are long-term unemployed. Are
the coal mines really going to open up, or isn’t this a problem that
you don’t really solve by giving unemployment benefits. Shouldn’t
you try to figure out what is the problem? And in my judgment
we should have retraining and an early warning system to indicate
where jobs and skills were going to become obsolete.
So we get into the structural problems. But we have not been
making adequate progress. And this is the one thing that might alert
us to these things, and help us to understand not only what is really
structural and frictional unemployment, but that which can be cyclical.
There is no question but what it is easier to solve structural and fric
tional unemployment in periods of economic upturn than in a period
of economic downturn. That still does not get around the problem
that much unemployment is structural. We are still forcing people
into the hospitals to fill those skills, people who are not really skilled,
and we are doing the best we can to train them on the job. And this
is creating some serious safety and health problems, I would observe.
Yes, you can do on-the-job training, force it on in that way. But
in my judgment it is not the best way.
JOB VACANCY STATISTICS
171
Let me put two things on the record here. I have a fifth category
of job creation, new enterprise, and particularly some new business
coming into the community. And it is interesting to see how it
affects the jobs already in existence. When a new enterprise comes
in it frequently does not take new employees. It takes people already
having jobs. But in the process other jobs open up elsewhere. So I
think that is another category.
And then I want to put this on the record. I do not know that
there is anything that can be done about it. But I am intrigued
with this business of on-the-job training as it relates to the problem
of job vacancies and unemployment and specifically to this thing of
using the want ads. Many employers say they don't even advertise
for this skill because it just isn't available. They say they know
that there is only one way they are going to get it, and that is to
actually train the man on the job.
I placed in the Congressional Record a couple of months ago a
study by an economist who estimated that we are spending between
$14 and $16 billion a year in the private sector on on-the-job training.
This is a hard statistic to get. But I suspect if we could determine
how much on-the-job training a company spends, particularly if we
had benchmarks, it would be very revealing. The statistic on quits
seems to indicate when the labor market is tight and when it eases
up. I think maybe these kinds of additional data would indicate a
similar thing.
Thank you, Mr. Chairman.
Chairman P r o x m i r e . Thank you, Mr. Curtis.
Mr. Goldfinger?
Mr. G o l d f i n g e r . I wanted to complete my thought on the issue
that I raised concerning the Rochester Negro problem.
It may well be that additional spending programs of a pinpointed
nature, aimed at creating jobs for unskilled workers with low levels of
education may be necessary, in addition to training and vocational
education programs.
Chairman P r o x m i r e . I see Mr. Creamer is nodding, and I think
all of us would agree with that. I would think that the more informa
tion you get, the more data you get, the more detail you get, the more
knowledge you get, the quicker we go ahead with this program, the
more quickly we can persuade the House to appropriate this, and it
seems to me the better we will be equipped.
Mr. G o l d f i n g e r . But there is a danger in the aggregate. The
aggregate could tell us—or it would be misinterpreted to tell us—
to halt any expansion in fiscal policy or to cut back on Government
expenditure programs.
Representative C u r t i s . But let me tell you what happened. The
Manpower Training Act went in with the discipline that you cannot
train a man unless there is a job in sight. And then the Secretary
of Labor reported back and said that we are finding out that some of
these people do not even have the ability to read and write, so they
cannot be trained for a job in being unless they are taught to write
and read. And the Congress in its wisdom— and it was wisdom, I
felt—modified the Manpower Training Act so that we can take a
person and teach him to read and write as long as it is tied in, still
tied in with training for a job.
172
JOB VACANCY
STATISTICS
So this is an indication of how, when you get moving in this area of
getting information on the structural and frictional problems, we
come up with more correct answers.
Mr. G o l d f i n g e r . I agree with you. I think the Congress and the
administration have done well in this area of moving ahead and getting
modifications of MDTA.
Representative C u r t i s . Thank you. I think it is a shame we have
been so slow. And I still say we have not done the job.
Mr. G o l d f i n g e r . The tragedy is that we got started so late. But
we have been moving ahead since 1961 and 1962.
Representative C u r t i s . I would say on the contrary that that is
when we slowed down. And it was because these aggregate eco
nomists thought they could solve these things by monetary and fiscal
policy and avoid the harsh consequences of dealing with these micro
economics.
Chairman P r o x m i r e . I want to thank you very much. This
has been a most enjoyable, interesting, and educational morning.
And I think that you gentlemen have all been very, very helpful.
I do not know if we have come to a definitive conclusion on the
basis of this, but it has been mighty well discussed on all sides.
This will conclude the hearings of the subcommittee. The sub
committee will stand adjourned. And the record will be open for a
week or so for correcting your remarks——
Representative C u r t i s . And extending your remarks.
Let me say—I know the chairman agrees—that if you have thoughts
or things that would help, please do extend your remarks, or if you
think you can say it better, please do so.
(Whereupon, at 12:40 p.m., the committee adjourned, subject to
the call of the Chair.)
(The material requested by the chairman during the proceedings
and later supplied by the Department follows—see p. 110:)
Hon. W i l l i a m P r o x m i r e ,
Chairman, Subcommittee on Economic Statistics, Joint Economic Committee, Congress
of the United States, Washington, D.C.
D e a r S e n a t o r P r o x m i r e : We are pleased to supply you with the information
on job vacancies which you requested in your letter of M ay 20 as follows:
Enclosure 1: The schedules used in the 1964-1965 program.
Enclosure 2: The schedules used in the 1966 surveys.
Enclosure 3: A table classifying the number of vacancies in the April 1965
surveys by occupation and industry for each area.
Enclosure 4: A table on the number and percent of vacancies offering wages
below the entry rates of pay in the area for the April 1965 surveys.
Because of the large number of individual occupations involved in the analysis
of this wage data, we have com bined the occupations into broad occupational
groups. As a result of these groupings, however, we could not indicate the spe
cific entry rates offered for the vacancies since these naturally varied over a wide
range within each occupational group. Because of the survey design, the analysis
of substandard vacancies was based only on actual vacancies rather than on an
inflation of the vacancies to a universe total. If it were to be subsequently demon
strated that smaller firms pay lower wages than larger firms, it would have the
effect of understating the percentage of substandard wage vacancies.
We have not enclosed the job vacancy rates which you requested in item 3 of
your letter, since information on em ploym ent by occupation is not available. As
for the rates by industry, the sample used in the fiscal year 1965 surveys was not
designed to provide data in this detail for all industries. While for some indus
tries such as manufacturing the data would be accurate, in others such as con
JOB VACANCY STATISTICS
173
struction and service it is subject to too large an error because of the smaller
samples involved. We have im proved the sampling design for our fiscal year
1966 surveys so that this type of detail can be provided. Therefore, we would
prefer not to publish job vacancy rates by industry until we have had an oppor
tunity at least to compare the second year’s data with the results of the first.
Operational uses of the data were discussed in the testim ony given by Mr.
Cassell (pages 13-15) and M r. Chavrid (pages 21-24) at the hearings held on
M ay 17. The continued high rate of response in the April 1965 round of surveys
attests to the fact that the surveys were well received. N ot reported in the
testim ony was a letter by the Birmingham Chamber of Commerce supporting the
Birmingham survey attached to all survey schedules sent to local employers.
In this letter, the Chamber stated that it
. . recognizes the costliness and wastefulness of this inability to match jobs
with trained personnel who are out of work. As a first step in a solution to this
problem your Chamber is supporting the Alabama State Em ploym ent Service in
making a job vacancy survey in the Birmingham area. W e believe it is in our
best interest to lend a helping hand in solving, on the local level, this econom ic
problem .”
Thank you for the opportunity given us to furnish your Subcomm ittee with
the additional job vacancy information and for your continued interest in the
program.
Sincerely yours,
R o b e r t C. G o o d w i n , Administrator.
JOB VACANCY
174
E nclosure
STATISTICS
1— S t a n d a r d S c h e d u l e F o l l o w e d
U
sed in
6 V
a r ia n t
Schedules
Budget Bureau No. 4*+-R 1255
Approval Expires 6-30-65
JV-i
U.S. DEPARTMENT OF LABOR
WASHINGTOND.C. 20210
6j 7
Please enter data re
quested and return in
accompanying envelope
by Decemoer 22.
by
1 9 6 4 -6 5 S u r v e y s
REPORT ON J03 VACANCIES
(State Agency Name
and
Address)
(Change Name and Mailing Address If Incorrect )
/
Information reported on this
form is strictly confidential,
and will not be revealed to any
unauthorized person nor pub
lished in such a manner that
data relating to individual
____ /
companies can be identified.
/ ___
BEFORE ENTERING DATA PLEASE READ EXPLANATIONS ATTACHED
/
A.
JOB VACANCIES, BY OCCUNATION. List below"‘sy. jo b ..t it le ,.a ll job vacancies
(as defined in attached explanations) in your establishment as of the close
of business on December 11. I f this is r.ot possible, report vacancies for
the nearest possible day. A job vacancy is a job opening that was unfilled
and immediately available to f u l l - , part-time, or temporary workers which
your firm was actively seeking from outside your company, i f no vacancy,
Job Vacancies
Occupational
Code
(Leave Blank)
(1)
Job Title
(2)
Total
Number
(3)
Number existinr»;
One
One work
week or
month or
longer
longer
( 5)
Comments
(6)
(Continue on baci< of fo rm i f more space is rleeded.)
Bl
C.
REFERENCE DATE: Job vacancies reported above were as of
(Date).
NUMBER OF EMPLOYEES: V/hat was the total number of employees who worked
during or received pay for any part of the pay period which includes the
12th of December? ___________________________
Signature
____________ _______________________ Title
(Firm representative responsible for this report)
JOB VACANCY
A.
STATISTICS
175
JOB VACANCIES, BY OCCUPATION (Continued)
Job Vacancies
Occupational
Code
(Leave Blank)
(1)
Number existing:
One
One work
week or month or
Total
longer
Number longer
Job Title *
.................... (3)
(5)
w
Comments
(6)
REMARKS:
Note to State agency: This continuation of employer collection fora, should be
printed back to back with page 1 of same form. Do same with "Explanations" at
tached to employer collection form. Reason: The employer should receive as
few sheets of paper as possible.
JOB
VACA N CY
STATISTICS
EXPLANATIONS FOR REPORT ON JOB VACANCIES
(Please read before entering data on report form)
A.
JOB VACANCIES are defined as current, unfilled job openings in your
establishment which are immediately available for occupancy by
workers from outside your firm and for which your firm is actively
seeking such workers. Included are part-time, full-tim e, permanent,
temporary, seasonal and short-term job openings.
“Actively seeking” is defined as current efforts to f i l l the job with
a worker from outside your firm through: (1) soliciting assistance
of public or private employment agencies, school or college placement
o ffic e s, labor unions, employee groups, business or professional orga
nizations, business associates, friends and employees in locating
suitable candidates; (2) using "help wanted’1 advertising (newspaper,
magazine, radio, television, direct mail, posted notice, e t c .) ;
(3) conducting recruitment programs or campaigns; (4) interviewing
and selecting "g a te ," "walk-in'' or "mail" applicants or workers
searched out of applicant f i l e s ; and (5) opening or reopening the
acceptance of applications from prospective candidates.
Do not include as vacancies 1) jobs held for employees who will be
recalled; 2) jobs to be fille d by transfer, promotion, or demotion;
3) jobs held for workers on paid or unpaid leave; 4) jobs fille d by
overtime work which are not intended to be fille d by new workers;
5) job openings for which new workers were already hired and scheduled
to start work at a later date; and 6) those jobs unoccupied because
of labor-raanagement disputes.
B.
REFERENCE DATE:,, Enter date for which vacancies are reported as Item
B on page 1.
C.
NUMBER OF EMPLOYEES: Enter the total number of employees on a ll
payrolls of your establishment who worked full-tim e or part-time or
received pay for any part of the pay period which includes the 12th
of December. Include persons on vacation and sick leave who re
ceived pay directly from your firm for the pay period reported, but
exclude persons on leave without company pay the entire period
and pensioners and members of the Armed Forces carried on the rolls
but not working during the pay period.
Column 1:
Leave blank.
Column 2:
List job t it le s for which job vacancies exist in your estab
lishment. Where possible, add modifier denoting material,
product, process or subject matter to make establishment
t i t le more specific, e .g ., "sssenibler, aircraft, wing parts"
or "stenographer, le g a l." Use a single entry to report
job vacancies with identical job t it le s where more than one
vacancy exists. If there are several classes or grades for
specific job t i t l e s , each class or grade should be listed
separately.
For office use only.
JOB
VACANCY
STATISTICS
Column 3:
For each job t it le listed in Column 2, report the
total number of job vacancies.
Column
For each job t it le enter the number of job vacancies
included in Column 3 which have existed one work
week or longer. I f none, enter "None." I f informa
tion cannot be provided, enter "X ." Do not leave
blank.
Column 5î
Of the numbers shown in Column k , enter the number
of job vacancies which have existed for 1 month or
longer. I f none, enter "None." I f information
cannot be provided, enter "X ." Do not leave blank.
Column 6 ;
Enter comments explaining why job vacancies are hard
to f i l l . I f additional space is needed, use "Remarks"
section on back of form.
IF YOU HAVE ANY QUESTIONS, PLEASE TELEPHONE:
____________________ _____ ____________________ at
(Name)
___________
(Phone No.)
When form is completed, please return in the accompanying s e lfaddressed stamped envelope by December 22.
Thank you for your cooperation.
Note to State agency: This 2nd page of "Explanations" should be
printed back to back with fir s t page of "Explanations."
JOB
178
JV-2
U.S. DEPARTMENT.OF LABOR
Washington, D.C. 20210
Please enter data re
quested and return in
accompanying envelope
by December 22.
VACA N CY
STATISTICS
2
6
3
5
REPORT ON JOB VACANCIES
Budget Bureau No. M4-R 1255
Approval Expires 6-30-65
(State Agency Name
and
Address)
( Change Name and Mailing Address I f Incorrect )
Information reported on this form
is strictly confidential, and will
not be revealed to any unauthori
zed person nor published in such a
manner that data relating to indi
vidual companies can be identified.
BEFORE ENTERING DATA PLEASE READ EXPLANATIONS ATTACHED
/
/
A.
JOB VACANCIES, BY OCCUPATION. List below, by job t i t l e , a ll job vacancies
(as defined in attached explanations) in your establishment as of the close
of business on December 11. I f this is not possible, report vacancies for
the nearest possible day. A job vacancy is a job opening that was unfilled
and immediately available to f u l l - , part-time, or temporary workers which
your firm was actively seeking from outside your company. I f no vacancy,
write "None" in Column 2, complete items 3 and C, and return the report.
Rate
of pay offered
Job Vacancies
i Indicate
Number existing:
whether
per hour,
Occupational
One
One work
week, or
month or
Code
week or
Total
month
Amount
(Leave Blank)
Number
longer
longer
Job Title
(7)
(6)
(5)
(2)
(3)
(1)
W
J
I
(Continue on back oi? form ii : more space! is needed .)
C.
NUMBER OF EMPLOYEES: What was the total number of employees who worked dur
ing or received pay for any part of the pay period which includes the 12th
of December?____
__
Signature
Title
(Finn representative responsible for this report.)
JOB
A.
VACANCY
STATISTICS
179
JOB VACANCIES, 3Y OCCUPATION (Continued)
Job Vacancies
Occupational
Code
(Leave 31ank)
(1)
Job Title
(2)
Total
Number
(3)
Number existing:
On3 work
One
month or
week or
longer
longer
(5)
'M
R.
of pa;y offered
Indicate
whether
per hour,
week, or
month
Amount
(7)
(e)
REMARKS:
Note to State agency: This continuation of employer collection form should be
printed back to back wit.* page X of same form. Do same with "Explanations” at
tached to employer collection form. Reason: The employer should receive as
few sheets of paper as possible.
JOB
180
V ACA N CY
STATISTICS
EXPLANATIONS FOR REPORT ON JOB VACANCIES
(Please read before entering data on report form)
A.
JOB VACANCIES are defined as current, unfilled job openings in your estab
lishment which are immediately available for occupancy by workers from out
side your firm and for which your firm i s a c tiv e ly seeking such workers.
Included are full-ti»ae, part-time, permanent, temporary, seasonal and short
term job openings.
"Actively seeking" is defined as current efforts to f i l l the job with a
worker froraoutside your firm through: (1) soliciting assistance of public
or private employment agencies, school or college placement o ffic e s, labor
unions, employee groups, business or professional organizations, business
associates, friends and employees in locating suitable candidates; (2)
using "help wanted" advertising (newspaper, magazine, radio, television,
direct mail, posted notice, e t c .); (3) conducting recruitment programs or
campaigns; (4) interviewing and selecting "g a te ," "walk-in" or "mail"
applicants or workers searched out of applicant f i l e s ; and (5) opening or.
reopening the acceptance of applications from prospective candidates.
Do not include as vacancies 1) jobs held for employees who w ill be recalled;
2) jobs to be fille d by transfer, promotion, or demotion; 3) jobs held for
workers on paid or unpaid leave; 4) jobs fille d by overtime work which are
not intended to be fille d by new workers; 5) job openings for which new
workers were already hired and scheduled to start work at a later date;
and 6) those jobs unoccupied because of labor-management disputes.
B.
REFERENCE DATE: Enter date for which vacancies are reported as Item B on
page 1. Enter date even though you report no vacancies.
C.
NUMBER OF EMPLOYEES: Enter the total number of employees on a ll payrolls
of your establishment who worked full-tim e or part-time or received pay
for any part of the pay period which includes the 12 of December. In
clude persons on vacation and sick leave who received pay directly from
your firm for the pay period reported, but exclude persons on leave with
out company pay the entire period, and pensioners and members of the
Armed Forces carried on the ro lls but not working during the pay period.
Enter this figure even though you report no vacancies.
Column 1:
Leave blank.
Column 2:
List job t it le s for which job vacancies exist in your establishment.
Wnere possible, add modifier denoting material, product, process or
subject matter to make establishment t it le more specific, e .g .,
"assembler, aircraft, wing parts" or "stenographer, le g a l." Use a
single entry to report job vacancies with identical job t it le s where
more than one vacancy exists. If there are several classes or
grades for specific job t i t le s , each class or grade should be listed
separately.
For office use only.
JOB
VACANCY
STATISTICS
Column 3:
For each job t it le listed in Column 2, report the total number
of job vacancies.
Column 4:
For each job t it le enter the number of job vacancies included
in Column 3 which have existed one work week or longer. If
none, enter "None.1' I f information cannot be provided, enter
"X .11 Do not leave blank.
Column 5:
Of the numbers shown in Column 4, enter the number of job
vacancies which have existed for 1 month or longer. If none,
enter ’’None.” If information cannot be provided, enter "X .”
Do not leave blank.
Column 6:
For each job vacancy listed in Column 2, enter on the line
opposite that vacancy, the rate of pay offered for the job to
which the job vacancy relates. The entry of a single rate
of pay is preferred; however, where a range of pay rates is
offered depending on the varying educational training and ex
perience qualifications of prospective applicants, the entry
of the offered pay rate range (that i s , the low and high pay
rates offered) w ill be acceptable. Wherever possible, please
enter hourly pay rates. I f the wage offered for the opening
is on a piece work or commission incentive basis, please enter
the estimated average full-tim e weekly earnings the new worker
is expected to receive.
Column 7:
For each pay rate (or pay rate range) entered in Column 6,
enter the basis on which the offered pay rate is quoted in
Column 6 (for example, indicate whether the pay rate given
is per hour, per week, per month).
IF YOU HAVE ANY QUESTIONS, PLEASE TELEPHONE:
____________________ _________________________
( Name)
at _____________ ____________
( Phone No.)
When form is completed, please return in the accompanying self-addressed
stamped envelope by December 22. Please do so even though you report
no vacancy.
Thank you for your cooperation.
Note to State agency: This 2nd page of " Explanations" should be printed
back to back with fir st page of "Explanations."
JOB
182
DL_____
U.S. ;DE_
WASHINGTON, D.C.
V AC A N C Y
STATISTICS
Budget Bureau
« o u
iNo.
* w ♦ 4^-R 1255
. .1 Approval Expires &-30-65
». . . .
20210
Please or;cer data re
quested and return in
accompanying envelope
by December 1.
1 2
LJ
REPORT ON JOB VACANCIES
(State Agency Name
and
Address)
Information reported on this
form is strictly confidential,
and w ill not be revealed to
any unauthorized person nor
published in such a manner
that data relating to indivi
dual companies can be identi
fied.
I
BEFORE ENTERING DATA PLEASE READ EXPLANATIONS ATTACHED
(Change Name and Mailing Address If Incorrect)
/
1
A.
JOB VACANCIES, BY OCCUPATION. List below, by job t i t l e , a ll job vacancies
(as defined in attached explanations) in your establishment as of the close
of business on November 20. If this is not possible, report vacancies for
the nearest possible day. A job vacancy is a job opening that was unfilled
and immediately available to f u l l - , part-time, or temporary workers which
your firm was actively seeking from outside your company. If no vacancy,
write "None*1 in Column 2.
Job Vacancies
Number
existing
Occupational
one
Code
Total
month or
Comments
Number
longer
(Leave Blank
Job Title
(3)
(5)
U)
( 1)
B.
C.
(Continue on back of form i f more space is needed)
REFERENCE DATE: Job vacancies reported above were as of~^_____
( Date).
NUMBER OF EMPLOYEES: V/hat was the total number of employees who worked
during or receivedpay for any part of the pay period which includes the
12th of November?____________________________
Signature ________________________________________ Title _____________________
(Firm representative responsible” for this report.)
JOB VACANCY
A.
STATISTICS
183
JOB VACANCIES, BY OCCUPATION (Continued)
Job Vacancies
Occupational
Code
(Leave Blank)
(1)
Job Title
(2)
Number
existing
one
Total month or
Number longer
(H)
Comments
(5)
“~1
REMARKS:
Note to State agency: This continuation of employer collection form should bs
printed back to back with page 1 of same form. Do same with "Explanations” at
tached to employer collection form. Reason: The employer should receive as
few sheets of paper as possible.
JOB
184
V ACA N CY
STATISTICS
EXPLANATIONS FOR REPORT ON JOB VACANCIES
(Please read before entering data on report form)
A.
JOB VACANCIES £3- defined as current, unfilled job
establishment which are immediately available for
from outside your firm and for vhich your firm is
such workers. Included are full-'time, part-time,
seasonal and short-term job openings.
openings in your
occupancy by workers
actively seeking
permanent, temporary,
"Actively seeking" is defined as current efforts to f i l l the job with a
worker from outside your firm through: (l) soliciting assistance of
public or private employment agencies, school or college placement
o ffices, labor unions, employee groups, business or professional organi
zations, business associates, friends and employees in locating suitable
candidates; (2) using "help wanted" advertising (newspaper, magazine,
radio, television, direct mail, posted notice, e t c .); ( 3 ) conducting
recruitment programs or compaigns; (b) interviewing and selecting ''gate,11
"walk-in" or "mail" applicants or workers searched out of applicant file s
and ( 5 ) opening or reopening the acceptance of applications from prospec
tive candidates.
Do not include as vacancies l) jobs held for employees vho w ill "be
recalled; 2) jobs to be fille d by transfer, promotion, or demotion;
3) jobs held for workers on paid or unx>aid leave; 4) jobs fille d by
overtime work vhich are not intended to be fille d by new workers;
5) job openings for which new workers were already hired and scheduled
to start work at a later date; and 6) those jobs unoccupied because
of labor-management disputes.
B.
REFERENCE DATE:
on page 1.
Enter date for which vacancies are reported as Item B
C.
HUMBER OF EMPLOYEES: Enter the total number of employees on a ll payrolls
of your establishment who worked full-tim e or part-time or received pay
for any part of the pay period which includes the 12th of November. In
clude persons on vacation and sick leave who received pay directly from
your firm for the pay period reported, but exclude persons on leave
without company pay the entire period, and pensioners and members of
the Armed Forces carried on the rolls but not working during the pay
period.
Column 1 :
Leave blank.
Column 2 :
List job t it le s for which job vacancies exist in your establish
ment. Where possible, add modifier denoting material, product,
process or subject matter to make establishment t i t le more spe
c ific , e .g ., "assembler, aircraft, wing parts" or "stenographer,
le g a l." Use a single entry to report job vacancies with identi
cal job t it le s where more than one vacancy exists. I f there
are several classes or grades for specific job t it le s , each
class or grade should be listed separately.
For office use only.
JOB
V ACANCY
185
STATISTICS
Column 3 :' For each job t it le listed in Column 2, report the total number
of job vacancies.
Column k :
Of the numbers shown in Column 3 > enter the number of job
vacancies which have existed for 1 month or longer. I f none,
enter "None." I f information cannot be provided, enter "X ."
Do not leave blank.
Column 5:
Enter comments explaining why job vacancies are hard to f i l l .
I f additional space is needed, use "Remarks" section on back
of form.
IF YOU HAVE ANY QUESTIONS, PLEASE TELEPHONE:
at
(Name)
(Phone No.)
When form i$ completed, please return in the accompanying self-addressed
stamped envelope by December 1.
Thank you for your cooperation.
Note to State agency: This 2nd page of "Explanations" should be printed
back to back with fir s t page of "Explanations."
JOB VACANCY
186
STATISTICS
Budget Bureau No. kk-R 1255
Approval Expires 6-30-65
JV-4
U.S. DEPARTMENT OF LABOR
WASHINGTON, D.C.
20210
Please enter data re
quested and return in
accompanying envelope
by December 22.
51
(State Agency Name
and
Address)
!T
REPORT OK JOB VACANCIER
(Change
Name and Mailing Address I f Incorrect)
(Cha]
Information reported on this form
is strictly confidential, and
w ill not be revealed to any un
authorized person nor published
in such a manner that data re
lating to individual companies
can be identified.
BEFORE ENTERING DATA, PLEASE READ EXPLANATIONS ATTACKED
A~. JOB VACANCIESTbY 0CCÜPM?ÏÔN" List below, by job t i t l e , a ll job vacancies (as
defined in attached explanations) in your establishment as of the close of
business on December 11. I f this is not possible, report vacancies for the
nearest possible day. A job vacancy is a job opening that was unfilled and
ilnmediately available to f u l l - , part-time, or temporary workers which your
firm was actively seeking from outside your company. I f no vacancy, write
"None" in Column 2, complete items B and C, and return the report.
Occupational
Code
(Leave Blank)
.
B
C
Job Vacancies
Job Title
(2)
.
Total Number
____ .13).........
Comments
cn
(Continue on back of form i f more space is needed
REFERENCE DATE: Job vacancies reported above were as of
(Date)~
NUMBER OF EMPLOYEES: What was the total number of employees wh; worked during
or received pay for any part of the pay period which includes the 12th of
December?
Signature
___________________________________________Title_____
(Firm representative responsible for this report. )
JOB
A.
VACANCY
STATISTICS
187
JOB VACANCIES, BY OCCUPATION (Continued)
Occupational
Code
(Leave Blank)
m
'
Job Vacancies
Job Title
(2)
Total Number
(3) ...
Comments
w
...
REMARKS:
Note to State agency: This continuation of employer collection foim should be
printed back to back with page 1' of same form. Do same with ''Explanations’* at
tached to employer collection fo m . Reason: The employer should receive as few
sheets of paper as possible.
...
188
JOB
VACA N CY
STATISTICS
EXPLANATIONS FOR REPORT ON JOB VACANCIES
(Please read before entering data, on report form)
A-
JOB VACANCIES are defined as currant, -unfilled job openings in your
establishment which are immediately available for occupancy by workers
fjjom outside your firm and for which your firm is actively seeking
such workers. Included are full-tim e, permanent, temporary, seasonal
and short-term job openings.
"Actively seeking11 is defined as current efforts to f i l l the job with
a worker from outside your firm through: ( l ) soliciting assistance of
public or private employment agencies, school or college placement
o ffices, labor unions, employee groups, business or professional organi
zations, business associates, friends and employees in locating suitable
candidates; (2) using "help wanted" advertising (newspaper, magazine,
radio, television, direct mail, posted notice, e t c .) ; (3) conducting
recruitment programs or campaigns; (h) interviewing and selecting "gate”,
"walk-in" or "mail" applicants or workers searched out of applicant
f i l e s ; and (5) opening or reopening the acceptance of applications from
prospective candidates.
Do not include as vacancies (l) jobs held for employees who w ill be re
called; (2) jobs to be fille d by transfer, promotion, or demotion; (3)
jobs held for workers on paid or unpaid leave; (4) jobs fille d by over
time work which are not intended to be fille d by new workers; (5) job
openings for which new workers were already hired and scheduled to start
work at a later date; and (6) those jobs unoccupied because of labornanageaient disputes.
B.
REFERENCE DATE: Enter date for which vacancies are reported as Item B
on page 1 . Enter date even though you report no vacancies.
C.
NUMBER OF EMPLOYEES: Enter the total number of employees on .a ll payrolls
of your establishment who worked full-tim e or part-time or received
pay for any part of the pay period which includes the 12th of December.
Include persons on vacation and sick leave for which they received pay
directly from your firm for the pay period reported, but exclude per
sons on leave without company pay the entire period, and pensioners
and members of the Armed Forces carried on the ro lls but not working
during the pay period. Enter this figure even though you report no
vacancies.
Column 1:
Column 2 ;
Column 3:
Leave blank. For office use only.
List job t it le s for which job vacancies exist in your establish
ment. Where possible, add modifier denoting material, product,
process or subject matter to make establishment t i t l e more
specific, e .g ., "assembler, aircraft, wing parts" or "stenographer,
le g a l." Use a single entry to report job vacancies with identi
cal job t it le s where more than one vacancy exists. I f there are
several classes or grades for specific job t it le s , each class
or grade should be listed separately.
For each job t i t l e listed in Column 2, report the total number
of job vacancies.
JOB
Column 4 :
VACA N CY
STATISTICS
189
Enter comments explaining vhy job vacancies are hard to f i l l .
I f additional space is needed, use "Remarks" section on back
of form,
IF YOU HAVE ANY QUESTIONS, PLEASE TELEPHONE:
____________ ^
__________ ______________
(Name)
at
______________________________
(Phone No.)
When form is completed, please return In the accompanying self-addressed
stamped envelope by December 22. Please do so even though you report no
vacancies.
Thank you for your cooperation.
Note to State agency: This 2nd page of ’’Explanations" should be printed
back to back with fir s t page of "Explanations. ”
JOB VACANCY STATISTICS
190
JV-5
U.S. DEPARTMENT OF LABOR
Washington,D.C. 20210
Please enter data re
quested ai'id return in
accompanying envelope
by December 22.
Budget Bureau Ko. 44-R 1255
Approval Expires 6-30-65
REPORT ON JOB VACANCIE
(State Agency Name
and
Address)
Information reported on this
form is strictly confidential,
and w ill not be revealed to any
unauthorized person nor pub
lished in such a manner that
data relating to individual
I __
___ /
companies can be identified.
BEFORE ENTERING DATA PLEASE READ EXPLANATIONS ATTACHED______________________________
^Change Name and Mailing Address I f Incorrect)
/
/~
A»
JOB VACANCIES, BY OCCUPATION. List below, by job t i t l e , a ll job vacancies
(as defined in attached explanations) in your establishment as of the close
of business on December 11. I f this is not possible, report vacancies for
the nearest possible day. A job vacancy is a job opening that was unfilled
and immediately available to f u l l - , part-time, or temporary workers which
your firm was actively seeking from outside your company. If no vacancy,
write "None" in Column 2 , complete items 3 and C below, and return your
report._________________
____
Job Vac.-:inciüS
D.O.T.
Code
( Leave
Blank)
Job Title
(1)
3.
C.
(2)
Total
Parttime
Number
only
(3T
T hT
ALL Vacancies Existing
Gne
One work
month or
week or
longer
longer
(5)
7sr
(Continue on back of form i f more space is needed.)
REFERENCE DATE: Job Vacancies reported above were as of _____________ (Date).
NUMBEit OF EMPLOYEES: What was the total number of employees who worked during
or received pay for any part of the pay period which includes the 12th of
December?
Signature
_____________ ________________________
Title
(Firm representative responsible for this report.!
JOB VACANCY
A.
JOB VACANCIES, BY OCCUPATION
D.O.T.
Code
( Leave
Blank)
(1)
Job Title
(2 )
STATISTICS
191
(Continued)
Job Vacancies
Total
Parttime
only
Number
(4)
(3)
ALL Vacancies Existing
One
One work
month or
week or
longer
longer
(6)
(5)
REMARKS:
Note to State agency: This continuation of employer collection form should be
printed back to back with page 1 of same form. Do same with "Explanations”
attached to employer collection form. Reason: The employer should receive as
few sheets of paper as possible.
JOB
192
VACA N CY
STATISTICS
EXPLANATIONS FOR REPORT ON JOB VACANCIES
(Please read before entering data on report form)
A.
JOB VACANCIES are defined as current, unfilled job openings in your
establishment vhich are immediately available for occupancy by workers
from outside your firm and for which your firm is actively seeking
workers. Included are fu ll-tim e, part-time, permanent, temporary,
seasonal and short-term job openings.
"Actively seeking" is defined as current efforts to f i l l the job with
a worker from outside your firm through: (l) soliciting assistance
of public or private employment agencies, school or college placement
o ffices, labor unions, employee groups, business or professional orga
nizations, business associates, friends and employees in locating
suitable candidates; ( 2 ) using "help wanted" advertising (newspaper,
magazine, radio, television, direct mail, posted notice, e t c .); ( 3 )
conducting recruitment programs or campaigns; (U) interviewing and
selecting "gate"; "walk-in" or "mail" applicants or workers searched
out of applicant f i l e s ; and ( 5 ) opening or reopening the acceptance
of applications from prospective candidates.
Do not include as vacancies l ) jobs held for employees who w ill be
recalled; 2 ) jobs to be fille d by transfer, promotion, or demotion:
3 ) jobs held for workers on paid or unpaid leave; ii) jobs fille d by
overtime work which are not intended to be fille d by new workers;
5 ) job openings for which now workers were already hired and scheduled
to start work at a later date; and 6) those jobs unoccupied because
of labor-management disputes.
B.
REFERENCE DATE: Enter date for which vacancies are reported as Item
Enter date even though you report no vacancies.
B on page 1.
C.
NUMBER OF EMPLOYEES: Enter the total number of employees on a ll
payrolls of your establishment who worked full-tim e or part-time
or received pay for any part of the pay period which includes
the 12th of December. Include persons on vacation and sick leave
for which they received pay directly from your firm for the pay
period reported, but exclude persons on leave without company pay
the entire period, and pensioners and members of the Armed Forces
carried on the rolls but not working during the pay period. Enter
this figure even though you report no vacancies.
Column 1:
Leave blank.
Column 2 :
List job t it le s for which job vacancies exist in your estab
lishment. Where possible, add modifier denoting material,
product, process or subject matter to make establishment
t it le more specific, e .g ., "assembler, aircraft, wing parts"
or "stenographer, le g a l". Use a single entry to report job
vacancies with identical job tit le s where more than one
vacancy exists. I f there are several classes or grades for
specific job t it le s , each class or grade should be listed
sspiirtVv/OlLy.
For office use only.
JOB
VACANCY
193
STATISTICS
Column. 3 • For each job t it le listed in Column 2, report the total number
of Job vacancies.
Column 4 :
For each job t it le enter the number of vacancies included in
Column 3 which are for positions in which the weekly hours
to be worked are less than the customary straight-time work
week for the occupation. I f the job vacancy is open to either
full-tim e or part-time workers, please report i t as a f u l l
time vacancy in Column 3. I f none, enter "None." I f the infor
mation cannot be provided, enter ”X ". Do not leave blank.
Column 5:
For each job t i t l e enter the number of job vacancies included
in Column 3 which have existed one work week or longer. I f
none, enter "None." I f information cannot be provided, enter
"X ". Do not leave blank.
Column 6:
Of the numbers shown in Column 5 , enter the number of job
vacancies which have existed for 1 month or longer. I f
none, enter "None." I f information cannot be provided, enter
"X ". Do not leave blank.
IF YOU HAVE ANY QUESTIONS, PLEASE TELEPHONE:
(Name)
(Phone No.)
When form is completed, please return in the accompanying self-addressed
stamped envelope by December 22. Please do so even i f you report no
vacancies.
Thank you for your cooperation.
Note to State agency: This 2nd page of "Explanations" should be printed
back to back with fir s t page of "Explanations.11
JOB VACANCY
194
JV-6
U.S.
STATISTICS
Budget Bureau No. kk-R 1255
department of labor
20210
!
1
Please enter data requested and return in
accompanying envelope
by December 22.
2 3
M
5
6
7 8
REPORT ON JOB VACANCIES
(State Agency Name
and
Address)
(Change Name and Mailing Address I f Incorrect)
Information reported on this foim
is strictly confidential, and w ill
not be revealed to any unauthorized
person nor published in such a
manner that data relating to indi
vidual companies can be identified.
BEFORE ENTERING DATA, PLEASE READ EXPLANATIONS ATTACHED
A. JOB VACANCIES, BY OCCUPATION. List below, by job t i t l e , a ll job vacancies (as
defined in attached explanations) in your establishment as of the close of
business on December 11. I f this is not possible, report vacancies for the
nearest possible day. A job vacancy is a job opening that was unfilled and
immediately available to f u l l - , part-time, or temporary workers which your
firm vas actively seeking from outside your company. I f no vacancy, write
"None'* in column 2, complete items B and C below and return your report.
D.O.T.
Code
(Leave
Blank)
Job Title
ir e i
—
Job Vacancies
ALL Vacancies Exi: .ti^ f
Total
Temporary only
3 dayj over 3 days One week One month
Number
or less to 4 months or longer or longer
HL
HI
(«>)
.. .
(Continue on back of form i f more space is needed).
REFERENCE DATE: Job vacancies reported above were as of
_(Date).
KUIvIBER OF EMPLOYEES: What was the total number of employees who worked during
or received pay for any part of the pay period which includes the 12th of
December?________________
Signature____________ ____________________________ Title ____________________
(Firm representative responsible for this report.)
JOB
,A.
V ACANCY
195
STATISTICS
JOB VACANCIES, BY OCCUPATION (Continued)
D.O.T.
Code
(Leave
Blank)
Job Title
J H
Job Vacancies
Total
Temporary only
over 3 days
i days
or less to h months
Nimber
E L
(W
ISL
Vacancies Existi:
One week One month
or longer or longed
ÆL
BEMAHS:
Note to State agency: This continuation of employer collection fozm should be
printed back to back with page 1 of same form. Do same with "Explanations'*
attached to employer collection form. Reason: The employer should receive as
few sheets of paper as possible.
H e
JOB
196
V ACA N CY
STATISTICS
EXPLANATIONS FOR REPORT ON JOB VACANCIES
A.
JOB VACANCIES are defined as current, unfilled 'ob openings in your estab
lishment which are immediately available for occupancy by workers from out
side your firm and for which your firm is actively seeking such workers.
Included are full-tim e, part-time, permanent, temporary, seasonal and short
term job openings.
"Actively seeking" is defined as current efforts to f i l l the job with a
worker from outside your firm through (l) soliciting assistance of public
or private employment agencies, school or college placement o ffic e s, labor
unions, employee groups, business or professional organizations, business
associates, friends and employees in locating suitable candidates; (2)
using "Help wanted" advertising (newspaper, magazine, radio.,- television,
direct mail, posted notice, e t c .); (3) conducting recruitment program or
campaigns; ( k ) interviewing and selecting "ga te ," "walk-in" or "mail" applicants
or workers searched out of applicant file s ; and (5) opening or reopening
the acceptance of applications from prospective candidates.
Do not include as vacancies l ) jobs held for employees who w ill be recalled;
2) jobs to be fille d by transfer, promotion or demotion; 3) jobs held for
workers on paid or unpaid leave; ^) jobs fille d by overtime work which are
not intended to be fille d by new workers; 5) job openings for which new
workers were already hired and scheduled to start work at a later date; and
6) those job occupied because of labor-management disputes.
B.
REFERENCE DATE: Enter data for which vacancies are reported as Item B on
page 1. Enter date even though you report no vacancies.
C.
NUMBER OF EMPLOYEES: Enter the total number of employees on a ll payrolls of
your establishment who worked full-tim e or part-time or received pay for any
part of the pay period which includes the 12th of December. Include persons
on vacation and sick leave for which they received pay directly from your
firm for the pay period reported, but exclude persons on leave without company
pay the entire period, and pensioners and members of the Armed Forces carried
on the rolls but not working during the pay period. Enter this figure even
though you report no vacancies.
Column 1:
Leave blank.
Column 2:
List job t itle s for which job vacancies exist in your establish
ment. ' Where possible, add modifier denoting material, product,
process or subject matter to make establishment t it le more spe
c ific , e .g ., "assembler, aircraft, wing parts" or "stenographer,
le g a l." Use a single entry to report job vacancies with identical
job tit le s where more than one vacancy exists. I f there are several
classes or grades for specific job t it le s , each class or grade
should be listed separately.
Column 3:
For each job t it le listed in Column 2, report the total number of
job vacancies.
For office use only.
JOB
VACANCY
197
STATISTICS
Column
Enter the number of job vacancies for short-time jobs, i . e . , those
expected to last for 3 days or less.
Column 5:
For each job t i t l e , enter the numbor of job vacancies included in
Column 3 which represent on.nin^L, i or temporary workers. A tempo
rary job is a job expected to l.ii>c irtore than 3 days but not longer
than 4 months.
Column 6:
For each job t i t l e , enter the number of job vacancies included in
Column 3 which have existed ono work wock or longer. If none,
enter "None.” If information cannot bo provided, enter "X ." Do
not leave blank.
Column 7:
Of the numbers shown in Column 6, enter the number of job vacancies
which have existed for 1 month or longer. I f none, enter "None.”
I f information cannot be provided, enter "X ." Do not leave blank.
IF YOU HAVE ANY QUESTIONS, PLEASE TELEPHONE:
(Name)
(Phone No. )
When form is completed, please return in the accompanying self-addressed stamped
envelope by December 22. Please do so even though you report no vacancy.
Thank you for your cooperation.
Note to State agency: This 2nd payv of "Explanations" should be printed back
to back with fir st page of "Explanations."
JOB
198
V ACA N CY
STATISTICS
Budget Bureau No.
Approval Expires
JV-7
U.S. DEPARTMENT OF LABOR
WASHINGTON, D.C.
20210
Please enter data re
quested and return in
accompanying envelope
by December 22.
(State Agency Name
and
Address)
REPORT ON JOB VACANCIES
(Change Name and Mailing Address I f Incorrect)
1255
Information reported on this form
is strictly confidential, and w ill
not be revealed to any unauthorized
person nor published in such a
manner that data relating to indi
vidual companies can be identified.
BEFORE ENTERING DATA, PLEASE READ EXPLANATIONS ATTACHED
Part-time
Full-time
Over h months
Less than
1 week
11 week but
1less than 1
! mrin-hVi
1 month or
longer
JOB VACANCIES t BY OCCUPATION. List below each individual Job vacancy (as de
fined in attached explanations) existing in your establishment as of the close
of business on December 11. I f i t is not possible to report for that date, do
so for the nearest possible date. Use a separate line for each individual job'
vacancy and enter on that line a ll information about the Job vacancy requested
in Columns 2 through 6b. A job vacancy is a job opening that was unfilled end
immediately available to fu ll-tim e, part-time, or temporary workers which your
firm was actively seeking from outside your company. I f you have no job va
cancies, write "None" in Column 2, complete items B and C below, and return
your report.
How long has How long is Is this a JWhat rate
job vacancy job exacte i full-time! of pay is
or part- pifared for
to last?
existed?
(Check one) (Check one) time job?| this job?
(Check one!)
A
!
. {Indicate
o
aa
;
Whether
gay rate
o «ÖCO•
Occ.
j is per
to
coo
Code
! hour,
a? u. k *
»ö K a>-dj week,
(Leave
è o
Blank)
Jteci^nt month
Job Title
-p
(2 )
(5b)
. teail (6b)
(1)
(?c )I1**: (kb] {h e)
'i
i
!
ì
!
ì
(Continue on back <D f
fC
rm i f mo!re £»pace is needLed).
i
.
i
1 .
»
.. i — .....
!
i
1
1
KQKE3R OF EMPLOYEES: What was the total number of employees who worked during
or received pay for any part of the pay period which includes the 12th of
December?_______________
Signature_________ ________________________
Title
_____
_________
(Firm representative responsible for this report).
JOB
A.
VACANCY
199
STATISTICS
JOB VACANCIES, BY OCCUPATION (Continued)
How long hc.L How long ic
job vacancy job expecte:i
to last?
existed?
(Check one) (Check one)
1
CT
C
c J
-p
a
& • o
o >d ^ n
liis
s.
oo n
|‘.:S s' V. U)
f-i
) tiJ in u Ja)
'ö ui (U
in rj
a>
! a> O
6
; -i.—
1Cr-i —: rO1—1è -p
LOcc.
Code
(Leave
Blank)
(1 )
S
£ ,s:
Job Title
a»
w>
(j>
l—
l r-i
(2 )
Î-.
o
(3 a )Î3 b )(3 c ),(l« i)
(h-b)ikc) 1
Is this t.
full-time
or parttime Job?
(Check one)
What rate
of pay is
offered for
this Job?
l i p dicate
w.:lether
a)
¿i
0)
;!-J
•p
Î
»
Ö
PH
P ty rate
s per
hour,
week,
Amount month
(6b)
(6a)
i
1
i
i
c
>
1
i
1
1
REMARKS;
Note to State agency: This continuation of employer collection form should be
printed back to back with page 1 of some form. Do came with "Explanations" at
tached to employer collection form. Reason: The employer should receive as few
sheets of paper as possible. Also: Horizontal lines in item A should be repro
duced as shown on schedule.
200
JOB
VACANCY
STATISTICS
EXPLANATIONS FOR REPORT ON JOB VACANCIES
(Please read before entering data on report form)
A.
JOB VACANCIES are defined as current, unfilled Job openings in your
establishment vrtiich are immediately available for occupancy by workers
from outside your firm and for which your firm is actively seeking
such workers. Included are full-tim e, part-time, permanent, temporary,
seasonal and short-term job openings.
"Actively seeking" is defined as current efforts to f i l l the job with
a worker from outside your firm through: ( l ) soliciting assistance
of public or private employment agencies, school or college placement
o ffice s, labor unions, employee groups, business or professional
organizations, business associates, friends and employees in locating
suitable candidates; (2) using "help wanted" advertising (newspaper,
magazine, radio, television, direct mail, posted notice, e t c .); ( 3 )
conducting recruitment programs or campaigns; (k) interviewing and
selecting "gate", "walk-in" or "mail" applicants for workers searched
out of applicant f i l e s ; and ( 5 ) opening or reopening the acceptance
of applications from prospective candidates.
Do not include as vacancies l) jobs held for employees who w ill be
recalled; 2) jobs to be fille d by transfer, promotion, or demotion;
3) jobs held for workers on paid or unpaid leave; 4) jobs fille d by
overtime work which are not intended to be fille d by new workers;
5) job openings for which new workers were already hired and scheduled
to start work at a later date; and 6) those jobs unoccupied because
of labor-management disputes.
B*
REFERENCE DATE: Enter date for which vacancies are reported as Item
B on page 1. Enter date even though you report no vacancies.
C.
NUMBER OF EMPLOYEES: Enter the total number of employees on a ll pay
rolls of your establishment who worked full-tim e or part-time or re
ceived pay for any part of the pay period which includes the 12th of
December. Include persons on vacation and sick leave for which they
received pay directly from your firm for the pay period reported, but
exclude persons on leave without company pay the entire period, and
pensioners and members of the Armed Forces carried on the ro lls but
not working during the pay period. Enter this figure even though
you report no vacancies.
Column 1:
Leave blank.
Column 2 :
Enter the job t i t l e for each individual job vacancy existing
in your establishment. Where possible, add modifier denoting
material, product, process or subject matter to make estab
lishment t i t le more specific, e .g ., "assembler, aircraft,
wing parts" or "stenographer, le g a l". Use a separate line for
each individual job vacancy. I f job vacancies exist for
several classes or grades within a specific job t i t l e , each
class*or grade should be separately identified and listed.
For o ffice use only.
JOB
VACANCY
STATISTICS
201
Columns 3&, b, c:
For each job vacancy listed in Column 2, enter on the
line opposite that vacancy a check mark in the appro
priate one of these three columns to indicate how long
that job vacancy (as—defined above)- has existed.
Columns Ua, b , c:
For cach job vacancy listed in Column 2, enter on the
line opposite that vacancy a check mark in the appro
priate one of these three columns to indicate how long
the job to which that job vacancy relates is expected
to la st.
Columns 5a, b :
For each job vacancy listed in Column 2, enter on the
line opposite that vacancy a check mark in the appro
priate one of these two columns to indicate whether the
job to which that job vacancy relates is a full-tim e
or part-time job. A part-time job is defined as one
where the weekly hours to be worked are less than the
customary ctruight-time work week for the occupation.
I f the job vacancy is open to cither full-tim e or
part-time workers, please report i t as a full-tim e job.
Column 6a:
For each job vacancy listed in Column 2, enter on the
line opposite that vacancy the rate of pay offered for
the job to which the job vacancy relates. The entry of
a single refce of pay is preferred; however, where a range
of pay rates is offered, depending on the varying educa
tional, training and experience qualifications of pro
spective applicants, the entry of the offered pay rate
range (that i s , the low arid high pay rates offered; will
be acceptable. Wherever possible, please enter hourly
pay rates. I f the wage offered for the opening is on
a piece work or commission incentive basis, please
enter the estimated average full-tim e weekly earnings
the new worker is expected to receive.
Column 6b:
For each pay rate (or pay rate range) entered in
Column 6a, enter the basis on which the offered pay
rate is quoted in column 6a (for example, indicate
whether the pay rate given is per hour, per week,
per month).
REMARKS:
On back of form enter comments explaining which of the
job vacancies listed you consider hard to f i l l and
reasons for d ifficu lty. Add ony other remarks that may
be pertinent.
IF YOU HAVE ANY QUESTIONS, PLEASE TELEPHONE:
_______ _____ __________________________ at_________________________
(Name)
(Phone No.)
When form is completed, please return in the accompanying self-addressed
stamped envelope by December 22. Please do so even though you report
no vacancies.
202
JOB
V AC A N C Y
STATISTICS
Thank you for your cooperation.
Mote to State agency: This 2nd page of "Explanations" should be printed
hack to back with fir s t page of "Explanations".
JOB
VACA N CY
STATISTICS
E n c lo su r e 2 — St a n d a r d Sc h e d u l e F o l l o w e d
U
U .S .
sed in
by
203
3 V
a r ia n t
Sc h edules
1966 P r o g r a m
D EPA RTM EN T OF LABOR
BUR E AU OF L A B O R S T A T I S T I C S
BUR E AU OF E MP L O Y M E N T S E C U R I T Y
V A t N I N G T O N , O. C.
20210
Budget Bur e a u
Expires:
8/31/66
Approval
No.
44-6609
Date:
REPORT ON JOB VACANCIES
I N F O R M A T I O N F U R N I S H E D ON T H I S R E P O R T I S S T R I C T L Y C O N F I D E N T I A L ANC « I L L NOT BE
NOR P U B L I S H E D I N SUCN A MANNE R THAT DATA R E L A T I N G TO I N D I V I D U A L C O M P A N I E S CAN
INSTRUCTI ONS
AC C O MP A NY I HG
OH THE BACK
E H V E L O P E BY
OF T H I S FORM B E F O R E
A P R I L 8, 1 9 6 6 TO:
( C HA N6 E
NAME
OR A O D R E S S
E N T E R I N 6 THE DATA
( NA ME AND A O D R E S S
IF
OH T H I S R E P O R T .
OF A GE N C Y )
R E V E A L E D TO ANY
BE I D E N T I F I E D .
AHO
I NCORRECT)
R E T URH
THE
U N A U T H O R I Z E D P E R S ON
P L E A S E R EAD THE
R E P OR T
IH
THE
DO NOT USE
Oa t e
Area
State
(2)
(1 )
SIC
(3)
10
(<)
Base
(5)
Si ze
TC
(6)
NC ORD
-ILL 1 * 1 i l l i ° l
1. TOTAL NUMBER OF EMPLOYEES:
e n t e r t h e t o t a l n u mb e r o f e m p l o y e e s
WORKED OR R E C E I V E D P A Y > O R ANY P ART OF THE PAY P E R I O D WHI CH I N C L U O E S
THE 1 2 t h OF MARCH ( S E E I H S T R U C T I O H S ON BACK OF T H I S F ORM) .
ES
Cards
±11
(12)
#K
M3i
who
2. JOB VACANCIES BY OCCUPATION:
L I S T BELOW, BY J OB T I T L E . A L L JOB V A C A N C I E S ( A S D E F I N E D ON THE BACK OF T H I S
F ORM) I N YOUR E S TA BL I S H M E H T . AHO THE RATE OF PAY O F F E R E D . AS OF THE C L OS E OF B U S I N E S S . A P R I L 1 . 1 9 6 6 ( OR THE
NEAREST P O S S I B L E DAY).
A JOB V ACANCY I S A J O B O P E N I N G THAT I S U N F I L L E D AND I M M E D I A T E L Y A V A I L A B L E TO F U L L - .
P A R T - T I M E . OR T E MP OR AR Y WOR KE RS W H I C H YOUR F I R M I S A C T I V E L Y S E E K I H G FROM O U T S I O E YOUR COMPANY.
NUMBER
VACANT
ONE MONTH
OR L ONGER
O C C U P A T I ONAL
CODE
( L E A V E BL ANK)
P E R HOUR,
DAY. WEEK,
MONTH. ETC
(LEAVE
BL ANK )
(4)
( CONTI NUE
3.
PU i St
C HE CK
THIS
BOX
ON
If
( NA ME OF S T A T E A G E N C Y )
SI GNATURE
(Person
responsible
BACK
OF
FORM
IF
YOU WOULD L I K E
IN F I L L I N G
flTr
this
MORE
SPACE
IS
NEEDED)
THE A S S I S T A N C E
OF THE
AN Y OF YOUR J O B V A C A N C I E S
r eport ! )
TITLE
D AT E FOR WHI C H J OB
V A C A N C I E S WERE R E P O R T E D
2 ()4
J0B
VACANCY STATISTICS
2. JOB VACANCIES BY OCCUPATION: (Continued)____________________________
NUMBER
VACANT
ONE MONTH
OR LONGER
OCCUPATIONAL
CODE
(LEAVE BLANK)
PER HOUR,
DAY, WEEK,
MONTH, ETC.
INSTRUCTIONS FOR REPORT ON JOB VACANCIES
1.
Number o f E m ployees: Enter in the box to the right of item 1 the total number of em ployees on all payrolls of your e sta b lish
ment who worked fu ll-time or part-time or received pay for any part of the pay period which includes the 12th of March. Include
persons on vacation and sick leave who received pay directly from your firm for the pay period reported, but exclude persons on
leave without company pay the entire period, and pensioners and members of the Armed Forces carried on the rolls but not
working during the pay period. P le a se enter this figure even though you report no v a ca n cies.
2.
Job V a can cies, by Occupation: Enter the number of current, unfilled job openings in your establishm ent as of the clo se of
b u sin e ss, April 1, 1966 which are immediately a vailab le for occupancy by workers from outside your firm and for which your
firm is actively seeking such workers. Included are full-tim e, part-time, permanent, temporary, se a so n a l, and short-term job
openings.
"A c t i v e ly seeking** is defined as current efforts to fill the job with a worker from outside your firm through: (1) so licitin g a s
sistance of public or private employment a g e n c ie s, school or college placement o ffic e s , labor unions, employee groups, busi
n e ss or profession al organizations, b u siness a s s o c ia te s , friends and em ployees in locating suitable can didates; (2) using
" h e lp wanted” advertising (newspaper, m agazine, radio, television , direct m ail, posted n otice, e tc .); (3) conducting recruit
ment programs or cam paigns; (4) interviewing and selectin g " g a t e , " "w a lk -in ” or "m a il** applicants or workers searched out
of applicant file s ; and (5) opening or reopening the acceptance of applications from prospective candidates.
Do not include as vacan cies (1) jobs held for em ployees who will be recalled; (2) jobs to be filled by transfer, promotion, or
demotion; (3) jobs held for workers on paid or unpaid leave; (4) jobs filled by overtime work which are not intended to be filled
by new workers; (5) job openings for which new workers were already hired and scheduled to start work at a later date; and
(6) those jobs unoccupied because of labor-management disputes.
Column 1: L eave blank.
For o ffice use only.
Column 2:
L is t the names of the occupations for which job vacancies exist in your establishm ent. Where p o ssib le , add modi
fier denoting material, product, process or subject matter to make establishm ent title more sp e c ific , e .g ., "a s s e m b le r , air
craft, wing parts’ * or "steno g ra ph er, le g a l.” U se a single entry to report job v acan cies with identical job titles where more
than one vacancy e x is ts . If there are several c la s s e s or grades for sp e cific job title s, each cla ss or grade should be listed
separately.
Column 3:
For each occupation listed in column 2, report the total number of job v a ca n cie s.
Column 4:
For each occupation, enter the number of job vacancies included in column 3 which have been vacant for one month
or longer. If none, enter " 0 ” . If information cannot be provided, enter " X .* * Do not leave blank.
Column 5: For each occupation, enter in column 5 the rate of pay offered for the job to which the vacancy relates
P le a se en
ter a single rate of pay wherever p o s s ib le . However, where a range of pay rates is offered, depending upon the education,
training, and experience of the job applicant, the entry lev els of the low and high pay rates offered will be acceptab le.
P le a s e enter hourly pay rates in column 5 wherever p o ssib le .
If the wage rate offered for the job vacancy is on a piecework or commission incentive b a sis, or if tips comprise part of the
worker’ s earnings, plea se enter the average fu ll-time earnings which a new worker can be expected to receive.
Column 6:
For each pay rate (or pay rate range) entered in column 5, please enter the b a sis on which the offered pay rate is
quoted, that is , whether the pay rate given is per hour, per week, per month, etc.
Column 7:
Leave blank.
For office use only.
IF YOU HAVE ANY QUESTIONS, PLEASE TELEPHONE:
(Na me )
( T e l e p h o n e Number J
When report is com pleted, plea se return it in the accompanying self-addressed stamped envelope by April 8 , 196 6 .
DO SO EVEN THOUGH YOU HAVE NO VACANCIES TO REPORT.
Thank you for your cooperation.
PLEASE
JOB
U .S .
VACANCY
STATISTICS
205
Budgat l a r a a u No.
Approval E i p l r o a :
DEPARTMENT OF LABOR
B U R E A U OP L A B O R S T A T I S T I C S
BUREAU OF EMPLOY MENT S E C U R I T Y
VASNINOTON. D.C.
20210
REPORT ON JOB VACANCIES
I N F O R M A T I O N t U R M I S N C O ON T H I S R E P O R T I I S T R I C T L Y C O N F I D E N T I A L ANO B I L L N O T BE R E V E A L E O TO A N Y U N A U T H O R I Z E D P E R S O N
NO R P U B L I S H E D I N S U C H A N A N N E R T H A T D A T A R E L A T I N S TO I N D I V I D U A L C O M P A N I E S CAN BE I D E N T I F I E D . _ P L E A S E R E A J T H E
I N S T R U C T I O N S ON T H E BACK OF T H I S FORM BE F O R E E N T E R I N G THE OATA ON T H I S R E P O R T . AND RE T URN THE R E P OR T I N THE
AC C OMP ANY I N 6 E N V E L O P E BY A P R I L 8. 1 9 6 8 TO:
( NAME AND A D D R E S S OF A 6 E N C Y )
______________________________________________
■tata
DO
Data
(1)
(2)
Slza
TC
NC
-ILL ILLH i
1. TOTAL NUMBER OF EMPLOYEES:
B O R R E D OR R E C E I V E D PAY
THE 1 2 t h OF MARCH ( S E E
E N T E R THE TOTAL NUMBE R OF E M P L O Y E E S
FOR A N Y P A R T OF THE PAY P E R I O O WHI CH I H C L U O E S
I N S T R U C T I O N S ON THE BACK OF T H I S F ORM) .
NOT USE
T 7 a a -
n e 1
ORO
ES
i l l
”
(4)
(3)
Carda
Wk
JLUL-LUI— L L U
I HO
2. JOB VACANCIES BY OCCUPATION:
L I S T BELOW, BY JOB T I T L E , AL L JOB V A C A H C I E S ( A S D E F I H E O ON THE BACK OF T H I S
FORM) I N YOUR E S T A B L I S H M E N T , ANO THE RATE OF PAY O F F E R E D , AS OF THE C L OS E OF B U S I N E S S , A P R I L 1. 1 9 6 6 ( OR THE
NEAREST P O S S I B L E DAY).
A J 0 8 V ACANCY I S A JOB O P E N I N G THAT I S U N F I L L E D AND I M M E D I A T E L Y A V A I L A B L E TO F U L L - ,
P A R T - T I M E , OR T E MP ORAR Y WORKE RS WHI CH YOUR F I R M I S A C T I V E L Y S E E K I N G FROM O U T S I O E YOUR C OMPANY.
RATE OF PAY
OF F E RE D
NUMBE R OF
JOB V A C A N C I E S
NUMBE R
VACANT
ONE MONTH
OR L ONGER
O C C U P A T I ONAL
CODE
( L E A V E BLANK)
PER HOUR
DAY, WEEK
MONTH,
E TC.
(4)
( CONTI NUE
•VACANCIES
3.
P L E A S E C H EC K T H I S
FOR
BOX
IF
( N A ME OF S T A T E A G E N C Y )
SI GNATURE
(Parson
responsible
ON
BACK
POSITIONS
OF
FORM
'lOU WOULD L I K E
IN
FILLIN G
for
this
IF
EXPECTEO
MORE
TO L A S T
SPACE
IS
(6)
THAN
OF THE
ANY OF YOUR J O B V A C A N C I E S
report)
TI TLE
NUMBE R OF
TE MP ORARY
J OB V A C A N C I E S
I NCLUDED IN
COLUMN ( 3 ) «
( 8)
NE E O E O)
NO L ONGE R
THF A S S I S T A N C E
(LEAVE
BLANK)
4 MONTHS.
□
DATE FOR WHI CH J OB
V A C A N C I E S WERE R E P O R T E O
JOB
V ACA N CY
STATISTICS
2. J O B V A C A N C I E S BY O C C U P A T I O N : (Continued)
NUMBER OF
JOB V AC A NC I E S
OCC UP AT ION AL
CODE
( L E A V E BLANK)
JOB T I T L E
TOTAL
(2 )
(I)
(3 )
RATE OF PAY
O FFERED
NUMBER
VACANT
ONE MONTH
OR LONGER
AMOUNT
P ER HOUR
DAY. WEEK.
MONTH.
ETC.
(4 )
(5 )
(6)
( L E A VE
BLANK)
NUMBER OF
TEMPORARY
JOB V A C A N C I E S
I NC LU DE D IN
COLUMN ( 3 ) *
C7) _
( 8)
INSTRUCTIONS FOR REPORT ON JOB VACANCIES
1. Number o f Em ployees: Enter in the box to the right o f item 1 the total number o f employees on all payrolls o f your establish
ment who worked full-time or part-time or received pay for any part o f the pay period which includes the 12th o f March. Include
persons on vacation and sick leave who received pay directly from your firm for the pay period reported, but exclude persons
on leave without company pay the entire period, and pensioners and members of the Armed Forces carried on the rolls but not
working during the pay period. P le a s e enter this figure even though you report no vacancies.
2. Job V acan cies, by Occupation: Enter the number o f current, unfilled job openings in your establishment as o f the clo s e o f
b u sin ess, April 1, 1966 which are immediately available for occupancy by workers from outside your firm and for which your
firm is actively seeking such workers. Included are full-time, part-time, permanent, temporary, seasonal, and short-term job
openings.
" A c t iv e ly se e k in g '* is defined as current efforts to fill the job with a worker from outside your firm through: (1) solicitin g assista n ce o f public or private employment agencies, school or colleg e placement o ffic e s , labor unions, employee groups, busi
n ess or p rofessional organizations, business a s socia tes, friends and em ployees in locating suitable candidates; (2) using
, " h e lp wanted” advertising (newspaper, magazine, radio, television, direct mail, posted n otice, etc .); (3) conducting recruit
ment programs or cam paigns; (4) interviewing and selectin g " g a t e ,” "walk-in** or "mail** applicants or workers searched out
o f applicant file s; and (5) opening or reopening the acceptance o f applications from p rospective candidates.
Do not include as vacan cies (1) jobs held for em ployees who will be recalled; (2) jobs to be filled by transfer, promotion, or
demotion; (3) job s held for workers on paid or unpaid leave; (4) jobs filled by overtime work which are not intended to be
filled by new workers; (5) job openings for which new workers were already hired and scheduled to start work at a later date;
and (6) those jo b s unoccupied because o f labor-management disputes.
Column 1:
Leave blank.
For o ffic e use only.
Column 2: List the names o f the occupation s for which job vacancies exist in your establishment. Where p o ss ib le , add modi
fier denoting material, product, p rocess or subject matter to make establishment title more s p e c ific, e .g ., "assem bler, air
craft, wing parts” or "stenographer, le g a l.” Use a single entry to report job vacancies with identical job titles where more
than one vacancy exists. If there are several cla s s e s or grades for s p e cific job titles, each cla s s or grade should be listed
separately.
Column 3: For each occupation listed in column 2, report the total number o f job vacancies.
Column 4: For each occu p ation , enter the number o f job vacancies included in column 3 which have been vacant for one month
or longer. If none, enter " 0 . ” If information cannot be provided, enter "X.** Do not leave blfenk.
Column 5: For each occu p ation , enter in column 5 the rate o f pay offered for the job to which the vacancy relates. P lea se en
ter a single rate o f pay wherever p oss ib le. However, where a range o f pay rates is offered, depending upon the education,
training, and experience o f the job applicant, the entry le v e ls o f the low and high pay rates offered will be accep table.
P le a se enter hourly pay rates in column. 5 wherever p ossib le.
If the wage rate offered for the job vacancy is on a p iece work or commission incentive b a sis, or if tips comprise part o f
the worker’ s earnings, p lea se enter the average full-time earnings which a new worker can be expected to receive.
Column 6: For each pay rate (or pay rate range) entered in column 5, p lease enter the basis on which the offered pay rate is
quoted, that is , whether the pay rate given is per hour, per week, per month, etc.
Column 7:
L eave blank.
For o ffic e use only.
Column 8: For each occu p ation , enter in column 8 the number o f vacan cies included'in column 3 which are for temporary p o si
tions. Temporary v a can cies are defined as positions which are expected to last no longer than four months. If none, enter
" 0 . ” If the information cannot be provided, enter " X . ” Do not leave blank.
IF YOU HAVE ANY QUESTIONS, PLEASE TELEPHONE:
( Na me )
( T e l e p h o n e Number)
When report is com pleted, p lea se return it in the accompanying self-addressed stamped envelope by April 8, 1966.
DO SO EVEN THOUGH YOU — ^ E NO VACANCIES TO REPORT. Thank you
out cooperation.
PLEASE
JOB
U .S .
VACANCY
STATISTICS
207
9 EPARTM EN T OF LABOR
Budgat Bureau No.
Approval Ex pl raa :
BUREAU OF LABOR S T A T I S T I C S
BUREAU OF EMPLOYMENT SE CU RI TY
«ASNINBTON. O .C .
20210
REPORT ON JOB VACANCIES
I N F I R M A T I O N F U R N I S H E O ON T N I S R E P O R T I S S T R I C T L Y C O N F I D E N T I A L A N D « I L L N O T BE R E V E A L E D TO A N Y U N A U T H O R I Z E D P E R S O N
N B R P U B L I S N E D I N S U C N A M A N N E R T N A T D A T A R E L A T I N G TO I N O I V I O U A L C O M P A N I E S CAN BE I D E N T I F I E D .
P L E A S E REAO TNE
I N S T R U C T I O N S ON THE B A C K OF T H I S FORM B E F O R E E N T E R I N G THE D A T A ON T H I S R E P OR T . A N D RE T URN THE R E P O R T I N THE
A C C O M P A N Y I N B E N V E L O P E BY A P R I L 8. 1 9 6 8 TO:
( N A M E A N O A D D R E S S OF AGE NC Y )
Stato
DO
Data
(1)
(2)
NOT USE
Araa
ID
(4)
Baaa
(6)
(9)
Slza
TC
NC
ORO
Cards
ES
m ILL 1 1 1 121 H I
1. TOTAL NUMBER OF EMPLOYEES:
V O R K E 0 OR R E C E I V E D P AY
THE 1 2 t h OF MARCH ( S E E
E N T E R THE TOTAL NUMBE R OF E M P L O Y E E S
FOR ANY P AR T OF THE P AY P E R I O D « H I C H I N C L U D E S
I N S T R U C T I O N S ON THE BACK OF T H I S F ORM) .
IIC
(3)
(12)
Vk
1121
«H0
2. JOB VACANCIES BY OCCUPATION:
L I S T BELOW. BY J OB T I T L E . A L L JOB V A C A N C I E S ( A S D E F I N E D ON THE BACK OF T H I S
FORM) I N YOUR E S T A B L I S H M E N T . AND THE RATE OF PAY O F F E R E D . AS OF THE C L OS E OF B U S I N E S S . A P R I L 1 , 1 9 6 6 ( OR THE
NE ARE ST P O S S I B L E DAY).
A J OB V AC ANC Y I S A i O B O P E N I N G THAT I S U N F I L L E D ANO I M M E D I A T E L Y A V A I L A B L E TO F U L L - .
P A R T - T I M E . OR TE MP ORAR Y WOR KE RS WHI CH YOUR F I R M I S A C T I V E L Y S E E K I N G FROM O U T S I D E YOUR COMPANY .
NUMBE R OF
J OB V A C A N C I ES
NUMBE R
VAC ANT
CNE MONTH
OR L ONGER
O C C U P A T I ONAL
CODE
( L E A V E BLANK)
P ER HOUR.
DAY, W EEK
MONTH.
ETC.
( 6)
(4)
(CONTI NUE
• V A C A N C I E S FOR
THE CUS T OMA RY
3.
P L E A S E CHECK T H I S
( N A M E OF S T A T E
BOX
OF
FORM
IF
MORE
SPACE
IS
YOU WOULD L I K E
IN
FILLIN G
THE A S S I S T A N C E
( 8)
NEEDED)
P O S I T I O N S I N WH I C H THE W E E K L Y HOURS TO BE WORKED
S T R A I G H T T I ME WORK WEEK FOR THE O C C U P A T I O N .
IF
AGEN CY)
ON BACK
(LEAVE
BLANK)
ARE
LESS
THAN
OF Th E
ANY OF YOUR J O B V A C A N C I E S
+
SI GNATURE
(Person
responsible
for
this
NUMBE R OF
PART-TIME
JOB V A C A N C I E S
I NCL UDED IN
COLUMN ( 3 ) *
report)
U
DATE FOR « H I C H JOB
V A C A N C I E S WERE R E P O R T E D
208
J0B
vacancy
2. J o b V A C A N C IE S B Y-O C CU PATIO N : (C ontinue)
O CC UP AT IO NAL
CODE
( L E A V E BLANK
s t a t is t ic s
_____ ________________________
NUMBER
VACANT
ONE MONTH
OR LONGER
JOB T I T L E
P ER HOUR.
DAY. WEEK,
MONTH.
ETC.
NUMBER OF
PART-TIME
JOB V A C A N C I E S
I NCL UDED IN
COLUMN ( 3 ) *
( 8)
INSTRUCTIONS FOR REPORT ON JOB VACANCIES
1. Number o f Em ployees: Enter in the box to the right o f item 1 the total number o f em ployees on all payrolls o f your establish
ment who worked full-time or part-time or received pay for any part o f the pay period which includes the 12th o f March. Include
persons on vacation and sick leave who received pay directly from your firm for the pay period reported, but exclude persons on
leave without company pay the entire period, and pensioners and members o f the Armed F orces carried on the rolls but not
working during the pay period. P lea se enter this figure even though you report no vacancies.
2. Job V a ca n cies, by Occupation: Enter the number o f current, unfilled job openings in your establishment as o f the clo s e of
bu sin ess, April 1, 1966 which are immediately available for occupancy by workers from outside your firm and for which your
firm is actively seeking such workers. Included are full-time, part-time, permanent, temporary, seasonal, and short-term job
openings.
" A c tiv e ly se e k in g " is defined as current efforts to fill the job with a worker from outside your firm through: (1) s olicitin g as
sista n ce o f public or private employment agencies, school or co lle g e placement o ffic e s , labor unions, employee groups, busi
n ess or p rofessional organizations, business a s socia tes, friends and em ployees in locating suitable candidates; (2) using
"h e lp w anted" advertising (newspaper, magazine, radio, telev ision , direct mail, posted n otice, e tc.); (3) conducting recruit
ment programs or campaigns; (4) interviewing and selectin g " g a t e ," "w a lk -in " or " m a il " applicants or workers searched out
o f applicant file s; and (5) opening or reopening the acceptance o f applications from p rospective candidates.
Do not include as vacan cies (1) jobs held for em ployees who will be recalled; (2) jobs to be filled by transfer, promotion, or
demotion; (3) jobs held for workers on paid or unpaid leave; (4) jobs filled by overtime work which are not intended to be
filled by new workers; (5) job openings for which new workers were already hired and scheduled to start work at a later date;
and (6) those jobs unoccupied because o f labor-management disputes.
Column 1: Leave blank.
For o ffic e use only.
Column 2:
L ist the names o f the occupations for which job vacan cies e xist in your establishm ent. Where p o ssib le , add modi
fier denoting material, product, pro cess or subject matter to make establishm ent title more sp e c ific , e .g ., "a s s e m b le r , air
craft, wing p a r t s " or "sten o g ra ph er, l e g a l ." U se a single entry to report job vacancies with identical job title s where more
than one vacancy e x is ts . If there are several c la s s e s or grades for sp e c ific job title s, each c la s s or grade should be liste d
separately.
Column 3:
For each occupation listed in column 2, report the total number o f job vacancies.
Column 4: For each occu p ation, enter the number o f job vacan cies included in column 3 which have been vacant for one month
or longer. If none, enter " 0 . " If information cannot be provided, enter " X . " Do not leave blank.
Column 5: For each occu p ation, enter in column 5 the rate o f pay offered for the job to which the vacancy relates. P lea se en
ter a single rate o f pay wherever p oss ib le. However, where a range o f pay rates is offered, depending upon the education,
training, and experience o f the job applicant, the entiy lev els o f the low and high pay rates offered will be acceptable.
P lea se enter hourly pay rates in column 5 wherever p ossib le.
If the wage rate offered for the job vacancy is on a piece work or com mission incentive b a sis, or if tips comprise part of
the worker’ s earnings, p le a s e enter the average full-time earnings which a new worker can be expected to receive.
Column 6: For each pay rate (or pay rate range) entered in column 5, p lease enter the basis on which the offered pay rate is
quoted, that is , whether the pay rate given is per hour, per week, per month, etc.
Column 7:
Leave blank.
For o ffic e use only.
Column 8: For each occu pation, enter in column 8 the number o f vacan cies included in column 3 which are for part-time p o si
tions. part-time va ca n cies are defined as positions for which the weekly hours to be worked are le s s than the customary
straight-time work week for the occupation. If the job vacancy is open to either full-time or part-time workers, p lea se re
port it as a full-time vacan cy. If none, enter " 0 . " If the information cannot be provided, enter " X . " Do not leave blank.
IF YOU HAVE ANY QUESTIONS, PLEASE TELEPHONE:
_____________________________________________ ;________________________________________________
( Na me )
at
___ ______________________________________________
( T e l e p h o n e Number)
When report is com pleted, p le a s e return it in the accompanying self-addressed stamped envelope by April 8 , 1966.
DO SO EVE " T H0UGH YOU HAVE NO VACANCIES TO REPORT. Thank you for your cooperation.
PLEASE
JOB
V ACANCY
209
STATISTICS
U.S. DEPARTMENT OF LABOR
Budget Bureau No.
App r o v a i E x p i r e s :
BUR EAU OF L A B O R S T A T I S T I C S
B UREAU OF E M P L O YM E NT S E C U R I T Y
W A S HI NG T ON . D . C .
20210
HH-6609
8 / 31 / 66
REPORT ON JOB VACANCIES
INFORMATION FURNISHED
N O R P U B L I S H E D I N SU CH
ON T H I S R E P O R T I S S T R I C T L Y C O N F I D E N T I A L AND W I L L N O T B E
A M A N N E R T H A T D A T A R E L A T I N G T O I N D I V I D U A L C O M P A N I E S CAN
S T R U C T I O N S O N T H E B A C K O F T H I S F O RM
I N G E N V E L O P E BY A P R I L 8 , 1 9 6 6 T O :
BEFORE
(NAME
E N T E R I N G T H E D A T A ON T H I S
AND A D D R E S S O F A G E N C Y )
_______________________________ (C H AN GE NAME OR A D D R E S S
IF
INCORRECT)
R E V E A L E D T O ANY
BE I D E N T I F I E D .
REPORT,
AND
RETURN
THE
------------------------------------
THE
USE
Area
(2)
(1)
1
IN
DO N O T
Date
State
r
U NA UT H OR I ZE D PERSON
P L E A S E READ THE IN
REPORT
SIC
(4)
(3)
To
L
ACCOMPANY
Base
Si z e
TC
( 5)
NC
(8)
(7)
ORD
ES
( 9 ) ( 1 0)
(11)
( 6)
Cards
(12)
Wk
(1?)
1. T O T A L N U M B E R O F E M P L O Y E E S :
WORKED OR R E C E I V E D
OF MARCH
(SEE
PAY
FOR ANY
IN S TR U C TIO NS
e n t e r t h e t o t a l n u m b e r o f e m p l o y e e s who
P A R T OF THE P A Y P E R I O D WHICH I N C L U D E S THE I 2 T H
ON THE
BACK
2. J O B V A C A N C IE S BY O C C U P A T IO N :
FORM) I N YOUR E S T A B L I S H M E N T . AND THE
E ST P O S S I B L E DA Y ).
A JO B V AC A N C Y I S
OR T E M P O RA R Y WO RKE RS WHICH YOUR F I R M
OF T H I S
F O RM ) .
L I S T BELOW, BY JO B T I T L E .
A L L J OB V A C A N C I E S ( A S DE r I N ED ON THE BACK OF T H I S
R ATE OF PAY O F F E R E D . AS OF THE C L O S E OF B U S I N E S S . A P R I L I . 1 9 6 6 ( OR THE N E A R
A JOB O P E N I N G TH AT I S U N F I L L E D AND I M M E D I A T E L Y A V A I L A B L E TO F U L L - . P A R T - T I M E .
I S A C T I V E L Y S E E K I N G FROM O U T S I D E YOUR C OMPANY.
NU MBE R OF
JOB V A C A N C I E S
OCCUPATIONAL
CODE
( L E A V E B L A NK )
(CO NT IN U E
ON BACK
OF FORM
R ATE OF PAY
OFFERED
NU MBE R
V AC A N T
ONE MONTH
OR L ON G ER
P ER HOUR.
DAY . WEEK,
MONTH.
E TC .
(4 )
(6 )
IF
MO RE
SP ACE
IS
P L E A S E C H E C K THI S BOX IF Y OU WOU L D L I K E T H E A S S I S T A N C E OF T H E
( N A M E O F S T A T E A G E N C Y ) IN F I L L I N G A N Y O F Y O U R J O B V A C A N C I E S
S I GN a t u r e
( P e r son
r es ponsi ble
tor
this
report)
OTHER
JOB
O P E N IN GS »
NEEDED)
• OT HE R J O B O P E N I N G S ARE O P E N I N G S NOT I N C L U D E D I N COLUMN 3 AND R E L A T I N G TO J O B S
FOR F I L L I N G I M M E D I A T E L Y . BUT FOR WHI CH YOU ARE A C T I V E L Y S E E K I N G NEW WO RKE RS .
3.
(LEAVE
BLANK)
O CCUP IED
OR NOT
R E A DY
□
JOB VACANCY
210
STATISTICS
2. JOB V A C A N C IE S BY O C C U PATIO N : (Continued!
NUMBER OF
JOB V A C A N C I E S
O CCUP ATI ONAL
CODE
( L E A V E BLANK)
JOB T I T L E
TOTAL
(3)
(2 )
(1)
RATE OF PAY
OF FE RE D
NUMBER
VACANT
ONE MONTH
OR LONGER
AMOUNT
(4 )
PER HOUR
DAY. WEEK.
MONTH.
ETC.
(5)
m
(L EA VE
BLANK)
OTHER
JOB
OPENINGS*
(7)
(8 )
INSTRUCTIONS FOR REPORT ON JOB VACANCIES
1. Number o f E m ployees:
Eater La the box to the right of item 1 the total number of em ployees on all payrolls of your e sta b lish
ment who worked full-tim e or part-time or received pay for any part of the pay period which includes the 12th o f March. Include
persons on vacation and sick leave who received pay directly from your firm for the pay period reported, but exclude persons
on leave without company pay the entire period, and pensioners and members o f the Armed F orces carried on the rolls but not
working during the pay period* P le a se enter this figure even though you report no va ca n cies.
2. Job V acan cies, by Occupation: Enter the number o f current, unfilled job openings in your establishment as of the c lo se of
b u sin e ss, April 1, 1966 which are immediately availab le for occupancy by workers from outside your firm and for which your
firm is actively seeking such workers.
openings.
Included are full-tim e, part-time, permanent, temporary, se a so n a l, and short-term job
"A c t i v e ly se e k in g * 'is defined as current efforts to fill the job with a worker from outside your firm through: (1) so licitin g a sslsta n ce o f public or private employment a gencies, school or co lle ge placement o ffic e s , labor unions, employee groups, busi
n e s s or professional organizations, b u sin ess a s s o c ia te s , friends and em ployees in locating suitable candidates; (2 ) u sing
"h e l p wanted’ * advertising (newspaper, magazine, radio, te levision , direct m ail, posted n o tice, e tc .); (3) conducting recruit
ment programs or cam paigns; (4) interviewing and selectin g " g a t e ,” "w alk-in** or " m a i l ’ * applicants or workers searched out
o f applicant file s ; and (5 ) opening or reopening the acceptance of applications from prospective candidates.
Do not include as v a ca n cie s (1) jobs held for em ployees who will be recalled; (2) jo b s to be filled by transfer, promotion, or
demotion; (3) jobs held for workers on paid or unpaid leave; (4) jobs filled by overtime work which are not intended to be
filled by new workers; (5) job openings for which new workers were already hired and scheduled to start work at a later date;
and (6) those jobs unoccupied because o f labor-management disputes.
Column 1: L eave blank.
For office use only.
Column 2:
L ist the names o f the occupations for which job vacan cies e xist in your establishm ent. Where p o ssib le , add modi
fier denoting m aterial, product, process or subject matter to make establishment title more sp e c ific , e .g ., "a s s e m b le r , air
craft, wing parts” or "sten o g ra ph er, le g a l.” U s e a single entry to report job v acan cies with identical job title s where more
than one vacancy e x is ts . If there are several c la s s e s or grades for sp ecific job title s, each c la s s or grade should be liste d
separately.
Column 3:
For each occupation listed in column 2, report the total number o f job vacancies.
Column 4:
For each occupation, enter the number o f job vacancies included in column 3 which have been vacant for one month
or longer. If none, enter " 0 ” . If information cannot be provided, enter " X .* * . Do not lea v e blank.
Column 5: For each occupation, enter in column 5 the rate of pay offered for the job to which the vacancy relates. P le a s e en
ter a single rate o f pay wherever p o ssib le . However, where a range o f pay rates is offered, depending upon the education,
training, and experience of the job applicant, the entry le v e ls o f the low and high pay rates offered will be acceptable.
P le a se enter hourly pay rates in column 5 wherever p o s s ib le .
If the wage rate offered for the job vacancy is on a p iece work or commission incentive basis, or if tips comprise part of the
worker’ s earnings, p le a se enter the average full-time earnings which a new worker can be expected to receive.
Column 6: For each pay rate (or pay rate range) entered in column 5, p lea se enter the b a sis on which the offered pay rate is
quoted, that i s , whether the pay rate given is per hour, per week, per month, etc.
Column 7:
L e a ve blank.
For o ffice use only.
Column 8:
Enter in column 8, by occupation, the number o f openings (not included in column 3) which are "o th e r ” open ings.
"O th e r ” openings are defined as position s for which your firm is actively seeking new workers but which p osition s are cur
rently occupied or unavailable for immediate occupancy for such reasons a s : job unavailable until expected separation o f
present incumbent o ccu rs; work will not start until some future date; new branch to be opened in future; or anticipated in
crease in b u sin e ss. If none, enter " 0 ” . If the information cannot be provided, enter " X ” . Do not leave blank.
IF YOU HAVE ANY QUESTIONS, PLEASE TELEPHONE:
( N ame )
at
( T e l e p h o n e Number)
When report is com pleted, p lease return it in the accompanying self-addressed stamped envelope by April 8, 1966.
DO SO EV5N THOUGH YOU HAVE NO VACANCIES TO REPORT. Thank you for your cooperation.
PLEASE
E n c l o s u r e 3— N u m b e r o f J o b V a c a n c i e s b y O c c u p a t i o n a n d I n d u s t r y a n d b y A r e a , A p r i l 1965
IN D U STRY
Occupation
Total
Durable Nondur
Con T.C.P.U.i
able
struction
Trade
F.I.R.E.2
Service
Govern
ment
Other
BALTIMORE
Total______________________________________________________________
1,284
851
956
237
2,035
1,090
0
19
86
2
65
198
30
0
31
0
0
752
344
157
0
3
101
0
98
444
202
3
41
341
234
165
111
64
0
28
205
4
0
0
0
0
339
759
396
52
404
83
2
431
301
273
40
7
25
13
0
0
0
0
0
0
0
____
1,854
276
114
163
130
310
184
392
285
0
Professional and managerial..
__________________________ ____
Clerical and sales--------------------- ----------------------- ------- -----Service---------- ----------_ _ „ _____ ______________
_____
Skilled_________________________________________________________
Semiskilled. __ ________
__ ___ _________________________
Unskilled_______________________________________________________
Other- _
______
___ ___________ __________ ___ _
454
417
255
244
219
262
3
27
13
0
82
16
138
0
10
8
0
3
64
29
0
37
0
0
42
0
84
0
1
48
4
38
27
9
3
18
90
95
30
75
2
0
11
173
0
0
0
0
0
168
64
96
31
33
0
0
182
21
60
18
4
0
0
0
0
0
0
0
0
0
1,413
57
79
140
53
166
47
104
767
0
238
216
119
459
228
153
0
9
3
0
21
12
12
0
4
3
0
33
30
9
0
0
0
26
76
17
21
0
1
4
0
6
42
0
0
1
66
28
38
28
5
0
0
47
0
0
0
0
0
29
6
11
0
18
40
0
194
87
54
285
81
66
0
0
0
0
0
0
0
0
______________
655
34
7
0
1
166
64
108
266
9
Professional and managerial____
______________ ______________
Clerical and sales. _ ___ ______ ____________________________
Service. . . . __ ___ _______ ________________________ ___ _ _
Skilled__________________________________________________________
Semiskilled______________________________________________________
Unskilled____ _ __ ___ _ ___ ___ _________________ _____
Other__ _ ______ _______ _______ _ ___ _ ...................... .
288
224
60
29
33
21
0
2
3
0
23
3
3
0
2
5
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
10
90
29
4
25
8
0
5
59
0
0
0
0
0
67
6
28
1
5
1
0
202
60
3
1
0
0
0
0
0
0
0
0
9
0
BIRMINGHAM
Total______
___
_
______
_ ___
_____________________
c h a r l e s t o n , s .c .
Total______________________________________________________________
Professional and managerial______
______________ _____ ____
Clerical and sales______
___
_______________ ____
Service _ _ _ _
..
__ ________________ __ ________ _
Skilled_________________________________________________________
Semiskilled... _ ____
_
__ ____
___ _ _ _____ ____
Unskilled_______________________________________________________
_________
_
_ __________________ _______
Other-----c h arleston, w
Total____
__
______
.
va.
______________
211
___ __
STATISTICS
400
76
34
1
142
45
229
1
VACANCY
528
968
1,827
910
1,314
1,553
790
19
JOB
7,381
Professional and managerial___________ ___ ________________
_
Clerical and sales____________________________________ ___ ..
._ _______________ _
Service___ _______________ _ ______
Skilled______________________________________________ __________
Semiskilled______________
____ _____ ________ _____ ____
Unskilled_____
_
___
_ ___ ______ _______ _______ _
Other ______________________ ___ _________ _________ __ .
E n clo su re
3— N u m b e r
Occupation
CHICAGO
Total_______________________________________
Professional and managerial____________
Clerical and sales______________________
Service_________________________________
Skilled_________________________________
Semiskilled____________________________
Unskilled______________________________
Other__________________________________
KANSAS CITY
Total______________________________________
Professional and managerial. _--------------Clerical and sales______________________
Service_________________________________
Skilled_________________________________
Semiskilled____________________________
Unskilled______________________________
Other__________________________________
LOS ANGELES
Total______________________________________
Professional and managerial-----------------Clerical and sales---------------------------------Service________________________________
Skilled ______________________________
Semiskilled-------------------------------------------Unskilled______________________________
Other__________________________________
MIAMI
Total______________________________________
Professional and managerial___________
Clerical and sales----------------------------------Service_________________________________
Skilled_________________________________
Semiskilled--------------------------------------------
of
Job
V a c a n c ie s
i
212
\y
O c c u p a t i o n a n d I n d u s t r y a n d b y A r e a , A p r i l 1965— Continued
IN D U STRY
Total
T.C.P.U.1
Con
Durable Nondur
struction
able
Trade
F.I.R.E.2
Service
Govern
ment
Other
2
1,512
6,728
1,617
3,706
3,049
466
401
751
22
114
453
240
1
0
2
0
0
0
0
0
30
528
49
123
735
47
0
1,249
2,843
932
1,512
58
134
0
175
1,337
105
0
0
0
0
1,919
419
1,030
173
12
153
0
1,363
1,117
167
308
10
84
0
42
46
6
2
148
0
222
2,836
409
198
27
74
736
138
882
272
100
'
562
1,075
126
556
378
70
69
45
25
3
296
23
17
0
8
18
1
34
127
10
0
7
1
0
7
0
12
0
4
51
1
15
1
2
0
39
513
4
2
173
5
0
4
132
2
0
0
0
0
393
188
104
193
2
2
0
62
147
11
9
19
22
2
0
0
0
0
33
0
67
.
22,506
5,028
3,467
370
742
3,342
1,059
5,454
3,034
10
6,061
5,453
2,435
1,902
5,546
1,073
36
1,872
379
8
1,048
1,276
445
0
77
203
2
96
3,065
24
0
340
14
0
12
0
4
0
69
335
12
181
129
16
0
53
1,915
715
140
418
101
0
48
999
8
2
2
0
0
2,306
827
1,172
273
528
346
2
1,294
781
518
150
128
129
34
2
0
0
0
0
8
0
4,037
307
216
35
1,020
977
179
846
352
105
432
1,019
672
573
1,057
9
9
0
125
116
15
64
1
35
30
19
231
o
30
740
17
282
385
207
59
26
81
1
71
0
303
248
153
6
71
41
98
121
28
17
2
4
11
65
23
0
2 ;
o ;i
8 1!
1 !I
STATISTICS
1,982
626
550
87
4,275
2,871
475
1
VACANCY
8,885
5,805
7,593
2,398
6,507
4,287
1,133
224
JOB
27,947
Unskilled.
Other____
MILWAUKEE
Total________________ __________________________
Professional and managerial________________
Clerical and sales----------------------------------------Service---------------------------------------------------------Skilled ____________________________________
Semiskilled______ __________________________
Unskilled___________________________________
Other______________________________________
MINNEAPOLIS-ST. PAUL
Total___________________________________________
Professional and managerial________________
Clerical and sales___________________________
Service_____________________________________
Skilled_____________________________________
Semiskilled_________________________________
Unskilled__________________________________
Other__________________________ ____________
NEW YORK
Total___________________________________________
Professional and managerial________________
Clerical and sales___________________________
Service_____________________________________
Skilled_____________________________________
Semiskilled________________________________
Unskilled__________________________________
Other______________________________________
PHILADELPHIA
Total___________________________________________
Professional and managerial________________
Clerical and sales__________________________
Service____________________________________
Skilled_____________________________________
Semiskilled________________________________
Unskilled__________________________________
Other______________________________________
26
0
0
0
27
0
0
0
65
0
44
3
0
0
8,542
1,420
550
391
394
1,730
241
3,008
808
0
1,432
1,323
1,393
1,772
1,228
1,226
168
272
148
7
631
178
184
0
23
83
61
46
242
95
0
0
0
0
326
0
65
0
16
67
12
111
160
28
0
90
384
378
197
455
205
21
24
214
3
0
0
0
0
634
312
746
441
169
591
115
368
121
186
18
23
60
32
0
0
a
0
0
0
0
8, 061
860
1,478
224
554
1, 318
384
1,112
931
1,200
2,126
2, 876
445
329
763
317
1,205
372
206
19
97
149
17
0
720
550
3
99
34
72
0
0
0
0
37
0
187
0
4
207
2
0
341
0
0
139
878
145
25
105
26
0
42
340
0
2
0
0
0
386
449
175
50
41
7
4
463
246
101
19
93
8
1
0
0
0
0
0
0
1,200
46, 780
5,263
5,941
1,008
4,130
6,097
2, 881
8,814
12,628
18
12, 358
16.136
2,999
4,981
7,216
3, 090
0
1,090
371
40
2,223
1,227
312
0
487
1, 381
560
3,170
336
0
153
14
0
841
0
0
0
104
2,471
80
335
953
187
0
1, 535
2, 435
1,091
438
446
152
0
248
2,553
63
4
13
0
0
3,190
3,104
1,194
327
991
8
0
5, 551
3,789
524
253
416
2,095
0
0
18
0
0
0
0
0
15,037
2,343
2, 408
1, 356
523
2,879
428
3,114
1,986
0
2, 739
2,955
1,934
3, 407
3,125
877
0
433
146
3
296
116
2
240
1,492
262
0
41
0
0
1,163
0
39
151
0
0
327
6
0
73
1,135
1,115
349
100
107
0
59
365
1,277
591
227
262
750
521
451
584
273
7
82
0
0
0
a
0
0
0
0
1,119
381
261
0
7
152
0
3
1
0
0
0
0
75
STATISTICS
71
0
VACANCY
48
0
JOB
281
3
to
I—I
00
214
E nclosure 3— N u m b e r
of
J ob V acancies
by
O ccupation
and
I ndustry
and
By A r e a , A pril 1965— Continued
IN D U STRY
Occupation
PORTLAND, OREG.
Total
Con T.C.P.U.1 Trade
Durable Nondur
struction
able
567
343
58
212
1
0
0
0
0
-
265
71
204
1
25
0
1
■"
24
202
22
161
568
863
521
383
304
229
25
48
19
0
267
101
96
0
5
37
1
35
75
49
0
3
0
0
19
0
0
0
2
12
40
401
254
36
37
16
44
57
32
2
0
0
—
- ■ .■
■ ,. ... ---- ■
Total_______________________________________________________________
5,661
1,750
1,778
125
135
411
144
814
386
118
Professional and managerial______________________________________
Clerical and sales_______________________________ _______ ________ .
Service. _ __ _________________________________________________
Skilled. .
_____________________________ ______________________
Semiskilled__________________________ ___________________________
Unskilled
_ _
_____________ _ __________________________
Other______________________________________________________ ____
196
623
697
1,028
2,570
542
5
20
60
7
491
873
299
0
35
11
0
270
1,325
137
0
0
0
0
100
25
0
0
0
4
0
0
127
4
0
2
89
186
34
98
2
0
4
108
0
16
0
16
0
16
258
428
23
44
45
0
103
71
16
86
71
39
0
16
22
60
8
7
0
5
4,015
201
289
785
258
699
401
788
594
0
0
11
0
220
53
501
0
0
10
3
11
234
0
0
14
291
166
138
57
25
8
7
386
4
0
2
0
2
160
58
161
109
143
157
0
275
216
44
30
26
2
1
0
0
0
0
0
0
0
Professional and managerial
Clerical and sales
......
Service__
_______ __ ______ ___ _____________________________
___ ______________________________________
S killed...
. .
Semiskilled— ___________ ____ ________________________
Unskilled_______________________________________________________
Other___
_ __________________________________________________
PROVIDENCE
RICHMOND
Total_______________________________________________________________
Professional and managerial___________________ _________________
Clerical and sales__________ _____ _______________ ________________
Service__
_
_________ ____________________________
Skilled-..
__________________________ _________________
Semiskilled—
. . .
. . _____________ ______
Unskilled-_____ ____________________
..........................................
Other __ .
______________________ _____________ _______ ______
581
1,053
379
565
670
755
12
Source: U.S. Department of Labor, Manpower Administration, Bureau of Employment Security, U.S. Employment Service, Washington, D.C., May 31,1966.
81
7
0
47
32
34
0
44
74
1
10
123
36
1
1 Transportation, communications, and public utilities,
2 Finance, insurance, and real estate.
0
175
0
83
25
0
8
0
0
2
0
50
24
0
1 :-r:__ - __-
STATISTICS
531
VACANCY
271
Other
JOB
Govern
ment
2,893
Total__________ ___________________________________________________
772
Service
F.I.R.E.*
E nclosure
4— N um b er
of
V acancies for W hich W age A n alysis C ould B e M ade and N um ber
V acancies W hich W ere S u bstand ard , b y A r e a ,1 A pril 1965
Baltimore
Total
Substandard
Num
ber
Per
cent
Num
ber
Substandard2
Num
ber
Per
cent
Num
ber
Substandard3
Num
ber
Per
cent
Num
ber
Substandard 4
Num
ber
Num
ber
Per
cent
Minneapolis-St. Paul
Substandard 8
Num
ber
T hese
Num
ber
Substandard 6
Num
ber
Per
cent
Per
cent
12,711
2,196
17.3
1,170
253
21.6
4,646
841
18.1
3,703
431
11.6
1,102
262
23.8
2,090
409
19.6
Professional and
managerial_______
Clerical and Sales. -Service---------- -------Skilled____________
Semiskilled .
Unskilled__________
3,962
3,901
1,174
1,504
1,408
762
918
507
205
272
190
104
23.2
13.0
17.5
18.1
13.5
13.6
373
277
219
133
52
116
173
31
18
26
1
4
46.4
11.2
8.2
19.5
1.9
3.4
953
1,628
382
696
588
399
274
177
47
150
130
63
28.8
10.9
12.3
21.6
22.1
15.8
1,594
921
184
417
514
73
181
131
31
39
46
3
11.4
14.2
16.8
9.4
8.9
4.1
219
316
207
165
68
127
49
52
77
43
13
28
22.4
16.5
37.2
26.1
19.1
22.0
823
759
182
93
186
47
241
116
32
14
0
6
29.3
15.3
17.6
15.1
.0
12.8
Source: U.S. Department of Labor, Manpower Administration, Bureau of Employ
ment Security, U.S. Employment Service, Washington, D.C., May 31, 1966.
1 These data are based only on actual vacancies reported rather than on an inflation
of vacancies to a universe total.
2 Primarily draftsmen, registered nurses, laboratory technicians, and medical tech
nicians.
3 Primarily electrical engineers, registered nurses, electrical and mechanical drafts
men, nurses aids, and chassis assemblers.
* Primarily civil engineers and general office clerks.
•Primarily general office clerks and police officers.
« Primarily registered nurses, practical nurses, and highway technicians.
215
STATISTICS
Total, all groups_______
VACANCY
Num
ber
of
JOB
Occupational group
P ercent
Miami
Los Angeles
Chicago
and
@FedFRASER