95th Congress 1
2d Session I
JOINT COMMITTEE PRINT
PUBLIC HEARINGS BEFORE THE
NATIONAL COMMISSION ON EMPLOYMENT
AND UNEMPLOYMENT STATISTICS
Hearings held in Washington
May 9 and 10, 1978
JOINT ECONOMIC COMMITTEE
CONGRESS OF THE UNITED STATES
OCTOBER 3, 1978 \
Printed for the use of the Joint Economic Committee
U.S. GOVERNMENT PRINTING OFFICE
WASHINGTON : 1978
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20402
Stock No. 052-070-04690-2
JOINT ECONOMIC COMMITTEE
(Created pursuant to sec. 5(a) of Public Law 304, 79th Cong.)
RICHARD BOLLI NG, Missouri, Chairman
LLOYD BENTSEN. Texas, Vice Chairman
HOUSE OF REPRESENTATIVES
HENRY S. REUSS, Wisconsin
WILLIAM S. MOORHEAD, Pennsylvania
LEE H. HAMILTON, Indiana
GILLIS W. LONG, Louisiana
PARREN J. MITCHELL, Maryland
CLARENCE J. BROWN, Ohio
GARRY BROWN, Michigan
MARGARET M. HECKLER, Massachusetts
JOHN H. ROUSSELOT, California
JOHN SPARKMAN, Alabama
WILLIAM PROXMIRE, Wisconsin
ABRAHAM RIBICOFF, Connecticut
EDWARD M31. KENNEDY, Massachusetts
GEORGE McGOVERN, South Dakota
JACOB K. JAVITS, New York
WILLIAM V. ROTH, JR., Delaware
JAMES A. McCLURE, Idaho
ORRIN G. HATCH, Utah
JOHN R. STARK, Executive Director
LOUIS C. KRAUTHOFF II, A88ssistant Director
RICHARD F. KAUFMAN, Assistant Director-General Counsel
LLOYD C. ATKINSON
WILLIAM R. BU'ECHNER
THOMAS F. DERNBURG
ROBERT HI. ATEN"
KENT H. HUGHES
L. DoLuGLAS LEE
GEORGE R. TYLER
CHARLES H. BRADFORD
M'. ARK R. POLICINSKI
PAUL B. MANCHESTER
DEBORAH NORELLI MATZO
M. CATHERINE MILLER
STEPHEN J. ENTIN
LETTERS OF TRANSMITTAL
SEPTEMtBER 26, 1978.
To the Members of the Joint Economic Conmmittee:
Transmitted herewith are the transcripts of the initial public hear-
ings conducted by the National Commission on Employment and Un-
The Joint Economic Committee has always maintained a deep
interest in the evolution of the statistics on employment and unemploy-
ment to meet changing legislative needs. For that reason we have
been pleased to participate as advisers to the National Commission on
Employment and Unemployment Statistics, whose mandate covers this
Because the public hearings held by the Commission provide in-
formative and valuable material from several different sources, the
committee has agreed to publish the transcripts in order to provide
widespread dissemination. I believe that members of the Joint Eco-
nomic Committee and other Members of Congress will find them most
The views expressed in the transcripts are those of the witnesses and
do not necessarily represent the views of the members of the Joint
Economic Committee or the committee staff.
Chairman. Joint Econom c Cam m itee.
SEPTEMBER 19, 1978.
Hon. RICHARD BOLLING,
Chairman, Joint Econom ; Commitnntee, U.S. Congr's., Washington,
DEAR MR. CHAIRMAN: Transmitted herewith are the transcripts of
the initial public hearings conducted by the National Commission on
Employment and Unemployment Statistics.
The Joint Economic Committee has maintained a continued interest
in the formulation of statistics on employment and unemployment. As
you are well aware, these data are under increasing scrutiny because
past, legislation has placed insupportable demands on these statistics.
In the initial process of examining various alternatives to existing
methods of data collection and presentation, the Commission on Enm-
ployment and Unemployment Statistics held public hearings. Wit-
nesses included persons from congressional, academic, government.
and public sectors. Their combined testimony gives the Joint Economic
Committee a valuable and broadly based compendium of information.
The committee's undertaking to publish these hearings will enable
a wide-ranging audience to review the material. The expected feed-
back from interested parties should provide another source of im-
portant insight in our studies. Public dissemination also will focus
attention on the complexities and ramifications implicit in any changes
recommended by the Commission.
The transcripts were prepared for publication under the direction
of Sar Levitan, the Chairman, Marc Rosenblum and Lois Black of the
The views expressed in the hearings are those of the respective wit-
nesses and do not necessarily represent the views of the Joint Economic
Committee or any of its individual members.
JOHN R. STARK,
E., u1 t '1' Director. Joint E roiomic? Committee.
NATIONAL CO11MiISSION ON EMPLOYMENT AND
UNE3M PLOYMENT STATISTICS,
Washington. D.C., August 16, 1978.
Mr. JOHN R. STARK,
Ee'rcutive Db-ctor, Joint Economn c Conmmittee,
U.S. Congrc.,s., Washington, D.C.
I)EAR MR. STARK: I am sending you transcripts of the initial public
hearings conducted by the National Commission on Employment and
Inemniployment Statistics. The public hearings were an educational
experience for us, and we hope to learn even more from the feedback
when the transcripts are printed by the Joint Economic Committee.
The record also should be of interest to other committees.
The material has been placed in three volumes as follows:
Volume I. Hearings held May 9-10, 1978. in Washington, D.C.
Volume II. Hearings held on May 23,1978, in New York City;
June 13, 1978. in Chicago: and June 20,1978, in San Francisco.
Volume III. Hearings held on July 11, 1978, in Atlanta; July
26, 1978, in Washington, D.C., and written submissions by wit-
nesses who could not appear in person to testify.
Our joint agreement to publish the material will help the Commis-
sion in its task to investigate and improve our system of labor force
statistics. Due to the growing use of these data in the formation and
implementation of government policy, the Commission strongly be-
lieves the issues should be presented to the public and not just a small
charmed circle of economic experts.
Thank you again for your continued interest in the work of the
Letters of Transmittal ......................... ii
Opening Statement of Chairman Sar A. Levitan ... 1
CHRONOLOGICAL LIST OF WITNESSES
TUESDAY, MAY 9, 1978
Boiling, Hon. Richard, chairman of the Joint
Economic Committee .......................... 2
Fellner, William, Herbert Stein and
Marvin Kosters, The American
Enterprise Institute for Public
Policy Research ... . ......... . . . . .... 15
Coleman, Edwin J., chief, regional economic
measurement division, U.S. Department
of Commerce, Bureau of Economic Analysis.. 38
Barrett, Nancy Smith, director, Research
on Women and Family Policy,
The Urban Institute ...................... 52
Ossofsky, Jack, executive director,
The National Council on the Aging ........ 65
Kalachek, Edward, professor of economics,
Washington University .................... 78
Motley, Brian, professor of economics,
University of Kentucky ................... 88
Bergmann, Barbara, professor of economics,
University of Maryland ................... 115
Yee, Albert, Census Advisory Committee
on Asian and Pacific Americans
Population for the 1980 Census ........... 128
Cardenas, Gilbert, professor of economics,
Pan American University .................. 140
Carter, Lewis, national labor director,
National Association for the
Advancement of Colored People ............ 151
WEDNESDAY, MAY 10, 1978
Pines, Marion, director, Baltimore
Metropolitan Manpower Consortium;
accompanied by Mark R. Horowitz,
staff analyst ........... . . . . . . . ... 160
Beals, L. Alan, executive vice president,
National League of Cities ................ 181
Moore, Geoffrey, National Bureau of
Economic Research ............ .. .......... 197
Ruggles, Richard, professor of economics,
Yale University . ......... . . . . . .... 216
Harrington, Hon. Michael, a U.S.
Representative from the State of
Massachusetts, representing the
Northeast-Midwest Coalition of
Congressmen; accompanied by James
Costello, legislative assistant .......... 230
Arnold, Richard, director, research and
statistics, Utah Department of
Employment Security; accompanied by
Samuel Morgenstein, staff member,
Interstate Conference of Employment
Security Agencies ........................ 252
Leach, Hon. Daniel E., vice-chair,
United States Equal Employment
Opportunity Commission ................... 268
Mister, Melvin, director, Institute
for Urban and Regional Economic
Analysis, U.S. Conference of Mayors ......
Urban League ...........
Mangum, Garth, professor of economics,
University of Utah .............
Appendix A. Additional Submission by
Alan Beals, executive director,
National League of Cities .......
Appendix B. Methods for Collecting
Employment and Unemployment
Statistics, Additional Submission
by Melvin Mister, director, Insti-
tute for Urban and Regional Analysis,
U.S. Conference of Mayors ................
. 0 0 *
Digitized by the Internet Archive
NATIONAL COMMISSION ON EMPLOYMENT
AND UNEMPLOYMENT STATISTICS
(Created pursuant to Sec. 13 of Public Law 444,
Sar A. Levitan, Chairman
Bernard E. Anderson
Glen G. Cain
Michael H. Moskow
Samuel L. Popkin
Joan L. Wills
James T. Mclntyre, Jr.
Jacob K. Javits
James M. Jeffords
Carl D. Perkins
Harrison Williams, Jr.
Arvil V. Adams, executive director
Robert Guttman, general counsel
Wesley H. Lacey, administrative officer
Curtis Gilroy, staff economist
Joseph Hight, staff economist
Janice Madden, staff economist
Marc Rosenblum, staff economist
Diane Werneke, staff economist
Lois Black, research analyst
Gary Solon, research analyst
TRANSCRIPT OF PUBLIC HEARINGS
TUESDAY, MAY 9, 1978
NATIONAL COMMISSION ON EMPLOYMENT
AND UNEMPLOYMENT STATISTICS
The Commission met pursuant to notice, at 9:30
a.m., in room 5437, 200 Constitution Ave., N.W., Sar A.
Levitan, Chairman, presiding.
Present: Bernard E. Anderson, Jack Carlson,
Michael H. Moskow, Samuel L. Popkin, and Joan L. Wills.
Also present: Arvil V. Adams, executive director;
Marc Rosenblum, staff economist; and Wesley H. Lacey,
OPENING STATEMENT OF CHAIRMAN LEVITAN
CHAIRMAN LEVITAN: This is the first of a series
of hearings that the National Commission on Employment
and Unemployment Statistics is going to hold. There
is, therefore, a strong desire to tell a captive
audience about all the good things we have done and
what we are going to do.
But since we have gathered here not to make
speeches but to listen to good advice, we are going to
start with a gentleman who has been a user, inter-
preter, and a policymaker of employment and unemploy-
Mr. Boiling, we are honored to have you with us.
Mr. Boiling is the Chairman of the Joint Economic
Committee and a statutory advisor to the Commission.
Mr. Boiling, we will be three times as good as the
Speaker of the House is. We give 15 minutes instead of
STATEMENT OF THE HONORABLE RICHARD BOLLING,
CHAIRMAN, JOINT ECONOMIC COMMITTEE,
UNITED STATES CONGRESS
HR. BOLLING: Mr. Chairman, I may not even take
the 15 minutes. I have thought a long time about what I
should say at this meeting because I think this is a
very important Commission, and I think that the work it
does will be critical to the future decisions that
policymakers at all levels make. And I wish I felt
wiser about how this Commission was going to solve some
of the problems that I see ahead.
As is usual, I think in American politics--and
maybe I am negative on the subject--commissions like
this are not included in what I am about to say, but
most groups tend to avoid the very difficult problems
they face and deal instead with the ones that are
perhaps easier because they are more finite.
And I have thought a lot about employment and
unemployment for a very long time, much longer than I
have been in Congress and I will not share with you my
experience in the '30s as a very young man, but I think
the problem that we have today is that we do not have
the vaguest idea what we are talking about when we talk
Of course, we define it and Julius Shiskin can
give you a very precise statement of what it is. So, I
decided to raise the things that worry me, not neces-
sarily the things that you can deal with because I
think if we do not deal with the things that worry me
and all the others too, the policymakers are still
going to be operating in the dark.
Now, when the Commission organized, I had the
privilege of being present then too and each one of
these, I think, was mentioned rather briefly and I am
going to mention them very briefly.
I do not understand how we deal with the problem
of individual employment in the current situation which
is too great a change from the days of depression, when
there are so many different confirmations of employment
in families, unless we know a great deal more than we
do about those confirmations.
Now, I do not understand even how to get at that
except through, perhaps, a census. There may be a
great deal more information on that than I read about,
but I do not think we get a true measure of employment
and unemployment just on individual employment statis-
tics. At least as far as I know on the Hill, we do not
seem to have an adequate awareness of the potential.
Now, that leads to a very fundamental political
problem, and we are really talking about politics,
making policy in the end. We in the Joint Economic
Committee are not arguing about policy issues. We are
arguing about factual issues, and we do not have the
I do not want to harken back to the past, but when
I first went on that Committee more than 25 years ago,
or about 25 years ago, we sort of stipulated what the
issues were, what the facts were, and then proceeded to
have awful fights about the issues.
I do not exactly know how we have gotten so far
out of touch with the facts, but the major argument
that we are having today is over a percentage. What is
unemployment? What is full employment?
Now, I just think that we have got to get beyond
that or we cannot deal effectively with the policy of
dealing with unemployment and full employment. I do
not know how you go about doing that. I have no
I just know that unless the policymakers have some
kind of a factual base from which to work that the
arguments are a waste of time. The right-wing Repub-
lican and the moderate to left-wing Democrat, they are
just shooting over each others' heads; they are not
talking about the same thing.
Now, that is something that is talked about. Two
other things are not. One was mentioned briefly in
that organization meeting and that is the other
economy. Everybody has a different name for it. Some
of it obviously is illegal. I do not know what you
want to call those economies, but they are there, and I
suspect there is one that is probably semi-legal and I
suspect that the 5 million people that the census
thought they missed in the last census were probably
largely involved in that economy.
And if we do not know who they are and where they
are, I do not know how we are going to deal realis-
tically with one of our objectives, which is the prob-
lem of poverty, because, at least in my district, I
strongly suspect that a good many of those people that
were missed are blacks and whites in the central city
that are very hard to find, even if you walk the
streets. The census people did not find them.
And I think we have to deal with that. I gather
that that is a more prevalent thing in some European
countries than we are familiar with, although I have
not had a chance to read the literature.
Then, there is a third thing--and I am sure there
are many that I have missed--that is sort of taboo
because it is politically firey, at least taboo to
policymakers and we come up with the most peculiar
solutions when we come up with solutions at all. And
that, of course, is aliens, legal and illegal.
And that links in with a whole series of foreign
policy problems that I do not have the vaguest inten-
tion of starting to go into here, some immediate and
close and some distant and very, very painful and com-
But if we do not face that situation in some kind
of a realistic way, I do not see how we can deal with
the total overall problem of employment and unemploy-
ment. And all of these things are obvious. I do not
have any answers. I do know that we need answers.
And then there is one more and I will stop, and I
think this has been a subject of so much controversy
that I hardly need more than to say it. What is
poverty? Who is poor? I think it is a controversy
that has to not be resolved but put in some kind of a
perspective that is accepted. I do not mean accepted
in total or accepted by everybody, but I think we have
to try to get some kind of an acceptance of what this
year's or next year's American level of decency is.
And I know that sounds a little bit like a
political speech, but unless you can get the conser-
vative segments of the society to at least in part
support social programs, at least accept them, they do
not over time, in my opinion, work very well.
Social security has been mostly accepted. It was
not in the beginning. I think by and large it has been
a success and, although it is contested occasionally by
occasional candidates, it is different from the poverty
There are other illustrations and maybe I am
entirely wrong in my view of how the American society
and economy works and I will not bore you with my own
particular philosophy on it very long, but I do think
there has to be some acceptance.
I think the mistakes that legislators and policy-
makers sometimes make is to think that if they win a
battle, they have won a war over a social program.
And, in fact, all they have done is perpetuate the war
and set up, perhaps, the situation that will cause them
to lose the battle.
And I think it is intolerable that we really do
not know where the poverty is, the things I have men-
tioned, aliens, the other economy--in full--we do not
fully know. And we have not come to some kind of an
acceptance of the notion of what is enough.
Now, I will give you one illustration. I had a
good deal to do with the passage of the early civil
rights bills, without which there would not have been
later civil rights bills, for reasons that will be
obvious to some of you. The only way on earth that we
could pass those was to begin with a proposition that
was not de minimis--it was terribly important; it had
to do with voting--but that could be sold to honorable
And I think we have to approach some of these
problems that you face in terms of developing a factual
base from that point of view and that is why I think
that these three or four things that I have mentioned
are so important. And I am sure I have been totally
unhelpful, but I have at least had an opportunity in a
useful forum to say some things that I think are impor-
Thank you, Mr.
think, since we have all appeared on the other side of
the table from you, maybe we can get a chance to repay
MR. BOLLING: You have full opportunity.
CHAIRMAN LEVITAN: Thank you, Mr. Boiling.
MR. MOSKOW: Yes I would like to ask a question.
Focusing on your point about facts, of course, one of
the problems in looking at statistics is that everyone
wants more. There is almost like an infinite demand
for more labor force statistics, and I think with good
reason because they serve an important purpose.
But one of our problems is going to be to adjust
the priorities as to where we would recommend resources
be allocated, what the most important factual needs
are. And one of the areas that has been discussed a
great deal is, of course, the need for additional and
improved local and state unemployment statistics.
Now, I wonder if you would want to say a few words
about that and how high a priority you would place on
that area in terms of the allocation of future
MR. BOLLING: Well, I started to include a part of
that question in my own opening statement and I decided
not to because I did not think it was really sort of
fair to throw that one in too.
I am not at all sure, as I think about it--it may
be permanently fixed--but I am not at all sure that the
policymakers have been entirely wise in using that
device. But, not to avoid your question, merely to
raise the fact that I am not sure that it is wise to
have that particular set of tools in action.
I do not believe in mechanistic approaches very
much because I have never really seen them work very
well in my own experience. Of course, I may be wholly
But leaving that and accepting the situation,
clearly we are not going to be able to make the trigger
mechanisms that we have in some of these laws. And I
studied the enormous amount of money that is involved
and started to talk about that too.
Unless we have better local figures--an ability to
come up with more precise and accurate and unarguable
local numbers--and that means that they must have given
the present situation considerable priority--because
what we now have is a system that almost inevitably
makes a political issue every time an automatic func-
tion takes place because the ones that did not get more
than what they think they should have complain that
they did not get enough, and nobody is really in a very
So, what you are doing again in the system that I
question is you are just setting up more and more con-
flict instead of getting more and more resolution. So,
I think there has to be a very high priority, given the
use of that sort of mindless approach, nonselective
CHAIRMAN LEVITAN: Mr. Carlson?
MR. CARLSON: Your fourth point, dealing with an
acceptable definition, at least a minimum definition
that there tends to be a great number of people that
would be supported on poverty, could you get us a
little better feel of what you mean? We do have
poverty definitions now based on a nutritious but low-
cost diet and then some multiple of that gives us our
definition of poverty that we have now, and that gives
us what? Around $6,000 for a family of four, in that
range, plus or minus $500.
Are you thinking of a much more detailed defini-
tion, or multiple definitions or what?
MR. BOLLING: I am thinking of a variety of
definitions because I think any figure is bound to be
inaccurate, untrue. I have the good fortune to have
been born in the City of New York, growing up in north
Alabama in the depression. I worked in the far west;
my mother was from Wisconsin, and I represent Missouri.
So, I have absolutely no illusions about the
variety that exists in this country. To make the point
32-931 0 78 2
very clear, I have said--and nobody has successfully
contradicted me, at least to my satisfaction--that
there is a greater difference between the regions of
the United States than there is between France and
Germany, and I am aware that they have different
I do not know how in the devil you can come up
with a figure like $6,000. I do not know how you can
get to regionalism, but I do know that I do not find
One of the things that bothers me--I was born in
New York--and my automatic reaction to the problems in
New York was that the federal government could let New
York go under. That did not mean we shouldn't put some
strings on our aid--but one of their dilemmas up there,
as was rather articulately stated by a former Post
reporter, who is now a freelancer in New York whose
name I have forgotten, a woman--what is happening in
New York is killing the little people, not the well
That is because they have such an incredibly high
cost. Now, I do not know how you come up with one
figure. I am really talking about regionalism and
taking into account a variety of other things
including--I am perfectly willing to go to the point of
trying to figure out what family income is, but I do
not know how to do it.
I am not very helpful.
CHAIRMAN LEVITAN: You suggest, as I understand
it, that a trigger mechanism for fund allocation cannot
operate because we do not have proper local data. I
agree with you about the paucity of local data, but
isn't it true that it is a question of political
Because, as I understand it, right now as Congress
is considering CETA, we have lots of numbers. BLS, and
the Employment and Training Administration are bringing
up all sorts of simulations, and I think Congress is
acting in a very sophisticated manner. They look at
what the numbers say and they look at "how would that
help my state or district most?" And that
is the best number.
So, is it a question of lack of numbers or is it a
question of the way Congress operates? And I do not
know how to change Congress.
MR. BOLLING: Well, I do not either. I, as you
know, have spent a lot of time trying to change it.
CHAIRMAN LEVITAN: You have done all right.
MR. BOLLING: With some success and some failure.
But I think that is right; I think that is valid up to
a point, but I do know--and perhaps I am picking out
too many individual cases. The political process oper-
ates like every other one. Part of it is trust.
Now, some of the people that I trust feel very
strongly that some of the local figures are not
accurate rather than not favorable. Now I have to tell
you that I have not had the opportunity to go behind
their complaints in detail, and I cannot pretend that I
get to validate everything that I think by a complete
But my impression is that there are some people
who feel, that I trust--and that does not necessarily
mean that they are just Democrats--feel that they are
not getting a fair shake out of the figures because the
figures are not precise enough.
Now, I am not in a position to argue with you
about that, Mr. Chairman, because you know a great deal
more about this particular thing than I do.
CHAIRMAN LEVITAN: But, Mr. Boiling, is it realis-
tic for Congress to ask BLS to determine unemployment
in 6,000 small areas? BLS or Census cannot prevent
Congress from requiring the data. Again the fault
comes back to you in Congress.
MR. BOLLING: But don't forget what I said at the
beginning that I almost included a very serious ques-
tion of whether that type of trigger mechanism made
sense, in my opening statement. And I have a very
serious question about that.
MR. POPKIN: Is that within our scope?
MR. BOLLING: Well, I am not sure it is.
MR. POPKIN: No, I asked because it is something
that I have started to think about when we look at the
data. I am asking my boss here, Sar.
CHAIRMAN LEVITAN: Yes, it is.
MR. POPKIN: Are we to be considering whether or
not it makes sense to even talk about unemployment rate
by Congressional district or by small areas when you
have so much commuting?
CHAIRMAN LEVITAN: It is within our scope. I will
give you the chapter and verse later.
MR. POPKIN: No, I did not mean to be critical.
This is a very important issue, and I did not know ---
MR. BOLLING: I understand what I am raising too,
and I am doing it on purpose. I understand I am asking
you to add to the difficulty of your task and you may
decide not to because it really is our rabbit. But
this is an opportunity to get some people to look at
it, who may look at it in a slightly different way than
CHAIRMAN LEVITAN: I can promise you, Mr. Bolling,
that we are not going to avoid the difficult problems,
which was your first sentence, but I can also promise
you that we will not know the answers to some of the
MR. BOLLING: Well, of course. I am not asking
for perfection, for heaven's sake.
MR. CARLSON: May I just pick up on that.
Politically to trigger the funds into certain areas,
can you have the trigger so general that it would be on
major substate regions as opposed to 6,000 communities,
so you may be talking about 250 or 300 breakdowns, or
maybe just by state?
MR. BOLLING: All right, let me appear not to
answer that question but to answer it. I have been in
Congress quite a while and under a bunch of adminis-
trations and I have found that the rules and regula-
tions could be administered in ways that tended to be
more objective rather than less objective. And they
could be administered in the reverse, regardless of
what they were.
And usually there was some device used to make it
possible to appear to be fair. That leads me to a
negative, which comes out positive, if I may sound very
strange. That leads me to believe that in the end, it
is the quality of administration, the way in which
matters are administered, that determines how these
programs work in terms of their ultimate purpose which
is to reach certain people, certain kinds of people.
And if that conclusion is correct, if the manipu-
lation of the political process in our system is too
great, then I find myself coming back to the notion
that it is better to use the nonmechanistic device of
good administration rather than the pseudo-objective
MR. CARLSON: But that puts a lot of burden on the
Civil Service to ---
MR. BOLLING: Of course, it does.
MR. CARLSON: --- and they will be criticized by
those that do not happen to like the direction it
MR. BOLLING: That is correct.
MR. CARLSON: --- as opposed to the mechanism
being criticized or the Congress being criticized for
having set up the allocation according to a mechanism.
Well, the Congress is almost
unbelievably capable of evading criticism in individual
terms. The institutions are criticized and regularly,
but the members generally succeed in keeping themselves
popular. So, again, through a negative I arrive at a
positive and go for good administration.
That is one of the reasons why I support something
that has nothing to do with this, that has to do with
Civil Service reform, so that we can have better
But what I am saying is: I do not think you can
escape the administration.
MR. CARLSON: Don't you really shift the need for
data away from the Congress to the Civil Service who
must then have the data?
MR. BOLLING: Yes.
MR. CARLSON: But the data is still needed.
MR. BOLLING: That is deliberate. I think we are
more skillful at evading the meaning of data than most.
CHAIRMAN LEVITAN: But aren't you also very skill-
ful in not providing the wherewithal to collect the
numbers? Congress passed laws which appropriated
billions of dollars without worrying about data. It
may be a good slogan, billions for the unemployed and
not a penny for the bureaucrats. By doing that, aren't
you giving the bureaucrats or the officials the power
to make policy because by allocating the funds on very
flimsy data, they make the policy rather than Congress?
MR. BOLLING: Well, I would like to be a little
bit more honest about it, a little more accurate about
it so people understood what was going on instead of
our pretending to do something that I agree with you we
are not really doing.
That is why in the beginning I questioned the
How does one carry that message
to the other 534 members?
MR. BOLLING: About the same way I have been
trying to carry other messages, just by working at it.
And I am not suggesting that--I want to be very careful
about this without appearing to suggest that you should
not assume all the burdens that make reasonable sense--
but I am suggesting that this be the prime burden of
your message. I am just trying to say that I think it
is something that has to be somewhere in any considera-
tion of the problem.
MR. CARLSON: But then in a sense you are really
saying the data needs, at least in the great detail
that people have suggested, fall on the Civil Service
under your ideal arrangement as opposed to the Congress
itself, but it is still needed.
MR. BOLLING: Right--in the sense you are perfect-
ly accurate under my ideal arrangement, but that is not
what we have. I am trying to present alternatives as I
see them and I think they are real alternatives, I
happen to think, and I am not sure that this would be
very popular in Congress, that it would be better if we
worked harder at getting good administration and con-
cerned ourselves less with pretending that we were
precisely allocating funds. Maybe that is too cruel a
way to speak of my institution.
MR. CARLSON: Well, if you have a problem on the
next level below you and say that it is a Presidential
appointee who is there for an average of 22 months and
you cannot expect a long-sighted view--generally a
short-sighted view--so you are really pushing it down
further into the Civil Service.
MR. BOLLING: We are going to have an argument on
that. I do not agree with that. I believe a person
can come in at the political level with a minimum of
professional background and using competent assistance
be a very competent administrator.
I do not intend to use names, but I think I can
cite people who are far from expert in their field in
both administrations, in the recent administrations,
who have come in and have been enormously effective.
And I think I can cite people who were perhaps even
more successful in their previous careers who came in
and had all of the facilities that the other people did
and who were not worth a damn.
So, I am not prepared to forgive the political
MR. CARLSON: From our standpoint, if I were
designing a data system for decisionmakers, I think it
would be different for the civil servant than the
Presidential appointee or the Congress, and that is why
who the decisionmaker is becomes somewhat important as
to what the data system should be. That is why I am
asking the question.
MR. BOLLING: Well, I do not really--I think we
have to have some common data base that is made under-
standable to all, but I am not going to argue that.
CHAIRMAN LEVITAN: Mr. Moskow, do you want to join
to defend the political appointees?
MR. MOSKOW: Certainly not.
CHAIRMAN LEVITAN: Well, thank you very, very
much, Mr. Bolling, and I hope you come back to advise
us frequently and soon.
MR. BOLLING: I intend to participate.
CHAIRMAN LEVITAN: Thank you very much.
MR. BOLLING: Thank you, sir.
CHAIRMAN LEVITAN: Our next witnesses are three
gentlemen from the American Enterprise Institute who
have been on the other side of the street, in the
Executive Branch, quite recently, but not today. Dr.
Fellner, Dr. Kosters, and Dr. Stein.
You have a joint statement that you have prepared
for the Commission. I will let you decide in which
order you want to proceed and how you want to present
it in your statement.
STATEMENT OF WILLIAM FELLNER, MARVIN KOSTERS,
AND HERBERT STEIN, THE AMERICAN ENTERPRISE
INSTITUTE FOR PUBLIC POLICY RESEARCH
Mr. Chairman and Members of the Commission, we
welcome this opportunity to appear before the Commis-
sion and present our views. We would like to note that
the views expressed in this statement are our own and
do not necessarily reflect the views of the staff,
advisory panels, officers or trustees of the American
Enterprise Institute with which we are affiliated.
We are pleased that a National Commission on
Employment and Unemployment Statistics has been estab-
lished, and we believe that periodic examination of the
procedures, concepts and methodology involved in con-
structing our employment, labor force, and unemployment
statistics can play a constructive role.
We are also very much in accord with a process you
are following of commissioning careful analyses and
conducting a thorough examination of all of the issues
surrounding these statistics. The unemployment data
receive a great deal of public attention, and advice
and suggestions on approaches to improving them should
be based on a comprehensive assessment of the issues.
In our statement, we would like to summarize
briefly a number of points that we believe the Commis-
sion should keep in mind in its deliberations. In some
of these points, principles that we consider to be
essential to maintaining reliable and comprehensive
measures of labor market conditions are outlined. In
others, approaches that we think merit exploration are
The concepts of employment, unemployment and labor
force status are the central elements in the labor
market statistics gathered through the Current Popula-
tion Survey. The main focus of the analyses and of
approaches to improving the statistics should be on
sharpening these distinctions and clarifying interpre-
tations of these categories rather than on signifi-
cantly or fundamentally altering the basis for col-
lecting and presenting these data.
There are several reasons for this proposition.
First, as we have noted, these data receive wide public
attention and the concepts that have been in use for
most of the last 40 years are generally understood by
the public as representing labor force and job status.
Second, it is important to maintain a basic con-
tinuity in these data and concepts, both from the point
of view of public understanding and from the point of
view of analytic uses of these data.
Third, an attempt to transform the current unem-
ployment statistics into a welfare-oriented index would
be ill-advised, because such an index could not be used
to adequately reflect either labor market conditions or
the overall income and well-being of citizens and
We should adhere to the common sense notion that a
person who has a job should not be called "unemployed."
It would be a serious mistake, in our view, to move
toward attempting to encompass measures of welfare and
job status in a single index.
At the same time, we believe that an effort should
be made to develop income and welfare measures that can
be related to the statistics on labor force and employ-
ment status and that can supplement these data and
facilitate better interpretation.
For example, information on what, if any, other
income is received by families in which a member is
unemployed would be extremely useful. Both the amount
and the sources of other income would be valuable
information, such as, for example, whether the family
member unemployed is receiving unemployment compensa-
Information on household income would also be
useful for assessing the welfare implications for
certain groups of workers, such as teenagers, of rela-
tively low-wage jobs that may provide valuable work
experience and opportunities for advancement.
Another area in which opportunities for supple-
meeting current information to improve our ability to
interpret labor force, employment, and unemployment
patterns should be explored is the general area of
prior work experience.
Additional supplementary information, (1) for both
the employed and the unemployed on the length of the
period of current or prior employment, and (2) for the
employed whether job changes occurred without an inter-
vening period of unemployment, could improve our under-
standing of changes in jobs and employment status,
duration of unemployment, and the timing of periodic
labor market entry on the part of intermittent workers.
Additional information on wage or salary rates
would be valuable both for the employed and unemployed.
For the unemployed, data (1) on reservation wages, (2)
on possible rejected job offers along with the wage
that these offers entailed, and (3) on wages during
prior periods of employment would be valuable for the
interpretation of labor market conditions and job
We recognize, of course, that wages are only one
element in conditions of employment. Our emphasis on
exploring possibilities for enrichment of the data
along these lines flows from the proposition that while
the distinction between being employed and not being
employed is reasonably clearcut and susceptible to
operational definition, the distinction between being
unemployed and not in the labor force is more
That is, unemployment is a small fraction of those
not working, and information that would be helpful in
sharpening the distinction between workers in these
categories could be valuable in interpreting labor
market conditions. Improving our understanding of the
relationship between unemployment data and wages
associated with job acceptance could contribute impor-
tantly to our ability to interpret the unemployment
statistics in terms of resources available for utiliza-
tion at prevailing market rates.
Experimental work to explore the promise of
supplementing the statistics in the ways that we have
outlined would not involve significant expansion of the
resources currently devoted to this statistical pro-
Since other approaches to expanding the statis-
tical program may be recommended and considered by the
Commission, such as increasing the sample size to
obtain additional demographic or area detail, we wish
to note that we are not recommending a major expansion
of this statistical program.
If tradeoffs need to be made, we recommend that
the Commission consider the approach of less frequent
surveys in any particular geographic area or demo-
graphic group, while expanding the intensity of the
surveys in terms of supplementary information gathered.
In view of (1) the interest that is frequently
shown in data for specific geographic areas, (2) the
contribution that more detailed information could make
to supplementing our understanding of unemployment data
for allocation of federal programs and funding support,
we recommend that the commission investigate the possi-
bility of supplementing our data on unemployment by
making more use of data collected by the United States
At the present time, these data consist mainly of
numbers of workers receiving unemployment compensation
or filing new claims. Collection of more detailed
information on industry of last employment, wages, and
demographic characteristics, along with information on
other registrants at the Employment Service, could
vastly expand our information on unemployment collected
from a different and independent source. These data
have the merit of representing an entire universe
instead of being sample based, they are subject to
confirmation by employers, and they could help to
satisfy demands for more local area detail.
In conclusion, we want to express the hope that
the Commission will develop a specific and candid
report concerning the significance and interpretation
of our employment, unemployment, and labor force
statistics. The report of the Commission can con-
tribute to better public understanding of the meaning
and limitations of these data and point to constructive
directions in which they might be improved to further
enhance their usefulness.
Thank you for this opportunity to appear before
you and express our views. We would be happy to
elaborate further on any of the points we have outlined
or to discuss other issues that you might wish to
CHAIRMAN LEVITAN: If I might, I had the privilege
of reading your statement last night, and a few ques-
tions came to my mind as I read it. On page two of your
DR. FELLNER: Page which?
CHAIRMAN LEVITAN: On page two--and you, Dr. Stein,
repeated that point and emphasized it--you say that you
do not want any fundamental altering of the basis for
collecting and presenting these data. Does that mean
that you do not want any changes in the definitions of
the current usage that we have adopted some 40 years
DR. STEIN: Well, of course, if you say that we do
not want any, that is an extreme position. But I think
that we do want to keep to the meaning of these terms
as they are generally understood. I would not say that
going from an age cutoff of 16 to 17, in my mind, is a
radical change in the meaning of these terms.
But to say that we are going to include as unem-
ployed people who are working would be a radical
change. So, we would like to keep the concepts stable
and if there are new things that we want to measure, we
should recognize that they are new things and give them
their own names.
I guess what I am saying is that we should try to
call things by their right names.
CHAIRMAN LEVITAN: Well, we will get to the hard-
ship index a little later, if that is what you are
referring to. My inquiry referred to the several
reasons you offer for not altering the collection and
presentation of data.
The first one, we know what the figures mean.
Let's be specific. Would you want to consider--you do
not have to give a definitive answer right now--16 and
17 year olds? Ninety percent of them are in school and
might not be considered as part of the labor force if
they look for only a few hours of work.
Would you possibly consider that this is a change
in our economy, in the way people behave, and therefore
the same way as the BLS excluded in 1967 the counting
of 14 and 15 year olds? As you know, in earlier decen-
nial censuses they counted 10 year olds as gainfully
Is it possibly time now to change to 16 and 17?
DR. FELLNER: I would like to have the informa-
tion, Mr. Chairman, which you are describing here hypo-
thetically or realistically. I would not like to see
the unemployment concept changed for that reason. I
would like to have the information. I would like to
have many other pieces of information along with the
Change is very disturbing--major changes. I think
that even the 1967 and 1970 changes in the question-
naire are disturbing if you try to engage in any kind
CHAIRMAN LEVITAN: All right, Dr. Fellner, let's
go then to the concept of continuity. We count the
military outside the labor force. In 1973, we changed
to a voluntary armed force. Now the armed forces are
competing with private employers for the same young-
sters. Do you think that continuity is necessary
there, or does continuity possibly not reflect reality
any .more and, therefore, there is room for a change?
DR. FELLNER: That, of course, is a very well-put
question, Mr. Chairman, but there I think continuity
has been destroyed by events. So, it is arbitrary how
you want to pick it up, on which side of the adjustment
you take care of it. So, I have no strong convictions
about that. Looking to the future, if we believe that
military service is likely to stay voluntary, let's
adjust it to what the situation now is because there
would be less trouble with it if you do that than if
you adjust it in the other direction. This depends on
a guess concerning the future.
So, there is no good answer to that question,
CHAIRMAN LEVITAN: Well--and I do not want to take
up too much time on that--we can go through many
similar examples. If it is a system that was developed
in the great depression of the 1930s, and now you have
an entirely different society and economy, isn't it
possibly time for an overhauling of the system by which
we count the current labor force?
And to that extent, wouldn't it possibly be
desirable to change the classification of a lot of
marginal cases? I do not know how total aggregate
employment and unemployment would change. Given these
shifting factors, should continuity be a prime con-
sideration? We are now dealing with an old system, and
some of the concepts and definitions may have actually
outlived their usefulness.
DR. KOSTERS: I think we should also point out
that we say in our statement that we favor analyses and
approaches to improving the statistics that would
sharpen the distinctions and clarify their interpreta-
tion and so on. That is to say, it is not a statement
that we are ruling all sorts of marginal changes in our
You are very much aware of the discussion that has
gone on from time to time about under what conditions a
person should be regarded as unemployed, whether he
works one hour or five, or whatever, the conditions for
answering the question in such a way that he is placed
in the category of being unemployed--and there are a
variety of such conditions. In some cases job search is
a requirement and in other cases it is not.
The main point that we are making is that we do
not favor basic and fundamental changes in the notion
of what being unemployed or employed or in the labor
force means, but, on the borders, on the fringes of
these categories there is always room for making judg-
ments, probably for sharpening the distinctions that we
Now, in the case, for example, of the treatment of
the armed forces, there is in a sense a discontinuity
that would be introduced by a difference in treatment.
However, all of the basic data are there to reconstruct
the series in whatever way you wish, and this is a
different kind of discontinuity than restructuring the
series in some way that would change very much the way
we think of a person as being employed or unemployed.
CHAIRMAN LEVITAN: We will come back to concepts
later. I see some of my colleagues want to raise some
MR. POPKIN: Since the presentation was so abso-
lutely clear and made so much sense on almost every
single point and the one or two places where it was
fuzzy, I assume that was on purpose, and I would like
to ask: both in the paper and just now, you sort of
implied that we should not count persons with a job as
I had no idea that anybody had ever suggested
that, and I wonder what you mean. Where has it been
suggested that we count people with jobs as unemployed?
My boss here has spent years saying that some
people with jobs have hardship, but I do not know any-
body who has ever said that somebody with a job is
unemployed, and I am not sure what you are alluding to.
DR. STEIN: Well, I am very glad to hear it. I
think that is all we intend to say. If nobody wants to
do it, well, let's not do it.
MS. WILLS: Can you give more information on what
you do not want changed, aside from not counting people
who have jobs as unemployed? This concern about conti-
nuity, I hear about it and I read about it a great
deal. I am constantly reminded of how I do not want to
change over to the metric system myself because that is
disturbing to me.
But aside from economic analysis, which I am not
saying is unimportant, I frankly think we need a little
bit more clarification about why the validity of the
concept of continuity is so important, what it is you
do not want to see changed.
I have not, as a member of this Commission, heard
anybody talk about, for example, the hardship index as
being anything but a hardship index, which is not what
we are talking about in terms of an unemployment
DR. STEIN: Well, I think one aspect of the con-
cept of unemployment which I would not want to see
changed is that a person who is to be counted as unem-
ployed should be, not only not working, but he should
be seeking work. I would not want to include as unem-
ployed a person who is not seeking work.
Now, this issue comes up in relation to the treat-
ment of the so-called "discouraged worker."
MR. POPKIN: Are there any others?
DR. STEIN: Well, those are the ones that occur to
DR. KOSTERS: Well, there are other matters that
are closely related to that. There is, for example, a
section in the report on duration of unemployment and
there, I think, is a fuzzy area, in a sense, of the
conditions that were supposed to have been met by those
people who were unemployed, say, for 13 weeks or 26
That is, it is not at all clear that if they were
interviewed in each of those months during the duration
that they would have appeared as unemployed. They may
or may not. It is an area where there is considerable
vagueness. So that if one were to consider a measure
of unemployment that weighted by duration, it seems to
me one also ought to look very carefully toward
improving our information on what that duration
32-931 0 78 3
MR. POPKIN: I am not sure I understand what you
DR. KOSTERS: Well, I mean by that, for example,
for some workers who are counted as unemployed, they
need to have shown that they have tested the market in
some sense. They have looked for a job. For others,
that is not necessary, such as for people who are
waiting to go to a job within 30 days and so on.
People are asked also in the survey how long they
have been looking for work or a question to that
effect. Now, what is not clear in many instances is
whether during each of those four-week periods, what-
ever the criteria for current unemployment status are,
that they fulfilled that job-search, market-test kind
of criterion. They might have been ill. They might
have been out of the labor force or they might not.
I am not really suggesting that the statistics are
biased in one way or another. I am just saying there
is a good deal of imprecision in what we know about
DR. FELLNER: To the best of our knowledge, the
question that is asked is: When have you started
looking for a job? And that leaves quite a bit of
leeway there as to what happened since his last employ-
So far as we can figure out there is no guarantee
that the statistics on reentrants is consistent with
the statistics on the duration of unemployment. When
they are testing whether a person is reentering, the
question that is being asked apparently is: Why are
you looking for a job? And the answer is either
because I left my job, I lost my job, or, on the other
hand, because ---
CHAIRMAN LEVITAN: You mean the reason for not
looking for a job?
CHAIRMAN LEVITAN: The reason for not looking for
DR. FELLNER: No, he is unemployed.
CHAIRMAN LEVITAN: Oh, he is unemployed, I see.
DR. FELLNER: He is unemployed, so he is looking
for a job. The question is: Is he an entrant or a
reentrant? Or was he in the labor force all the time
since he left or lost his last job? Now, it seems to
us that the question, whether he is an entrant or a
reentrant, on the one hand, or jobloser or leaver, on
the other, is at present decided by asking him: Why
are you now looking for a job?
Then, he may answer because I lost my job, because
I left my job, or because I have now decided to get
myself a job or get back into a job. And that, of
course, need not be consistent with the data on the
duration of his unemployment. He may, for example,
truthfully answer that he is now looking for a job
because he left or lost a job rather far back in the
past, but he nevertheless should be regarded as a
reentrant if he started looking for a job only a short
while ago. The answers concerning the reason for
looking for a job and the answers from which the dura-
tion of unemployment is inferred may be inconsistent
with one another if we understand the procedure cor-
rectly, and I think that needs to be looked into. And,
generally speaking, what needs to be looked into, I
think, is the continuity of jobseeking over the period,
since the time when he started looking for a job.
So, there are a number of things that could be
sharpened and I think they should be sharpened. That
does not mean that I am deviating from the general
MS. WILLS: I would like to know--on page four you
pick up three things, reservation wages, more informa-
tion on possible rejected job offers. Why is it you
chose those three issues as something we need more
information on and are opposed to other kinds of possi-
abilities? That is one question.
Secondly, you raised later in the paper that if we
got more information through the Employment Service
that we would have better coverage to expand our infor-
mation base. Two things come to my mind. Why hasn't
that been done before from your perspective and what
will that really buy us?
And my third question and final one--we do have to
deal with tradeoffs obviously, and I think you heard
part of the Congressman's concerns just earlier--we as
a Commission do have to make some recommendations to
Congress about what statistics are to be used for the
allocation of funds and/or recommend that no statistics
be used to allocate funds.
If you assume for a moment that we have to deal
with the political realities of Congressman Boiling at
all, they are going to need some facts. How is it you
think your recommendations on continuity, these kinds
of information that you are asking us, will deal with
that political reality that we have to advise Congress
DR. STEIN: Well, I would like to say something--
Dr. Fellner may also want to--on the question about the
wages or these questions on page four.
It seems to me that the biggest gap in what econo-
mists would like to know about the unemployed is in
this area of the conditions on which they are willing
to work. That is, the thing that we measure is I think
not what an economist would be interested in knowing,
and we do not measure what an economist would be
interested in knowing because it is very hard to
measure what an economist would be interested in
knowing so we measure what is out there easily obtain-
But I think you would be interested in knowing:
What is the number of people out there who are not
working and who would be willing to work at a wage
which does not exceed the value of their possible
product? I mean, that is an important economic con-
But we have no way of getting it.
But this kind
of question would give us some clue to that.
MR. POPKIN: On that very narrow point, I was very
curious on page four why you didn't suggest also that
we ask people the reasons for rejecting the jobs they
were offered other than perhaps the salary.
DR. STEIN: That would be a good question, yes.
MR. POPKIN: Is there a reason you did not mention
DR. STEIN: I did not think of it.
DR. FELLNER: It was mentioned in a way, I think.
I think the way it is mentioned is that we said that
terms of employment are not only a matter of the wage.
CHAIRMAN LEVITAN: Dr. Kosters?
DR. KOSTERS: Could I pursue that point on wages.
I do not want to repeat what Herbert said about that
basic component, the conditions under which people are
willing to accept employment. It does seem to me when
one looks back at reviews from time to time of employ-
ment statistics and issues surrounding them and so on,
there is a tendency for them to reflect to some extent
the kind of issues that are pertinent to the time when
these committees consider the unemployment statistics.
And it seems to me in recent years there has been
a good deal of work and a good deal of increase in our
knowledge about issues concerning the nature of job
search and what it means in terms of workings of the
labor market. There has been a good deal of work by
Martin Feldstein and others in this country and by some
people in Canada on the impact of such things as higher
unemployment compensation payments and wider avail-
ability of them on labor force behavior, for example.
This is relevant, of course, to welfare reform issues.
What kind of impact do these kinds of supple-
mentary income payments have on labor force behavior?
It is also increasingly relevant in a world in which
there is more likely to be more than one worker in a
family so that you do not always have one worker on
which the family income depends.
So, it seems to me that that raises to the fore-
ground more sharply at this time the issue of the con-
ditions under which people would be willing to take
In an inflationary period, this issue is also
particularly relevant. But it seems to me that changes
in social conditions, changes in our income support
mechanisms and so on make this a more pressing issue
than it previously was, and it would be useful, it
seems to me, if the Commission could address it and
present some basic information on the state of our
knowledge on these issues so that it would be the use-
ful reference work and place to turn to for information
on these matters that some of the previous reports have
CHAIRMAN LEVITAN: Don't you think, Dr. Kosters,
that the suggestion you make on page three for a hard-
ship index--and I am delighted that you joined the
club--might take care of that? The present measurement
of employment and unemployment does not take into con-
sideration family income, and the fact is that a
majority of families have more than two earners. Many
also receive transfer payments. Maybe an additional
measure is needed to reflect reality in the labor
market. I therefore welcome your excellent idea about
the hardship index to supplement present data and fill
a gap in these data.
Would you agree with that? Does that fill your
needs, Dr. Kosters?
DR. KOSTERS: Well, let me say that I do not have
any objection whatsoever to a hardship index. I just
think we should not transform what we used to regard as
an unemployment rate into one. They are very different
concepts and what we mean by continuity, at least what
I mean by continuity, is to try to avoid transforming
the unemployment rate into something else.
Now, one can think of a lot of useful information
one can get. One can get, of course, information on
discouraged workers. Maybe we could get comparable
information, as Herb Stein has suggested, on dis-
couraged employers who are on the other side of the
But I think that there is a lot of information
that would be very useful concerning family income or
other sources of income and so on, and we could develop
a number of indexes and we should have the imagination
to develop separate new names for them as well, rather
than calling them the unemployment rate.
DR. STEIN: As I understand it, I do not think
that the hardship index will serve the need of these
figures about reservation wages. That is, there may be
people who are unemployed and whose incomes, or family
income is very low, but that does not give us any clue
to the wage at which they would be willing to work.
They may have very high expectations or demands.
So, it seems to me this is a different thing.
CHAIRMAN LEVITAN: Well, it would partially ful-
fill that need. That it would not completely fulfill
it, that is certainly correct.
DR. STEIN: Well, okay. I do not think it would
fill it very well because I think that the idea that it
would fulfill it to a high degree implies some assump-
tions about what we were calling earlier the elasticity
of supply of labor, which may be incorrect. So, I
think it is a separate question that deserves atten-
Of course, from time to time, the BLS has done
some work on this, as you know, and Dr. Fellner has
worked on these questions, that if they have reported
the wages at which some sample of the unemployed would
be willing to work--of course, are seeking work--but
this sample is very incomplete and the survey is taken
very infrequently and it is very hard to interpret.
They ran into a problem of very
considerable slippage there. That is to say they had a
very high nonresponse ratio to begin with and then they
had further slippages on the specific questions. This
may have to do with something that is very hard to
overcome, but I think that one should put one's mind to
overcoming it, namely, the difficulty that that kind of
information needs to be gotten from the person in ques-
tion and not indirectly through a member of his house-
CHAIRMAN LEVITAN: Dr. Fellner, do you think that
we should try to get data on reservation wages from
special studies, longitudinal studies, or would you
want to make them part of the Current Population
Survey? The reason I am asking this question is that
you know the Census people and BLS people are very
leery of asking questions because they depend on volun-
And there are just so many personal questions that
you can ask of volunteers. You alluded just now that
the responses may not be as reliable or that many would
refuse to respond. Would you think then that this
should be part of CPS or should the information be
obtained through special surveys?
DR. FELLNER: Well, I do not think that I am
particularly competent to answer that question. I
think that is the kind of question which people who are
actually engaged in the work know much more about than
I would. I think it would make sense not to make it
part of the regular monthly survey, but to do it fre-
To the best of my knowledge, it was done once.
Now, that may be my limited knowledge of it, but to the
best of my knowledge it was done once and rather
recently and with meager results due to no lack of con-
scientiousness on the part of the people who were doing
Due to no fault of their own, it is a very incon-
clusive survey. They do have some data which they have
not published so far on what the latest earnings of the
unemployed were. This did not get into the special
report. They do have some data on that and those also
should be published. And they should try to find
methods for reducing the difficulties into which they
MR. MOSKOW: You started asking part of the ques-
tion I was going to ask. There are some practical
limits as to how much you can get on a monthly survey,
according to the Census Bureau and BLS, and I was just
wondering--I guess there are two questions: Is there
any information that we are now getting on a monthly
survey that you think is unnecessary in the future and
you would delete and substitute the suggestions you
have made? Second, the alternative would be since
people want more information on discouraged workers and
a lot of other subgroups that may require surveying the
individual and not just the respondent in the house-
hold, would you lean towards some type of a twice-a-y
ear or three-times-a-year special survey of a group of
people, on a regular basis, to get this very detailed
DR. FELLNER: I would imagine that that would be
the answer, yes, that they would get that kind of
information perhaps including the reservation wages and
MR. MOSKOW: Yes.
DR. FELLNER: Through surveys that are not inte-
grated with the monthly survey and one would perhaps
transfer some questions that are now in the monthly
survey into the other survey which would be undertaken
CHAIRMAN LEVITAN: Dr. Stein?
DR. STEIN: Well, I would be willing to give up
the monthly survey entirely if you can get away with
it. I do not know why we have to know this every
month. There are a lot of things we do not know every
month, and if it would be economical to do a quarterly
survey, if that would enable you to do more questions,
I think that would be a permissible tradeoff.
I think that the monthly variations are small; we
do not learn anything from them; they just keep people
in a constant state of agitation.
DR. FELLNER: As you point out, we do not have any
monthly GNP either.
MR. MOSKOW: That is right. Some people thought
they wanted it, but they got over that.
CHAIRMAN LEVITAN: Dr. Kosters?
DR. KOSTERS: Could I comment just a bit further
I think that whenever I look into these questions
I see things that it would be nice to know and get more
data on. There are all kinds of pressures for expan-
sion of the amount of information one gets and at least
in our statement we came out in favor of not trying to
expand the Current Population Survey to fulfill every
kind of data need for all kinds of local areas and so
on, and in favor of possibly periodic, more intensive
And that is also one of the reasons why we have
suggested exploring at least the possibility of using
some existing data more intensively, the employment
service data, as an alternative in a sense to expanding
the number of surveys, questions on the survey and the
sample size that would be involved.
CHAIRMAN LEVITAN: Dr. Kosters, would you be
willing then to get some kind of uniformity in the
state employment data at the cost of greater federal
intervention? Because, as you very well know, in some
states insured unemployment may represent 50 percent of
total unemployed, as counted by the CPS, and in some
states it may be 70 or 80 percent.
Since the variation is so very great, would you
then be willing to have federal intervention to get
more uniform data?
DR. KOSTERS: Well, there are always two approaches
to that sort of thing. One approach is to try to get
federal uniformity, and the other is to take systems as
they are and to adjust for differences in the way they
are treated. I really do not have any detailed recom-
mendation on that. We have only suggested it as some-
thing that might be explored, whether there could be
better utilization of those data.
CHAIRMAN LEVITAN: Dr. Stein?
DR. STEIN: Well, I will proceed from that. I
would like to say something about the question that was
asked about the regional or local data in relation to
what I heard Congressman Boiling say, and I have not
followed it, but I must say I was staggered to think
that there was an interest in having unemployment
information on 6,000 regions of the United States.
I cannot see any sense to an unemployment number
for something which is less than what you might call a
labor 'market area, and I think you should tell the
Congressmen that the unemployment figure is supposed to
measure a condition of the availability or nonavail-
ability of jobs. That cannot be defined in 6,000 parts
of the United States.
CHAIRMAN LEVITAN: Quoting you, Dr. Stein, we
shall do so.
MS. WILLS: Just for a point of information, we
have been told that they now recommend, as of two days
ago, that we expand that to 10,000.
DR. STEIN: Well, you could protest.
MR. POPKIN: I would like to hear a word from each
of you, if you have anything to say, on the need for
more longitudinal data on some of these questions, even
if it is very expensive to trace people's labor market
behavior over a 10-year period?
I think that is very important and
valuable. I think that the only way we are going to
get some light on the question of the value of the tens
of billions of dollars we are spending on employment
and training programs is to see how people perform
after they have been through these things, and that
means you have to follow them for a long time.
And I think an evaluation of the causes of so-
called chronic or structural unemployment always
implies things about conditions of peoples' lives over
a period of time, and I think that we need to know more
CHAIRMAN LEVITAN: Dr. Kosters?
DR. KOSTERS: I would like to comment on that
While I believe it would be nice to follow people
around over time to learn something more about matters
such as the contribution of employment training pro-
grams and so on, I do not believe that the Current
Population Survey is an appropriate mechanism for that.
I do not think we should get into that at all in the
Current Population Survey, frankly.
I notice that you have a little section on that in
your outline, and my comment on that would be that I
think it would be very desirable if recommendations are
made for approaches other than the Current Population
Survey, that they be based on a thorough review of
studies that have already been made in this area, so
that any approach that is followed will be solidly
based on evaluative approaches that have proven
successful and on all of the evaluations that have been
done so far.
Many have been done and there should be some
information on the relatively more successful
approaches to evaluation. There are some now underway,
as I understand it, in connection with the Census
Bureau, but I do not believe that the Current Popula-
tion Survey is an appropriate vehicle for providing
information useful for that purpose.
CHAIRMAN LEVITAN: Do you think that a longi-
tudinal study like the one at Ohio State University
would be the proper vehicle for it?
DR. KOSTERS: It may be useful. I would not
really want to comment on what the most desirable
design would be. It is only one of a really quite
large number of explorations by this time into that
kind of area, and I would very much like to see a care-
ful analysis of the usefulness of the approaches that
have been used and that should be recommended for
further exploration in that context.
DR. FELLNER: I certainly would agree with that.
I think one thing that should be done and perhaps was
done--I am just not aware of any appraisal or analysis--
is to explore to what extent the data we get from the
present longitudinal studies are consistent with what
we get from the cross-sections.
There would be this difficulty, I think, that to
explore consistency you really would have to talk even
in the cross-section surveys to the person to whom the
data relate and not just to a member of his household.
You would have to talk to the person in question
to get anything worthwhile there, and there may be some
lack of comparability for that reason between the
Current Population Survey and surveys tracing the
destinies of individuals. That needs to be explored.
I still think it will be important to do that and to
devote quite a bit of effort to working out what the
difficulties are and how they could be overcome, yes.
CHAIRMAN LEVITAN: Dr. Stein, I have a question
for you, and this is my final one, unless my colleagues
might have some more. I am a charter reader of the
excellent newsletter, The Economist, which you are
authoring. If I remember correctly, we talked about
this earlier this summer. The Economist carried a
piece in which you identified a 7 percent unemployment
level as representing full employment.
DR. STEIN: Right.
CHAIRMAN LEVITAN: Now that BLS reported that
unemployment has declined to 6 percent, would you say
that we are having super-full employment, or would you
want to modify that statement?
DR STEIN: Well, of course, you tempt me strongly.
CHAIRMAN LEVITAN: Yes, I did.
DR. STEIN: I raised the possibility last fall
when I wrote that. My article was called "Full Employ-
ment at Last?", and it appeared in the Wall Street
Journal. One of the main points I was making was that
we did not know what full employment was. I would
stick with that very strongly.
Of course, the fact that we are at 6 percent would
not at all disprove the contention that 7 percent was
full employment because, after all, when I came into
the government in 1969, the unemployment was 3.3 per-
cent. Even my more ambitious colleagues thought 4
percent was full employment then, so it is possible to
have the unemployment rate lower than the full employ-
What I understood to mean by full employment was a
rate that you could not get below without an accelera-
tion of inflation, and I did not think that an accel-
erating inflation was a continuously endurable condi-
All right, so we are below 7 percent; we have the
inflation accelerated. Now, you can draw the conclu-
sion you like from that. I do not think it disproves
that 7 percent was full employment, but I will not
insist on it.
CHAIRMAN LEVITAN: But would you want to recant
I do not want to recant anything I
said at that point. I mean, I did not say that 7 per-
cent was full employment. I do not want to withdraw
the possibility that it was.
CHAIRMAN LEVITAN: Could it also be possible, when
you were saying that unemployment was really not 7
percent, that the seasonal adjustments were incorrect?
DR. STEIN: Well, I think that became a problem in
the next few months. When I wrote that in August, the
unemployment rate had been 7 percent for some months
and I think even with the revised figures, it still had
been 7 percent for a few months in August, and the
problem, as I recall, part of the divergence between
the old and new seasonal began to appear later in the
year, so that when the unemployment rate in the old
bases remained at 7 percent through September, October,
and November, then, of course, I felt confirmed in my
view and the seasonal adjustment kind of raised a ques-
tion about that view.
But I understand that there are still questions
about seasonal adjustment, so maybe I was right. Maybe
you will revise the seasonal and find that it was 7
percent all year long.
CHAIRMAN LEVITAN: I see.
Thank you very, very much for the help and thank
you for changing your schedule, Dr. Stein, to be with
us. We will be looking forward to your advice and help
over the next 17 months left before we deliver the
Thank you, gentlemen. We will have a 10-minute
(A brief recess was taken.)
CHAIRMAN LEVITAN: Our next witness is Mr. Edwin
J. Coleman, Chief, Regional Economic Measurement Divi-
sion. Mr. Coleman, would you present your statement,
STATEMENT OF EDWIN J. COLEMAN, CHIEF,
REGIONAL ECONOMIC MEASUREMENT DIVISION
BUREAU OF ECONOMIC ANALYSIS, U.S. DEPARTMENT
MR. COLEMAN: Thank you. I have a brief state-
ment, less than 15 minutes, that will state the BEA
position for the commission.
One of the major programs of the Bureau of
Economic Analysis is the preparation of personal income
and employment estimates at the national, regional and
local area levels. The national estimates have long
been in use as indicators of economic well-being.
In recent years, however, there has been an
increasing demand for more reliable county and metro-
politan area estimates by both private and government
agencies. Personal income and its derivatives at the
local area level are used for current economic analysis
and for economic projections. They play a major role
in the Federal General Revenue Sharing Program and in
the allocation of other Federal grants-in-aid.
The county employment series, developed as a
supplement to the personal income estimates, represent
the only annual estimates of employment by county in
industrial detail, currently being maintained.
The Bureau of Economic Analysis, in its prepara-
tion of State and local area income and employment
estimates, conducts no surveys of its own, but rather
relies heavily on data contained in the administrative
records of both State and Federal programs. The
remainder of the data comes from the various censuses
and from nongovernmental sources.
Much of the data generated as a byproduct of the
State unemployment program of the Bureau of Labor
Statistics, the social insurance programs of the Social
Security Administration, and the Federal tax program of
the Treasury Department provide an important basis for
the development of economic and social statistics.
Although much of the reported statistical infor-
mation is not directly or wholly suited for income
measurement, adjustments can be made to convert this
material into useful measures reflecting national and
regional economic developments.
The major alternative to the administrative record
approach to income and employment measurement would be
for BEA to collect the necessary information in surveys
of income recipients. The survey approach would pro-
vide data directly suited for the measurement of per-
sonal income and its employment adjunct, eliminating
the necessity of adjusting for definitional and con-
ceptual differences among the various inputs.
The cost associated with this alternative, how-
ever, would be prohibitive because a very large sample
would be necessary to permit reliable local area esti-
On the other hand, the use of administrative
records for statistical purposes is, or can readily be
made to be, both reliable and economical. Nonetheless,
it has to be emphasized that there is considerable room
for improvement in the quality and access to data basic
to the personal income and employment estimates at both
the local area and national levels.
One of the areas of concern is the degree of
establishment reporting in the ES-202, which purported-
ly covers approximately 90 percent of private payrolls
and upon which the bulk of the county estimates of
wages and salaries, employer contributions to private
pension, health, and welfare funds, and personal con-
tributions to social insurance as well as employment
The BLS guidelines for reporting of information of
multi-establishment firms are such that it is possible
for employment and payrolls of branch establishments to
be combined with that of the primary unit into a single
This deficiency can be remedied if establishment
reporting were made mandatory and BLS given access to
the Census Standard Statistical Establishment List
(SSEL), the use of which by other agencies is currently
inhibited by Title 13.
Access to the SSEL could give a complete geo-
graphically and industrially coded establishment list
to code the ES-202 reports. The use of the file would
serve as an important first step in the difficult
reconciliation of Census and BLS statistics.
32-931 0 78 4
Another prime concern with respect to the ES-202
is that the employment and payroll tabulations BEA
obtains from the ES-202 files are viewed as secondary
statistical byproducts of the UI program. As such, the
processing and tabulating of the data have a low
priority on the work schedule of the State agencies,
whose primary commitment is related to the unemployment
insurance aspects--taxes paid by employers and claims
by the beneficiaries.
BEA has a working arrangement with each of the
State Employment Security agencies whereby it receives
directly from each agency tapes and/or tabulations of
quarterly wages and monthly employment by county and
However, while most States try to be cooperative,
very often there are delays in the transmission of
these data to BEA because of the workload involving the
unemployment insurance itself, particularly at times of
high unemployment. These delays impair the quality of
the county income and employment estimates and reduce
their relevance as a current measure.
Such delays are not restricted to the sub-State
data. They frequently occur at the State level where
the data are received by BEA directly from BLS. Here,
too, delays often occur because of low internal priori-
ties. Delays in transmission of the State data under-
mine the reliability not only of the State quarterly
and annual personal income series, but also the
national series, since BEA must make estimates for the
missing States using less reliable methods or data.
At the heart of the problem is the division of
responsibility between the technical and administrative
operations of the ES-202 program which presently
exists. In past years, efforts on the part of users to
obtain improvements in the timing and quality of the
industry and geocoding, as well as the processing, of
the ES-202 reports failed due to an inability to
clearly earmark funds directly for statistical improve-
BEA endorses the recommendation that the technical
and administrative responsibility be consolidated under
one roof and that this roof be part of the BLS house.
We further recommend that the earmarking of funds for
improvements in the processing and tabulating of
employment and payrolls, as well as the upgrading of
the priority level of this phase of the ES-202 program,
be carried through to the State agencies.
There are two other recommendations concerning the
ES-202 that BEA would like to see implemented. The
first is a test of the current level of compliance by
matching the ES-202 files with the IRS Form W-2 which
now supersedes the Form 941 previously used by the
Social Security Administration. In 1972, when the
extended coverage provisions arising out of the 1970
amendments to the Federal Unemployment Tax Act took
effect, there was evidence of considerable noncompli-
ance with the UI laws in prior years. To the best of
our knowledge, there has been no check since to deter-
mine whether the level of compliance has improved.
A second recommendation is the inclusion of tabu-
lations by "legal form of organization" at least every
year. This data classification is of particular impor-
tance for the national accounts.
The BLS 790 establishment series in its present
form is of limited use to BEA in its State quarterly
income series. It is used extensively at the national
level. BEA relies on the monthly data on employment,
hours, and earnings gathered in the 790 not only for
the monthly national estimates of the wage and salary
component of personal income, but also as an indicator
of economic activity in months when the data normally
used in the preparation of the GNP estimates are not
Therefore, BEA is concerned about the large revi-
sions that occur annually when the UI data are inserted
into the 790 series as a benchmark.
We are inclined to believe that the revisions are
only partly a result of the ES-202's lag in reflecting
changes in industrial classification. More than
likely, the cause is in the nature of the 790 itself:
that is, the sample is based on the voluntary coopera-
tion of employers sampled; it is used as a link-rela-
tive; and industries other than manufacturing are
apparently poorly covered in the sample.
Other contributing factors are, no doubt, the lack
of a suitable benchmark for average weekly hours and
average hourly earnings and the lack of 790 data on the
hours and earnings of nonproduction and supervisory
An expanded 790, as recommended in the Wolfbein
report, could be utilized at the subnational level in
The collection of earnings data for all employees
would contribute to the improvement of the State
quarterly estimates. At present, BEA must rely upon
BLS 790 employment data for the preparation of the wage
estimates for the nonmanufacturing sector.
If earnings data were collected on a monthly
basis, even at the industry division level, for all
sectors, the quality of the quarterly estimates would
be upgraded considerably.
The inclusion of monthly data for all Standard
Metropolitan Statistical Areas (SMSAs), based on an
increased sampling rate to ensure quality data for the
smaller SMSAs would facilitate the preparation of a
quarterly income series for SMSAs and non-SMSA portions
of States, for which there is an increasing demand, a
demand that BEA has been unable to accommodate because
of the lack of sufficiently reliable data upon which to
base such estimates.
The inclusion of data on hours worked for all
industries, rather than, as at present, for manufac-
turing only, would enhance the possibility of the
establishment of a full-time equivalent employment
series at the subnational level.
Another key element necessary for producing this
measure is the derivation of estimates of the number of
farm and nonfarm proprietors on a full-time equivalent
The recommendation that more emphasis be placed on
producing information on the self-employed is whole-
heartedly endorsed by BEA. Although there is a large
body of data available on self-employment, it is almost
exclusively at the national level.
With the exception of recently available data from
the Social Security Administration's self-employment
file, BEA has very little information on the number of
self-employed or their income by State and local area
except on a episodic basis. The Social Security data
base has contributed greatly to the improved reliability
of the estimates of income and number of nonfarm pro-
However, it is not compatible with the concept of
fulltime equivalency since it includes only those self-
employed filing Schedule SE with their IRS Form 1040.
Exempted from filing for Social Security self-employ-
ment tax are those whose net earnings from self-employ-
ment are less than $400 as well as those who have
already paid the maximum Social Security tax for wage
and salary jobs.
The confusion in defining "proprietorship" in
farming has also impeded BEA's attempts to prepare
State measures of full-time equivalent self-employed
because of a lack of a definition of "proprietor" that
is consistent for both farm and nonfarm. A serious
effort to resolve this confusion is most important.
An alternative to expanding the BLS 790 series,
and possibly more economical, would be to adapt several
existing BLS series that already cover, although
limited in scope, some of the information that has been
suggested be added to the 790. Specifically, by
broadening the base of the Survey of Employer Expendi-
tures for Employee Compensation and the Employment Cost
Index Survey, these programs could yield reliable infor=
nation at the State and SMSA level similar to the
national data they now provide.
Two promising alternatives to the proposal by
Wolfbein that BLS derive commutation data by zip code
analysis of employee addresses obtained through its
surveys are: the Current Work History Sample/IRS Link
Project and the new Social Security reporting system.
Either alternative would place less of a reporting
burden on the business community than the implementa-
tion of the zip code recommendation.
The first alternative refers to the current
Census-BEA joint project to construct a bridge between
the Social Security Current Work History Sample and IRS
administrative records. This "bridge" would be used to
update the Census benchmark computer flows data.
The new Social Security reporting system, which
went into effect at the beginning of the year, utilizes
the annual W-2 reports filed with IRS, eliminating the
quarterly form 941. The W-2 file will be processed by
the Social Security Administration and IRS for their
For the first time, information about all W-2s
will be in computer-readable form. This change in
administrative procedures presents an excellent oppor-
tunity for the development of a statistical information
system capable of producing data on employment, wages,
migration and commutation patterns on an annual basis
and at the same time provide a check on the ES-202
The ultimate success in developing such a system
depends upon interagency coordination and modification
of existing confidentiality restraints, such as Title
13 of the Census legislation and the Tax Reform Act of
1976, which may limit access by the Department of Com-
merce and other agencies to data on the W-2 records.
Establishment reporting should become mandatory
and the priority level of the statistical uses of the
record system should be upgraded so that full advantage
of the information potential is realized in all of the
BEA generally supports those recommendations made
in the Wolfbein report to the National Commission on
Employment and Unemployment Statistics that would bring
about much needed improvements in the 790 establishment
survey and in the underlying ES-202 data base.
It is difficult to exaggerate the importance of
this BLS/BES data system to policymakers and to the
statistical community on which the policymakers rely
They are, however, programs currently in exis-
tence, the Survey of Employer Expenditures for Employee
Compensation, the Employment Cost Index Survey, the
Standard Statistical Establishment List, and the newly
implemented Social Security reporting system, which, if
expanded or made more accessible, might provide a good
supplement to the data presently available from the
BLS/BES system. The latter two sources of information
offer the most promise.
Given the necessary budgetary and legislative
support, they represent alternatives to expanding
selected areas of the 790 and might, in the long run,
prove more economical as well as more acceptable to
In any case, the dramatic increase in the use of
the wage and employment data, as well as personal
income, in formulas used to distribute billions of
Federal dollars to States and local areas requires that
significant improvements be made in these various pro-
grams. That is the end of my prepared comments.
CHAIRMAN LEVITAN: Thank you, Mr. Coleman.
MR. CARLSON: May I just pick up on your use of
the Social Security data--I guess it is the IRS data
with the Social Security--saying that that has powerful
opportunities for producing data on employment, wages,
migration, etc., and you say that there is a restraint
now and a problem of confidentiality.
As I understand it, that restraint is by indi-
vidual file, not by your specifying a run that could be
made and the results provided in summary form. Am I
incorrect on that?
MR. COLEMAN: No, you are not.
Agencies like BEA have a mortal fear of obtaining
individual information, and we generally try to have it
summarized to a county or industry level before we
obtain it. But much of the editing and analysis has to
take place in small areas, small counties in which,
even when you summarize the data to a county level, you
are faced with a confidentiality issue quite frequently.
So, the agency that has prime responsibility for
retention of the file for confidentiality, with good
reason, says, "Well, we will keep the file and we will
do the processing." I think this is a real problem for
the statistical agencies who really would abide by the
rules and would have very little opportunity to abuse
the confidentiality issue.
But the initiating agency is very conservative and
fearful of releasing the data. There is a blockage
developing in the transmission of administrative record
information between statistical agencies, even at the
Frequently, when they do summarize it for you, you
lose much of your capability to analyze the data.
MR. CARLSON: I guess the trick there would be to
have the analysis done within the agency which has the
confidentiality responsibility. In some other areas we
have found that somebody from the statistical agency
could, in fact, become the employee of the other
agency, having the skill necessary for that, and
qualify under the Act to do the analysis for you and
give it in summary form for whatever publication pur-
poses are needed.
MR. COLEMAN: Yes, that is a viable alternative.
We--BEA--need Census agents to look at data for editing
purposes. Our concern with regional statistics is that
because of the small geographic areas, we run into con-
fidentiality problems that people at the national level
just never dream of.
MR. CARLSON: But of all the things that you have
said, I think that you have come to conclude that this
offers one of the most promising areas, probably at the
least cost, of getting more richness of data on a local
MR. COLEMAN: Well, I am a firm believer in using
existing administrative record systems rather than
surveying the business community to death because I
think the business community is already to the point
where the noncompliance problem is becoming very
severe. So, yes, I would say that--but I would say
only as a supplement to the ES-202 program which BEA
considers probably the prime source of its income
information at the local area level.
MR. MOSKOW: I have just a few questions.
First, just as a point of clarification, your
testimony is presented on behalf of the BEA. I assume
that does not soley reflect the views of the Commerce
MR. COLEMAN: No, it does not. It reflects the
view of the Bureau of Economic Analysis.
MR. MOSKOW: Okay. The second thing is: I am
pleased to see the concern that you expressed about the
added cost of collecting data, both tax dollars as well
as the cost of supplying the information. You have
made some recommendations here. Have you made any cost
estimates as to what the cost would be in terms of tax
dollars for providing additional funds for these pur-
poses or in terms of the time for people filling out
MR. COLEMAN: No, I have not.
It would be considerably less than initiating
MR. MOSKOW: Yes.
MR. COLEMAN: Because the mechanism is in place.
MR. MOSKOW: Right. Is that something that the
BEA could provide for the Commission?
MR. COLEMAN: I think the only people that could
provide that information would be BLS since they are
conducting the surveys and they know what the costs are
for what they are doing. They could easily estimate--
or not so easily estimate--the margin of cost.
MR. MOSKOW: In terms of the data that you are
presently analyzing, they are used, if I understand
correctly, for distribution of general revenue sharing
I have samples of the uses of the
personal income data which incorporate the wage data
that we are discussing, and we estimate that something
like $35 to $40 billion of Federal funds use personal
income somewhere in an allocation formula, not neces-
sarily as the primary weight, but as a major weight.
So, this issue of regional data that we are
talking about is truly a serious one for Federal opera-
tions and the quality of the data is truly a serious
MR. MOSKOW: Let me get the $30 billion and $17
MR. POPKIN: That is entirely separate from that,
MR. COLEMAN: Yes, $17 billion is the CETA money.
I would be glad to provide the Commission with a break-
down. I did not bring it with me, but I will be glad
to provide you with a breakdown of the agencies using
the data and how we arrived at the figure of $35 or $40
MR. MOSKOW: I think that would be very helpful.
Let me ask you another thing. You do not utilize any
MR. COLEMAN: We use employment; we do not use
MS. WILLS: What recommendations--you do have some
in here--do you need the Commission's endorsement,
blessing, change in laws, etc., to implement these
recommendations from the Federal perspective?
CHAIRMAN LEVITAN: I would like to warn you,
Dr. Coleman, BLS and Census are right in back of you.
MR. COLEMAN: I know.
CHAIRMAN LEVITAN: So, any answer you give is at
your own peril.
MR. COLEMAN: Well, you know, one of the problems
with the 202 is it is a Federal-State cooperative sys-
MS. WILLS: Yes.
MR. COLEMAN: And when we attempt to get some of
the States to improve their data, the States say, "As a
State agency, we would like to see you (the Federal
agency) improve the data. We are a State agency, we
cannot do anything." They then put on their Federal
hat when we approach them from the Federal standpoint
and say, "Well, as a Federal program, you need to do
something." They say, "Well, we are controlled by our
governor who appoints the Commissioner of Employment
So, in Federal-State cooperative data systems,
there are real cost advantages in using Federal-State
cooperative systems, but one of the disadvantages is
that it is difficult to legislate solutions because it
is a mix of two levels of government.
For example, we have tried over the years through
0MB to get money directly to the R&A chiefs of the
Bureaus of Employment Security for the improvement of
the quality of the data, and it has always failed.
MS. WILLS: So--I do not want to put words in your
mouth, but see if I am hearing you correctly. One of
the things that you are suggesting is that there be, by
administrative fiat, or preferably by legislative fiat,
some sort of clarification on the role of the Bureau of
Labor Statistics as it relates to the research system
inside the state employment security agencies.
Do you consider that a critical factor?
MR. COLEMAN: Yes, if the research is focused on
the quality and the content of the estimates. It seems
reasonable to assume that if the Federal Government is
going to spend $40 billion, distribute $40 billion,
that it should earmark a specific sum of money for the
quality of the estimates being used.
MS. WILLS: Are there any other recommendations in
here that would require Federal legislative changes?
MR. COLEMAN: Yes, I would believe you would have
to change the legislation to--I am not sure that BLS
can of its own accord change the establishment report
to include the information that we had asked for on the
legal form of organization or the proprietors. They
may be able to do that; I do not know.
CHAIRMAN LEVITAN: Mr. Coleman, from what we have
heard so far, this seems to be a very crucial point for
the Commission's consideration, and I do not think it
would be very feasible to pursue that right now until
we explore it more fully.
Are there any additional papers that you would
want to submit to the Commission so that the staff and
the members could study it? And, then, I would like to
invite you to come down to discuss this with the Com-
mission in greater detail because the possibilities
that we are suggesting are very, very important for the
work of the Commission in terms of quality, at the
State and local level.
And when you start asking the question of why it
has not been done before, we know those things --- and
that is a proper subject for discussion, I suppose, in
a public hearing.
MR. COLEMAN: Fine. I would be most willing to
give it some thought, see what additional data or
information I might have that would be useful to the
Commission, and I am at the Commission's disposal to
meet at your discretion.
CHAIRMAN LEVITAN: Thank you, Mr. Coleman.
MR. CARLSON: Let me just ask: Inasmuch as you
put quite a bit of hope, if not faith, in the new
Social Security reporting system ---
MR. COLEMAN: You are attributing more hope than I
MR. CARLSON: You don't even have faith?
I guess this point you mentioned, this change in
administrative procedures, presents an excellent oppor-
MR. COLEMAN: Yes.
MR. CARLSON: It would be nice to know what poten-
tial is there, since going the administrative route may
have less burdensome qualities for the people who fill
out the forms. Are you the key person who can help us
on that or should we be contacting somebody who is
actually managing the systems?
MR. COLEMAN: Yes, I am not the key person on
that, but I think I can put you in touch with the
person who is.
MR. CARLSON: I think that would be worthwhile.
MR. COLEMAN: I clarify my point about the lack of
faith or hope in that the problem with administrative
records is that frequently program agencies do not have
the time or the resources to edit the local area data,
and we found that in working with the Social Security
files that this was a critical problem, that much of
the current work history sample that we pulled turned
out to be far dirtier than we had anticipated.
MR. CARLSON: But the potential seems --
MR. COLEMAN: The potential is there, yes.
MR. CARLSON: --- and lower cost seems to be
greater here ---
MR. COLEMAN: Yes, it is.
MR. CARLSON: --- than any of the other alterna-
MR. COLEMAN: I think the administrative record
approach has great potential.
CHAIRMAN LEVITAN: Mr. Coleman, we appreciate your
help and the Commission will be in touch with you when
you come back from your trip. Thanks again for adjust-
ing your schedule, too. Thank you very much, Mr. Cole-
Our next scheduled witness is Ricardo Zazueta. I
do not see him in the audience. Therefore, we will
continue with Nancy Barrett. Could you change it for a
DR. BARRETT: Yes.
CHAIRMAN LEVITAN: Thank you, Nancy.
Dr. Barrett, thank you for the prepared statement.
Will you proceed in your own fashion, summarizing it or
reading your statement, whichever way you prefer.
DR. BARRETT: Fine. Thank you, Mr. Chairman.
STATEMENT OF NANCY SMITH BARRETT, PROFESSOR
OF ECONOMICS, THE AMERICAN UNIVERSITY, AND
DIRECTOR, RESEARCH ON WOMEN AND FAMILY
POLICY, THE URBAN INSTITUTE
Dear Mr. Chairman and members of the Commission.
It is a pleasure to be here today both to congratulate
you on your preliminary work and to wish you well on
your important mission, as well as to identify what in
my mind are important issues that the Commission should
consider in its deliberations.
One major policy issue that is of immediate con-
cern is assessment of the degree of slack that current-
ly exists in the economy. This debate has raised the
possibility that our current measure of potential GNP
perhaps overstates the capacity of the economy to pro-
The problem is, of course, that a prolonged period
of economic slack has coincided with many structural
changes in the economy. These changes include demo-
graphic changes in the labor force, higher and more
comprehensive unemployment insurance benefits, higher
natural resource costs, environmental and safety regu-
lations, and the like.
Despite what many would have us believe, we simply
do not know how the economy would behave if it were
operating at capacity today. For instance, in the
slack labor markets of the past decade, the avail-
ability of unemployment insurance may have added to
measured unemployment to the extent that jobless
workers remained in the labor force rather than
dropping out or refrained from accepting or remaining
in a job far below their skill level.
What is not at all clear is how strong this effect
would be in a tight labor market where plenty of good
jobs are available. There is no evidence that indi-
viduals would opt for unemployment insurance when they
can find an acceptable job.
Inflation is no longer a test of how close we are
to capacity. Inflation today is much more the legacy
of inflationary expectations that perpetuate a self-
fulfilling cycle of wage and price increases than a
sign of undue pressure on the economy's productive
The recent pickup in the inflation rate is pri-
marily the result of an acceleration of food prices and
the cost of imported goods.
Despite these uncertainties, the question of how
much excess capacity is now in the economy is critical
to determining the appropriate macroeconomic policy
stance. How can the President and members of Congress
reach a wise decision on whether and how much to cut
taxes this year if economists cannot agree on a measure
of potential GNP?
I have written a background paper for this Commis-
sion that suggests new data needs in this area and
tries to sort out some of the issues associated with
interpreting existing data.
The second point I would like to raise is that
many of the structural changes I alluded to earlier
necessitate a reevaluation of the way we think about
the labor market.
To take only one example, many of our concepts are
based on the assumption that the primary labor force is
comprised of adult males. In 1977, however, 54 percent
of the civilian labor force was comprised of women and
youth, under age 25, compared with 45 percent in 1965.
This shift in the demographic composition of the
labor force means that we need to bring different data
to bear on the labor force experience of individuals
and to develop new models for explaining their behavior
in the labor force. Given the way our society is
organized, nonmarket options remain more significant in
the labor market decisions of women and young people
than for men. Conventional variables like unemployment
and wage rates need to be supplemented with such
measures as family status, fertility, educational
opportunities, and the like, that were in the past
considered beyond the purview of labor market analysis.
Further, life cycle information is particularly
important in assessing the probability of labor force
participation for women, necessitating improved longi-
tudinal data sets. Data obtained from the recently-
developed National Longitudinal Surveys has been par-
ticularly useful for analyzing the labor force behavior
of women and teenagers, but they suffer from small
sample sizes and the fact that they are restricted to
limited age cohorts.
We need much better and more timely data on the
educational status of teenagers and to develop some
linkages between those educational data and data on
teenage employment and unemployment.
Inadequate information about the determinants of
the labor force behavior of persons other than adult
males has resulted in a number of problems that plague
policymakers. The remarkable strength of labor force
growth through the 1974-75 recession and the continued
growth in the past two years, despite high unemploy-
ment, has puzzled economists and policymakers who are
trying to achieve what seems to be an impossible target
of 4.5 percent unemployment.
The official labor force projections of the Bureau
of Labor Statistics are consistently below the mark,
and budget watchers are confounded by the ever-
increasing numbers of women and youth who are potential
participants in federal job programs under CETA.
Even now there is talk of restricting eligibility
for public service jobs to "primary earners," but data
inadequacies preclude any reliable estimate of how
effective such a limitation would be and what would be
its distributive implications.
Changes in the demographic profile of the unem-
ployed have also rendered the concept of joblessness
ambiguous. When one thinks of the labor force as com-
prised of adult men who are the primary breadwinners
with women at home doing housework and young people in
school, then the distinction between unemployment and
joblessness is not very significant.
But when individuals have clearly defined social
and economic roles that span market and nonmarket work,
as do women and young people, then the question: "When
is a jobless person in the labor market?" becomes much
more difficult to answer.
And the difficulty is compounded when unemployment
statistics are collected from a household survey rather
than administrative data since the responses are influ-
enced by a wider range of interpretations.
The availability of unemployment insurance adds
another dimension of confusion. Under our system we
count UI recipients as unemployed even if they are not
seriously looking for work while we omit many jobless
persons who desperately want work but have given up
what they view as a fruitless, discouraging search for
All this adds up to an overwhelming need for
better ways to assess the labor market status of the
jobless and better information on discouraged workers.
The changing demographic composition of the labor
force is frequently cited as a cause of high unemploy-
ment and sluggish productivity growth, due to the
relatively low earnings of women and young people.
Better data in this area could improve our under-
standing of the causes of structural unemployment and
of why women, especially, continue to earn relatively
Further clarification of the dimensions of job-
lessness in our society is crucial to our assessment of
the amount of slack in the economy as well as to our
evaluation of the economic hardship that is suffered
when jobs are not available to all who want them.
A final issue is that most of our labor force data
are obtained from households despite the fact that we
have a potential gold mine of information that could be
collected from establishments. The BLS establishment
survey covers over 150,000 firms reporting on over 40
percent of the labor force and provides current infor-
mation on wage and salary, employment, hours, earnings
and labor turnover by industry and geographic location.
Except for employment, however, none of the data
are provided by sex. Space does not permit me to
elaborate on all the many reasons why we should get
establishments to report by sex. Apart from the well-
known superior reliability of establishment reports of
income, relative to household reports, some data such
as turnover rates are only available from firms.
Another need is for data on part-time employment
by industry. Presently, the only source of data on
part-time workers is the Current Population Survey, and
the definitions currently used in the CPS are ambiguous
Despite the obvious benefits of all the things I
suggested, they also have their costs, both budgetary
and substantive. Substantive costs are incurred when a
break in a data series results in a loss of historical
continuity and makes it impossible to compare the
present with the past. Expert witnesses will come
before this Commission asking for many new data series
and many changes in the old concepts and definitions.
But the Commission must be aware of the need for con-
tinuity and for keeping within reasonable budget
Given the many demands that will be made, I would
like to express what I view to be the most important
outcomes of your deliberations.
First, there is an overriding need to achieve a
national consensus on the acceptability of the official
unemployment statistic, or some other measure, as a
criterion for federal funds distribution. Literally
billions of dollars in funds ride on the numbers pub-
lished by BLS and their reliability and suitability for
this purpose is increasingly called into question.
Achieving a consensus means dealing with all of
the issues I raised earlier in addition to more tech-
nical problems like seasonal adjustment, defining geo-
graphic boundaries, and the like. Small changes that
seem inconsequential taken in isolation but that cumu-
late into significant ones as well as reconsideration
of big-picture issues are both important aspects of
this Commission's deliberations.
We will undoubtedly never achieve a similar con-
sensus on macroeconomic policy goals. Some people will
opt for upside risk, that is, keeping unemployment as
low as possible, without generating a serious
acceleration of inflation, while others will prefer the
risk on the downside, keeping inflation at bay without
generating a serious recession.
Despite these differences, however, we should be
able to get much more agreement than we now have on our
measure of economic potential to serve as a benchmark
against which actual GNP is compared.
In the past, we have too often viewed unemployment
as a measure of economic hardship. But now, demo-
graphic changes have greatly altered the traditional
support system of families and a more comprehensive
system of transfer payments based on entitlement blurs
the distinction between the working poor and the non-
Professor Levitan has suggested a new measure of
hardship. Although I disagree with many of the cri-
teria in his particular index, at least it is a
beginning in what I hope will be an open national
debate on the subject.
Finally, I think it is extremely important to bear
in mind how critical is the choice of data concepts and
the timely availability of data for analyzing social
and economic behavior and in designing and implementing
public policy. Data availability not only limits the
hypotheses one is able to test and the questions one
can answer, but data concepts often suggest the ques-
tions themselves or at least influence the researcher
or policymaker in the way questions are formulated.
At no time is the need for testing new hypotheses
and breaking out of traditional mindsets more crucial
than in a period of rapid change. I submit that the
U.S. labor market is undergoing such change and that if
we cling to outdated statistical concepts we will also
cling to outdated ways of thinking about the world.
This can only lead to bad research and bad public
policy. I urge you, for this reason, to be forward-
looking in your deliberations over the very important
issues contained in your mandate.
CHAIRMAN LEVITAN: Thank you, Nancy.
MR. MOSKOW: I have a couple of questions. As you
know, at present we have the best system of collecting
employment and unemployment statistics in the world, it
is my understanding.
Of course, we always want to improve it. But I am
just wondering if you think that we are going to make
quantum leaps? If we do improve our system, are we
going to make quantum leaps forward in our understand-
ing of the labor market and also in our social policy
deliberations, or is it really just going to be sort of
marginal improvements in those two things? Because we
are talking about spending a lot of dollars here.
DR. BARRETT: Yes. I think that you raised two
issues. One of them has to do with improving our data
collection system and the other has to do with thinking
about collecting different kinds of data or using dif-
ferent measures for describing and evaluating the per-
formance of the labor market.
I think that you really were talking about the
second aspect--at least that is what I was talking
about. There are some questions we do not ask about
the unemployed because we assume that they have been
continuously in the labor force ever since school.
And, in fact, that is not the case for women, or
for many women. We have very little information on the
experience of workers prior to their entering the labor
force or prior to their becoming unemployed, and I
think several people brought that up today.
We do not have a longitudinal picture of the
causes and effects of unemployment. We do not want to
throw out the unemployment statistic, but to assess the
performance of the economy or the hardship imposed on
the economy by substandard economic performance. There
are other important variables.
I disagree very much with Mr. Stein and his view
that if a person has a job that we should not worry
about that person any more. In fact, he went on to
say, rather inconsistently, that it is important to
collect income statistics on the unemployed to find out
if they really are hard up.
The implication was that some of these unemployed
people are so well-to-do that unemployment is not a
hardship for them, and we ought to know about that.
Similarly, it seems to me that we ought to know
something about the income of people that are employed.
That is not to say that we ought to throw out the
concept of unemployment entirely, but it might not be
the most relevant concept for a lot of the questions
that we ask about people. The unemployment concept is
quite specific, of great importance to economists, but
with only limited relevance to hardship or poverty. We
should try to avoid confounding poverty, or income-
related, problems with unemployment.
MR. MOSKOW: Let me ask you one other question.
What are your criticisms of Professor Levitan's hard-
ship measure that you mentioned in the paper here?
DR. BARRETT: Well, I think that he takes--with
all due respect, Sar--he takes a rather short-term
snapshot view of hardship. For example, as I under-
stand his proposal, he would not want to include in his
measure unemployment of what he calls secondary earners
which are usually, in most cases, women, married women.
And the reason that I think that is short-sighted
is that for the large majority of women in low income
families who are disproportionately represented among
the unemployed, their marital status is "iffy" at best.
The fact of the matter is that most of these women end
up for some portion of their lives the sole support of
Something like 40 percent of all children born in
the United States today, including upper, lower, all
income strata, will at some point be supported solely
by their mothers. Forty percent of all children will
at some point be supported solely by their mothers.
Now, it seems to me that if we think about unem-
ployment of women as causing negative attitudes towards
work, as causing a disruption of their employment con-
tinuity, as resulting in less work experience than if
they had been employed, then I say that any social
index, any index of hardship that does not take into
account the fact that this woman may later on be having
to support children and be looking for a job and be
handicapped because she was unemployed, because she had
a discontinuous work experience, and because she
developed bad attitudes toward work, that somehow this
ought to be counted.
And if, further, you want to elaborate on Sar's
general line of thinking, you think that these women
should not be given priority in public jobs programs
because they are viewed as secondary earners, that
seems to me to be a very bad social policy. It
develops a mentality among these women that they should
not be expected to work. Consequently, when they do
find themselves supporting children, they have to rely
on welfare or some other form of income transfer.
CHAIRMAN LEVITAN: Professor Barrett, we will
argue some other time, but for the time being I will
just plead "not guilty" to most of the charges that you
MS. WILLS: Do you have other concerns on the
hardship index? Other factors?
DR. BARRETT: No. I think by and large it is a
good idea. I just think that there are going to be
several problems with it. The first is that you are
never going to please everybody. I mean, politically,
it is a very difficult thing. The unemployment concept
is a lot less susceptible to the kinds of political
pressures that you are going to get when you try to
define who it is that is being hurt by some government
But I just cannot think of any kind of a hardship
measure that deliberately excludes a sizeable propor-
tion of the population and says that they are not dis-
advantaged by unemployment.
CHAIRMAN LEVITAN: Dr. Barrett, if I may, I will
send you the design that we are working on--measure-
ments which I think will take care of some of your
objections--I am afraid not all of them. But if you
will advise us how to eliminate all those problems ---
DR. BARRETT: Look, Sar, the thing about the hard-
ship index is the same thing I was saying about unem-
ployment. If you want to have a criterion for distri-
bution of federal funds, if you want to have a cri-
terion for evaluating the performance of the economy,
there has to be consensus, it seems to me, on what is
And if you have substantial disagreement from a
major element of the population, most of them people
who are concerned about the question of poverty among
female-headed families, poverty among women in general,
as well as discrimination, they will not buy your
CHAIRMAN LEVITAN: That is why I am trying to get
MR. POPKIN: Would you be so kind as just to
follow up with a note. I have not read your paper. I
do not think I have your Commission paper here. I
would like to see a list of specific suggestions from
you--this is just a small thing--the changes and the
exact questions and procedures you would use in this
CPS and BLS survey, in addition to all your other
things which we will be getting into later because you
alluded to the problems you had with the questions, but
you did not go on and say, "Here is what I think we
And I think you should follow up later, if you
would, please, with the specific suggestions for ques-
CHAIRMAN LEVITAN: Mr. Carlson.
MR. CARLSON: I, too, came late to your statement,
but I notice you are concerned about potential GNP.
Are you really recommending that the Commission do
something about the difficulty of coming up with a
standard definition of potential GNP?
DR. BARRETT: Yes, in my background paper, I made
very specific recommendations concerning data needs in
that particular area. The major concerns have to do
with determining the benchmark unemployment rate, that
is, what ought to be used as the standard, full employ-
Another one concerns determining the rate of
potential productivity growth, which is very tricky at
the moment because of the higher energy costs and the
allegation that somehow energy substitution is respon-
sible for the slowdown in the potential rate of produc-
tivity growth. We need to assess the growth of the
capacity of the economy to produce for that reason.
Now, as I suppose you know, the responsibility for
the potential GNP series rests with the Council of
Economic Advisers. This, in my mind, is not a very
good idea because the Council of Economic Advisers also
resides in the White House and is responsible for
articulating the economic policies of the government,
of the presidential administration.
So, it has a political view of the world as well
as an economic view of the world. And I do not mean in
the least to question the integrity or competence of
the individual members of the Council of Economic
Advisers. I myself was on the staff of the Council of
Economic Advisers last year.
CHAIRMAN LEVITAN: I could not ask for anything
When the Ford Council of Economic
Advisers revised down the potential GNP series by a
substantial amount, which essentially reduced the
gap--the official gap--between the actual and the
potential performance of the economy, this was a matter
taken in the academic community with quite a bit of
alarm. Those responsible for performance should not at
the same time determine the benchmark.
I think there was justification for revising that
potential GNP series down at the time, but it should
have been done by somebody not in the government, or at
least not a body with an explicitly political mandate.
In any event, there are data questions as well.
We do not have very good data on what the effect of
energy substitution has been on productivity growth.
In my view, there has not been much substitution away
from energy use, so a lot of the argument is sort of a
long-run argument rather than a short-run argument.
However, we do not have very good information on
that and we need better data. The same thing is true
with the interpretation of the unemployment rate as a
measure of capacity. We need much better information
than we have now on unemployment of hours and on unem-
ployment of skills.
It seems to me that if you have a person that is
in a job that does not fully utilize his or her skills,
for whatever reason, then your economy is operating
below potential. Art Okun and Wayne Vroman, for
example, at Brookings, have done a number of very
excellent studies on the process of what they call
worker upgrading of the business cycle. They show that
workers in a recession get downgraded in terms of the
capacity of the job to utilize their skills, and we
really need to get some better measure of the excess
capacity resulting from that phenomenon.
If lower wages for women and other groups result
from discrimination, rather than a lack of skills on
their part, then actual output as we measure it is
again below potential. We need to be able to assess
all of these things and I have made specific sugges-
tions in my paper on potential GNP for how we could get
data on that kind of thing.
CHAIRMAN LEVITAN: Dr. Barrett, we thank you. I
am sorry that most of the members here did not have a
chance to see your full, complete paper. Some of the
questions would be answered. We are a little amiss in
distributing them, but we will do so soon.
I think that since Mr. Zazueta is not here--what
was scheduled next, Marc?
DR. ROSENBLUM: We were then scheduled for lunch
and our afternoon speakers have not yet arrived.
CHAIRMAN LEVITAN: And what time are they expected
DR. ROSENBLUM: About 1:30.
CHAIRMAN LEVITAN: Which means that we will take a
break now for taking in some victuals.
(Whereupon, at 12:00 p.m., the hearing was
CHAIRMAN LEVITAN: Dr. Ossofsky, welcome to the
hearings. I appreciate very much your being able to
adjust your schedule and make it for 1:30, but we had a
witness, Mr. Ricardo Zazueta, who was supposed to be
here earlier and he did not show up. So, thank you
very much for coming earlier.
Dr. Ossofsky, I understand you have a statement.
DR. OSSOFSKY: I do.
CHAIRMAN LEVITAN: You can either read it or
include it in the record. You have all the time you
need, up to 15 minutes.
STATEMENT OF JACK OSSOFSKY,
THE NATIONAL COUNCIL ON THE AGING
DR. OSSOFSKY: Thank you, Mr. Chairman. I will
follow the statement, although I will make some inser-
tions as I go along.
I appreciate the opportunity which you and the
members of the National Commission on Employment and
Unemployment Statistics have given us.
I am Jack Ossofsky, Executive Director of the
National Council on the Aging, NCOA, a private, non-
profit organization which provides leadership and
guidance in the development of services for older
persons in hundreds of communities throughout the
Since 1950, the Council has worked to improve the
lives of older Americans by eliminating the problems of
aging and opening up opportunities for older people.
We are especially committed to those opportunities such
as work which encourage independence and self-suffi-
Through our National Institute on Age, Work and
Retirement, NCOA has provided research, information,
technical assistance and consultation on middle-aged
and older workers to employers, government agencies and
universities for over a decade. Our quarterly journal,
"Aging and Work," formerly "Industrial Gerontology,I"
has created a body of knowledge on all subjects related
to middle-aged and older workers.
NCOA is very pleased to appear before this Commis-
sion to discuss the relationship between labor force
statistics and middle-aged and older workers. The
importance of the Commission's charge--to revise the
way in which information on our nation's employed and
unemployed is collected, analyzed and reported--cannot
These statistics are the basis for the distribu-
tion of funds and the evaluation of federal programs,
such as the Comprehensive Employment and Training Act,
with the potential of helping many middle-aged and
older people return to the active labor force and
maintain their economic independence.
Thus, I am sure you will understand our concern
that the current approach to data gathering and report-
ing may not reflect the actual employment status of
middle-aged and older workers.
Current statistics show that older workers experi-
ence a markedly low rate of unemployment, but our
experience tells us this is false. The Bureau of Labor
Statistics (BLS) reported that, in 1977, the average
unemployment rates for persons 45-54, 55-64 and 65 and
over were 4.0 percent, 3.9 percent, and 5.1 percent,
These rates are lower than the average annual
unemployment rate of 7 percent and especially lower
than the rates for those 16-19 and 20-24, which were
17.7 percent and 10.9 percent, respectively. If we
were to rely solely on these official government sta-
tistics, we might perceive an optimistic picture of the
older worker in the economy.
Yet, numerous studies published both by NCOA in
"Aging and Work" and elsewhere question the accuracy of
the official statistics. In a 1975 monograph by
Elizabeth Meier for NCOA, using the Louis Harris survey
data, 10 percent of those aged 55-64 considered them-
selves to be unemployed, that is, in the labor force
but currently unable to find a job.
The figure of 10 percent is considerably above the
official government figures for this group during the
same time period of May through July 1974 reported as
slightly over 2 percent for males and 3 percent for
Part of the divergence between the official and
survey rates is caused by different definitions of
unemployment. BLS' definition of unemployment includes
only those actively seeking work during a certain
period. NCOA has stressed a number of times that these
figures do not include thousands of "discouraged"
workers who give up on finding work, remain unemployed,
but are considered to be outside the labor force.
As a recent "Aging and Work" article states,
"Advocates of a broader definition of unemployment
believe that present figures--which categorize dis-
couraged workers as not-in-the-labor-force--understate
by a considerable extent the true unemployment rate."
The not-in-the-labor-force categorization may be par-
ticularly misleading with regard to older women who
have given up the search for employment.
Though the category "discouraged worker" covers
all ages, it is older workers who are most likely to be
hidden in this group. During the first quarter of
1978, for example, workers aged 55 and above consti-
tuted 14.7 percent of the civilian labor force and 15.2
percent of the unemployed, but 32.5 percent of the
total number who were classified as discouraged.
But if we view the discouraged worker as someone
who has stopped looking for employment because of job
market conditions, and include those with the belief
that age is a barrier to employment, the proportion is
36.3 percent, two and one-half times the proportion of
older workers in the labor force.
The statistics regarding duration of unemployment
also support the contention that older workers are
likely to become "discouraged workers" once unemployed.
The two are interrelated: the long duration of unem-
ployment for the older worker can lead directly to the
"discouraged" status. The average mean duration of
unemployment in 1977 increased with age, as shown
Omitting discouraged workers from the official
unemployment count merely hides the true rate of unem-
ployment and deceives the public into believing that
this group is not in need of assistance. It is essen-
tial that the official definition of unemployment be
revised to include the discouraged, whether discourage-
ment results from job market or personal reasons, so
that the many older workers now omitted are enumerated
as desiring work.
This new definition of unemployment could assist
officials charged with developing programs and policies
to assist older workers. As it now stands, published
data give the impression that joblessness is not a
serious problem among older workers.
This serious failing can lead, and has led, to an
inequitable distribution of federal employment and
training program resources and caused a serious detri-
ment to the vast numbers of older people who still need
Unemployed older workers are also often hidden in
the "Not in the Labor Force" category because of retire-
ment. Dr. A. J. Jaffee of Columbia University in an
unpublished paper states: "One of the reasons for the
fact that unemployment is not so extremely high among
older workers is that those who cannot find jobs quit
looking and simply retire; they are not then classified
It is important here to note that in a recent
study of supplemental federal unemployment insurance
benefits, most of the people who had used their maximum
benefits and subsequently dropped out of the labor
force were 45 years and over; 70 percent were age 45
years and over, and 45 percent were 55 and over.
It appears from this study that the older the
individual, the greater the chance that he/she will
have used all entitled employment insurance and still
be unemployed or out of the labor force once all bene-
fits have expired. This increases the pressures for
many older jobseekers to elect pension and Social
Security benefits prematurely, at reduced levels and,
thus, cease to be counted in the labor market.
Others have been involuntarily retired though they
are still able and willing to work. The Louis Harris
survey commissioned by NCOA revealed that "over a third
of those who are retired said that they did not retire
by choice, but were forced to." This is a very large
group of older people who are also listed as "not in
the labor force" under the current statistics.
Clearly, we need to distinguish between voluntary
and involuntary retirement and, then, between those who
were forced to retire but still are able and willing to
work and those who are not.
Yet, there are no questions in the monthly Current
Population Survey that seek to ascertain if retirement
was voluntary or involuntary. In fact, there is no
breakdown for even the specific category of "retired"
under the "Not in the Labor Force" category in the A-3
If these data are to be useful, a subcategory of
retired should be reported, a breakdown made of volun-
tary or involuntary retirement, and the numbers of
those who want to go back to work counted.
Other issues which must be considered include:
In general, many tables published by BLS lump all
workers 65 and over into one category. The passage of
the 1978 Age Discrimination in Employment Act Amend-
ments, PL 95-256, will raise the age of protection to
70 for most private employees and eliminate it entirely
for federal workers.
Therefore, it is imperative that more detailed
information on age subgroups within the 65 and over
category be made. Only one table even attempts to do
Employment and poverty status by age is tabulated
in unpublished tables. This information should also be
reported on a regular basis to provide guidance to
policymakers and administrators in programs such as
CETA or the older worker community service employment
program. For the same reason, it would be desirable to
report duration of unemployment data by income, as well
Table A-18 which highlights duration of unemploy-
ment by sex, age, race, and marital status should be
published in a form that includes more detailed age
-.* *L La.&.; -l Q ^ f ^
-M --'. :: ; -:a "It* "--
4- *. i' .* ~ : : t
K + l *1. 1 .; -_*. ;'' "
STI -- . .. .!. . 1
,.- : .*-* .- - L- k. i -: E -: -
ir- i *. :^ E; .': .i'- .*-: ^ *L jn .l. -
I Air I I.
CT I -
4 ... --
a --Ii..** -.
'- ,., 1 1 1>.' *
n '--. .tm-- --
n* -. i'.* --
ir n.I.1": -., .
_ '- *
.. ." +- + ":' :- .. .: "1 .. ; .
.. .. .. ., !1 a i ....I l +. ,- : I.+ ,
J.* I.T : ' > ^ *-:* 4I4.' *j -' -i ; _
' -.- ""- -"'.. r-I
" 5: .- r - -.. .. i c-'."- - -n."
(.1j- * '.*- 4 1j4
'V -- ; .. A ^ :t L -. - -L. r Pc- -*:
*." F- -
"" -.+ LL'--1.
:. ,; ''i-a". .:u L1 ..1 If .
-< ~ l; iv: .at > 2':.'. * : 1: '2. r:- i gel"'' flit -'* -; c -
.. .: .. ..F .. .,. . _,... : > r n : ',. '- -:
_- "+' --- L ^ .. i- ._ i*." +'1 ^j ..t
Iv'-_--' :+* _
_t-. _-I,*: 2P : .-'" -L h ,
I1 - .
.1- .. .. eva-
:-it *'-" L' st_.m: -I J[ "-n r;_t: C I-
L-it* a]uu:"" _L11: :-L _:..$tC :rit*:rt 'z*-
* *> -,..s-
*t*:~-i.*":- 7+- n
'-'Ir. '.1: _
be able to differentiate and relate to other data that
you have. We have no particular emphasis on one method
or another, but would like to integrate it with other
data that you are gathering now and to differentiate a
little bit more clearly within the existing questions
that are currently being asked.
We do not pretend to be experts in the statistical
gathering field here nor in the methodologies used.
What we do claim some expertise in is that what we are
finding in life is not being reflected in data. Yet,
the policies that are being developed to reflect what
is going on in the communities are based on the data.
Therefore, we need to make some changes. If you
want to find another more appropriate way to do it, we
are very open to those suggestions. It just seemed to
us more appropriate to build it into your current
MS. WILLS: Do you think that by adding the dis-
couraged worker as a part of the count of unemployment
that that will solve many of the problems that you are
DR. OSSOFSKY: If you look at the way CETA's funds
are being utilized in the communities, for example, you
will discover that while middle-aged and older workers
represent perhaps half of the work force in our
country, they are lucky if they are reflected in 4
percent of the client groups of that program.
There are many tasks that need to be taken to
overcome it and one of them is the gathering of data.
That happens to be the purview of this Commission. We
will take other tasks to other committees and commis-
sions at the appropriate time.
This is not an issue or a battle that we can
resolve by simply taking one simplified answer to it,
but we at least need the hard data and we do not think
we are getting it sufficiently here.
CHAIRMAN LEVITAN: Dr. Ossofsky, why do you assume
that the enumerators discriminate against older people,
although I find it difficult to define older people at
age 40. Why do you assume that the enumerators, are
more prone not to count or interview them in the way
they do everyone else?
If anything, I would think that older people would
be easier to reach. The undercounting problem is cer- t
tainly much lower at that age. Why do you think that
CPS is not doing a proper job in this area?
DR. OSSOFSKY: I am not sure I can give you a very
good answer to that. All I know is that when we com-
missioned Lou Harris to do the study of older people,
we set up an advisory committee that included special-
ists in social research and gerontology who knew some-
t'ling about the conditions of older people.
Lou Harris, normally in a national poll, will
interview 1,800 people. In meeting our requirements,
he had to interview over 4,000--close to 5,000--we
oversampled, particularly to reach older blacks and
minorities in order to be able to draw some valid con-
I can tell you too that some 10 years ago when I
directed a program for the Office of Economic Oppor-
tunity on behalf of the National Council, called
Project Find, what we underscored at that time and in
the years that followed is that you have to knock on
every door in order to find the older people, particu-
larly those who are most vulnerable, those who have
given up hope and those who are isolated from the life
of the community.
But while I am not suggesting that your enumera-
tors do not do a good job for the normal segments of
our society, you have to look very hard and very
thoroughly to find the older people, and it may very
well be that what is adequate in seeking out and docu-
menting the needs of the rest of the population is not
good enough, not detailed enough in locating, gathering
the data, and finding adequate responses on the part of
There is also the question of how questions are
raised with older people. We know that if anybody
knocks on the door and asks certain kinds of questions,
he may find it more difficult to get a valid answer
unless the questionnaire and the whole issue is couched
in terms that make it comfortable for the person to
Now, I would not suggest that this is true for all
older people. I do not want to stereotype all old
people, but it is true for enough that we need to take
special means to make sure we are getting accurate
The other part of the answer is that it is at such
variance with what we who work with older people are
finding in the data that we uncover.
CHAIRMAN LEVITAN: I do not understand. Are you
mentioning that older people are having problems, or is
it a question of counting? Because, for example,
Dr. Ossofsky, you mentioned very eloquently in your
testimony that older workers seem to have longer
periods of unemployment, something which you can docu-
Now, I would not be surprised at that, because
older people very frequently retire and can still
collect unemployment insurance in a number of states.
DR. OSSOFSKY: Sometimes.
CHAIRMAN LEVITAN: No, isn't it a fact?
DR. OSSOFSKY: I do not know how much of a fact it
is. I know of many instances in which some older
people do that.
CHAIRMAN LEVITAN: Well, I know that in a number
of larger states they do allow retirement and then
collection of unemployment insurance.
DR. OSSOFSKY: If it is a mandatory retirement;
not if it is voluntary retirement. That would, how-
ever, not deal with the great volume that we are find-
ing here of people who are not eligible for retirement
yet. If we were to single out the people who were 65
or even 62, you might be able to make a pretty good
case, at least in those states.
How do you deal with the person who is 55? How do
you deal with the person who is 63 or 61 and not eligi-
ble for Social Security? So, I think we are dealing
with a much broader issue here on how we deal with the
growing numbers of older people in our society and one
of the reasons that we need to begin getting some good,
hard data now is that the trend in our population is
such that we had better begin shaping our data pretty
accurately to prepare for changes in our population in
the coming years.
CHAIRMAN LEVITAN: I assure you, Dr. Ossofsky,
that the Commission will be very sympathetic with the
concerns you expressed. But isn't it also true that
the availability of private pensions (which I didn't
hear you mention in your testimony) might also have
something to do with the employment and unemployment
status of older people in the labor force?
DR. OSSOFSKY: Yes, it has something to do with
it. But, how do you account for the fact that 4 mil-
lion older people, many of whom are drawing pensions,
made it clear that they are drawing those pensions
because that is their source of income at the moment,
they are entitled to them, but they would rather go
back to work?
How do you account for the fact that we now have
close to 50,000 people in the Title IX program who are
willing to give up in many cases not only Medicaid
benefits, because we are dealing with the older poor,
but other sources of income in order to go back to
The fact of the matter is that for great numbers
of this generation of older people the opportunity to
work and the additional income is much more important
and much greater potentially an income than the dollars
they get in pensions. In private pensions, of course,
they are only covering a small fraction of today's
generation of older people.
CHAIRMAN LEVITAN: Again, Dr. Ossofsky, I am not
quarreling with you and with your concerns, but this
Commission's interest is in statistics.
CHAIRMAN LEVITAN: And if you mention 4 million
and then you say in the next breath 50,000, it would
seem to me that is a very small percentage of 4 mil-
DR. OSSOFSKY: You underscore precisely the need
for expansion of the Title IX program. There are 4
million people ---
CHAIRMAN LEVITAN: That is not ---
DR. OSSOFSKY: I understand, but that is where the
difference comes from. You are asking me: If there
are 4 million people who want to work, how come only
47,000 are employed? Because that is the total number
of jobs currently provided by that particular program.
CHAIRMAN LEVITAN: Dr. Ossofsky, before the clock
rings again, do you have something really specific--
except for the discouraged worker, which everybody
seems to be mentioning here. What else can be done
either through CPS or through any organization that
would improve the statistics you want about older
workers? And I hope you would stop defining them at
DR. OSSOFSKY: If the prejudice and conditions of
discrimination would stop, we would not have to define
them that way. The circumstance of defining them stems
from trying to deal with the problem, not from a search
for a definition.
I am not sure that I could very accurately define
what an older worker is because I think that all you
learn about chronological age by an individual is how
many candles to buy for his cake or her cake. That is
about all that tells you.
We think that we have to get much more firm and
complete data about the middle-aged and older worker
for the reasons that we have cited. We think that they
are not being adequately represented in the figures
that you have available. Public policy is based on
those figures and resources are made available based on
those figures, and our own experience in data seems to
show that we are not finding them in sufficient
quantities as they are being found by others outside
And what we are urging you to do is to find a more
appropriate way to do so, and then among those you
interview to differentiate between the ages of the
discouraged worker and the courses of retirement as
well. I think it would give us perhaps a fair picture
of unemployment in our country, but also a more accu-
CHAIRMAN LEVITAN: Dr. Ossofsky, I will ask the
question in another way. You have been in this busi-
ness for a number of years, if I remember correctly.
The present Secretary of Commerce was a member of your
group, as are many other distinguished citizens. You
have a lot of resources.
Do you think that your organization could tell us
better ways of counting your constituency, or do you
want to leave it to the rhetoric that you have supplied
DR. OSSOFSKY: I would be happy to have some of
the specialists who have worked with us in the develop-
ment of specific techniques, who are better equipped
than I am in the technology of statistical data, the
gathering of such data, and polling methodologies, to
meet with your Commission, either formally or infor-
mally, to share with you the methodology we have used
in other data which has given us altogether different
results than those used by the Bureau of Labor Statis-
CHAIRMAN LEVITAN: Would you provide that for the
record of the Commission?
DR. OSSOFSKY: I would be happy to supply you with
the names of such people for the record and to set up
any meetings that would be of value to the Commission
to achieve that goal. We welcome the opportunity to
work with you in that regard and welcome your inquiry.
CHAIRMAN LEVITAN: I would appreciate your doing
so. Thank you very much.
DR. OSSOFSKY: Thank you.
MR. ANDERSON: I have a question about the treat-
ment of discouraged workers. As you know, the defini-
tion of unemployment at the current time is intended to
permit us to identify people who are willing and are in
fact actively engaged in seeking work. The burden of
your testimony, Dr. Ossofsky, is that those persons who
say they want jobs but are not doing anything to seek
jobs should be added to those who do not have jobs and
are looking for them in order to have a more comprehen-
sive measure of unemployment.
That would seem to suggest that what economists
call the labor market test should be disregarded alto-
gether, and I wonder if you think that is the case or
whether you think there might possibly be some way to
inquire into when an individual most recently sought
work as a way of trying to get at the willingness to
Do you think that a person, for example, who has
sought work within the last six months might be counted
as unemployed, or the last year? Or what about a per-
son who has not sought work in the past two to three
years? Should that person still be counted as unem-
ployed in your view?
DR. OSSOFSKY: Mr. Anderson, the most practical
way I can answer that question has to do with the fact
that when we have gone out and provided the opportunity
for viable employment for middle-aged and older
workers, many who had not been able to get a job for 15
years and 10 years suddenly agreed to take a job, wel-
comed it, because they had given up hope of ever
getting a job again.
It is very hard, therefore, to draw a line in a
way that would respond specifically to your question.
It may very well be there needs to be some relationship
between age and past attempt to seek employment, but we
have not found any period of time as a clear line of
demarcation in our own work with older people over the
last number of years that would dispel the notion that
given a chance to take the job, the worker who really
wanted that job did not do so.
CHAIRMAN LEVITAN: Dr. Ossofsky, we will look for-
ward to you supplying us with the technical information
on how to better count older workers. I am sorry we
cannot do anything about providing jobs for them.
DR. OSSOFSKY: Well, at least you can start by
giving us some hard data to maybe make some changes in
CHAIRMAN LEVITAN: We will do our best. Thank you :'
DR. OSSOFSKY: Thank you very much.
CHAIRMAN LEVITAN: Our next witness is Professor
Edward Kalachek from Washington University in St.
Professor Kalachek has already done yeoman's work
for the Commission by preparing an excellent paper on
the subject which he is going to address right now.
Professor Kalachek, you have 12 to 15 minutes to sum-
marize whatever you want to say.
EDWARD KALACHEK, PROFESSOR OF ECONOMICS,
WASHINGTON UNIVERSITY :
DR. KALACHEK: Thank you very much and let me
thank the Commission for the opportunity of expressing
my views before them today.
Most of our employment and unemployment statistics
are cross-sectional and were devised as monitoring [i
devices. They were created to answer questions about
levels and incidence. Questions like: What is the
unemployment rate? How does it differ between whites a:
and blacks, between the young and the old, and between
factory and office workers?
Over time, these cross-section series have been
expanded and improved. As a result, they have measured
labor market performance with increasing detail and
precision, thus facilitating the identification of
problem groups and problem areas.
t This very success in identifying problems, to-
Sgether with continuing efforts at treatment by the
federal government, have accentuated the need for new
and different types of information. The key questions
have gradually changed from "what" and "how many" to
S"why" and "how" and "does it matter?"
e The construction of public policy has increasingly
required inferences on the stability of populations, on
the impact of processes, and on behavior responses.
For example, how many of those who are poor or
n unemployed today are likely to be poor or unemployed
tomorrow? What is the impact of persistent unemploy-
Sent or nonlabor force status as a teenager on employ-
ment stability and wages as an adult? What is the
impact of training on subsequent employment and wages?
ow will labor supply change if wages or nonemployment
incomes are altered?
The traditional cross-section series have been
Used in ingenious and invaluable ways to seek answers
to such questions. However, they were created to meas-
k tire, not to explain. As tools for explaining labor
Snarket behavior and outcomes, they have often proven
inefficient or inadequate.
As a result, government agencies, beginning in the
nid-1960s, have sponsored a number of special purpose
longitudinal series specifically designed to facilitate
the investigation of change and the uncovering of
These longitudinal series or panels involve
repeated interviews over time with the same sample of
people. They permit observation of the experiments
continuously being conducted in the real world where
ages, income, training, job opportunities, health, and
ge are continuously altering.
Our experience with this panel data has been quite
limited; tapes from the National Longitudinal Survey
md from the Income Dynamics panel containing three
successive observations on the same individuals have
)een publicly available only for the last five years.
!he research and policy analysis done during this brief
)eriod is impressive in both amount and significance,
i however, particularly given the problems of early users.
The promise shown by this analysis is sufficiently
great that it can support the following assertion: if
significant advances are to be made during the next
decade in our knowledge of how the labor market
operates and of how it responds to policy initiatives,
that knowledge will most likely come from work utiliz-
ing longitudinal data.
The special purpose panels have already demon-
strated their high priority place in the system of
employment and unemployment statistics. Their creation
has moved the portfolio of statistics in the right
direction, but has not moved it far enough.
The special purpose panels have a small sample
size and they reinterview infrequently. They conse-
quently are not optimal for fully investigating prob-
lems which during any time period affect only a small
portion of the population or which persist for only
modest time intervals; problems like job turnover or'
Happily the instrument for analyzing these prob-
lems is readily at hand. It is the Current Population
Survey (CPS). The CPS is clearly the most promising
frontier for the further extension of longitudinal
While CPS is a cross-section survey, its sampling
format felicitiously creates subpanels of 16 months'
duration. CPS thus collects longitudinal data as a
byproduct of an ongoing cross-section survey. All that
is required to open this seeming treasure house for
exploitation is the quite modest effort of regularly
matching observations across survey periods creating
what are called matched or gross flows tapes.
Since the duration of the CPS panel is short and
the length of the CPS questionnaire is limited, the
gross flows tapes will not be substitutes for the
special purpose panels. They will be complements. A
large sample size and eight interviews within 16 months
makes CPS the ideal instrument for investigating short-
run labor force dynamics.
Given the strong and continuing interest in the
causes and repercussions of unemployment, it is ironic
that the most powerful vehicle for analyzing the unem-
ployment experience has long been readily at hand, but
has been virtually unexploited.
Let me enumerate just a very few of the questions
CPS panels can be used to investigate: What is the
impact of the seasonality of industry on labor force
participation and unemployment? Where do seasonal
industries obtain workers in the boom season? How much
unemployment is due to industry-specific cycles? How
much unemployment is due to tight production management
closing down or laying off workers for brief periods
when production seems to be outstripping sales?
How many workers who enter unemployment through
layoff return to their original employer? What dif-
ferentiates workers who return to the same employer
form those who find jobs elsewhere? What types of
workers, by age, sex, race, education, do expanding
Does it vary depending on the size of the expan-
sion and the stage of the business cycle? Such topics
are only the tip of the iceberg. The recurring CPS
interview schedules can periodically be supplemented by
special questions. This capability for supplementation
is perhaps the most attractive aspect of CPS viewed as
It means that the CPS gross flows tapes can be
used in a timely and flexible fashion to investigate
questions of immediate policy concern. As an example,
let me cite the work currently being done by matching
replies to the Job Search Supplement of May 1976 with
interview data from subsequent months.
As you are aware, the responsible government agen-
cies have long been reluctant to make gross flows data
publicly available. Limitations and biases in the data
have been cited as the reason for this reluctance.
However, all statistical series inevitably have
limitations and biases. Work already accomplished at
BLS and elsewhere make it quite clear that the biases
in the gross flows data can be readily circumvented by
Developing controls and corrections for these
biases is not an unsurmountable task. Indeed, it is
not even a particularly difficult task. The task would
have been accomplished long since if the motivation
were present. I would strongly urge the Commission to
recommend a high priority for the regular preparation
of matched CPS tapes for public use.
CPS has evolved without any thought being given to
longitudinal analysis. If it begins to be used as a
panel, it should be possible to make a number of small
changes which will greatly augment its potential.
One such change, also useful for cross-sectional
analysis, would be to collect information on hourly
earnings three or four times a year rather than once.
Finally, if gross flows tapes are generated, the
absence of adequate information experienced at work and
on the employer would be the major remaining gaps in
our portfolio of longitudinal data. Panels like cross-
sectional surveys have concentrated on the individual
and on the household as isolated units, as sole sources
of information, and as information sources primarily on
demographics, family background, labor market status,
wages, and earnings.
However, understanding labor market actions and
outcomes requires information on the work site as well
as on the worker. Our knowledge of wage determination
of the labor supply decision, of the retirement deci
sion, of job turnover, and of the determinants of unem-
ployment would be greatly enriched if we knew mor
about the characteristics of the work place and of th(
NLS and others have obtained some pertinent infor-
mation by directly asking respondents work-relate
However, the typical respondent may know little
about the size of the establishment or firm, its turn-
over rates, production processes, personnel policies,
or comparative wage levels. Even where he is quite
knowledgeable, as in the case of working conditions,
responses could be better interpreted if we had a
number of separate judgments.
The closing of these data gaps will not be a
simple and straightforward matter like the generation
of gross flows tapes. Rather, it will require ambiti-
ous and innovative planning. What is needed is a two-
stage sampling format in which establishments are the
first stage and workers the second stage sampling unit.
Information could then be obtained from the
worker, from his employer, and from his union and
matched onto one record. A two-stage probability
sample is not unprecedented.
The National Longitudinal Survey of the high
school class of 1972 sampled first schools and then
students chosen from those schools. This procedure
enabled it to match the student's school record with
information obtained directly from the student by
CHAIRMAN LEVITAN: Thank you, sir.
MR. MOSKOW: Professor Kalachek, I am not sure if
you were here this morning, but I believe it was Marvin
Kosters who discussed longitudinal surveys a bit, and I
believe--if I am correctly quoting him--his view was
that he believed in longitudinal surveys and expanding
them, but he thought before any steps were taken that
we should have a thorough review of all of the existing
longitudinal surveys, the Ohio State University data,
and others that have been constructed, and the work
that has been done on these.
Do you share that view now, that we should have a
thorough review of all of these before we make any
recommendations, such as the ones that you have men-
tioned in your paper here?
DR. KALACHEK: Let me differentiate my response in
the following fashion. When I am talking about creating
matched tapes using CPS data, what in effect is true is
that the information already exists. The surveys have
already taken place.
What I personally would regard and have long
regarded as one of the great treasure houses of poten-
tial information on short-run labor force dynamics is
lying there virtually unexploited. Given that for 20
years economists have been recommending the exploita-
tion of this data and given my own strong belief that
nowhere in the whole area of statistics would the rate
of return on expenditure be higher, I would very
strongly urge the Commission to recommend that the
responsible government agencies proceed immediately
with the process of matching.
Now, I am not out of sympathy with Professor
Kosters' desire for a thorough review of existing
panels as a means of providing guidance for future
developments. Over the past decade we have created a
number of panels. Only two have been operative long
enough and generated enough data to permit a signifi-
cant quantity of analytic work.
Given these two panels and others which are now
becoming more available', there is very little question
in my mind that knowledge of what is going on in the
labor market and policy relevant research is increas-
ingly going to be based on panel data.
This creates a strong case for a continuing exami-
nation of the merits and limitations of panel data and
for coordination between panels. I tempered my lan-
guage carefully on this issue in my recommendations to
the Commission. The thought of recommending a small
government section or committee to review and coordi-
nate panels simply bothers me.
I believe, on the basis of long personal experi-
ence--though it is a matter I cannot document--that if
we had created a central statistical agency back in the
mid-1960s to coordinate the development of panels, we
would not have experienced the rich and productive
flowering panels which has occurred.
Nonetheless, I think that it would be a highly
advisable thing while minimizing bureaucracy to take a
really serious look at the nature or biases which arise
in panels, at the differences in results between
panels, and at comparisons between panel and CPS
I think that far too little methodological work
has been done in this area. It would now be useful to
have someone responsible for analyzing potential biases
in panels, and for coordination between panels. I
would depart from Professor Kosters mainly in believing
that review coordination and contemporaneous action are
all feasble and desirable. Thus, I see no reason to
delay the matching of CPS tapes, or the recommending
and creation of a two-stage employer-employee survey.
MR. MOSKOW: If I could return just to the
matching of the CPS for just a minute, I assume this is
something you would like to see done on a regular
MR. MOSKOW: Like a quarterly basis or twice a
ear or something of that sort, once a year?
DR. KALACHEK: You should understand that once the
mechanism is established, matching this will not be a
terribly expensive or complicated operation. I would
like to see continuous matching.
Each rotation group should be matched throughout--
first month, second month, third month, fourth month,
and then picked back up the 13th, 14th, 15th and 16th
This would be an enormous source of potential
information on labor force dynamics. If we matched
ith some of the special supplements, the information
base becomes even larger.
My own feeling is that complete matching across
rotation groups is the desirable outcome. It will give
us the most experience with this data and the ability
to learn most quickly. Clearly, I would be quite con-
tent with some lesser degree of matching. Something is
always better than nothing.
My impression, however, is that the difference in
cost is not enormous and the difference in benefits
CHAIRMAN LEVITAN: What is your concept of enor-
DR. KALACHEK: Once you establish the program, it
is a mechanical matter of matching observations and
creating files, which would be quite inexpensive.
CHAIRMAN LEVITAN: In other words, what you are
saying is that it is a question of hundreds of thous-
ands and not millions or multimillions of dollars.
DR. KALACHEK: It is cheaper than that. To create
and publish gross flows tables, a highly desirable
objective, you will need a research program for obtain-
ing forced consistency between the gross and the net
flows, for deriving hypotheses as to these relation-
ships, and for correcting rotation group bias.
This would involve a large one-time expenditure
for research of hundreds of thousands of dollars. The
matching of CPS tapes is a far simpler matter. I do
not pose as an expert on computer costs, but costs here
on an annual basis are way below $100,000; possibly
below $10,000. Matching CPS tapes will be one of the
great statistical bargains.
CHAIRMAN LEVITAN: Ms. Wills?
MS. WILLS: See if I am correct in what I think I
am learning. What you are suggesting does not require
any new legislation. It does not require redefinition
of the statistics. It does not require the development
of a hardship index.
In other words, we do not need to create something
new. Now, if I am correct on that, I am curious to
know, quite frankly, what kind of power you think this
Commission would have to recommend back to the Census
and to BLS that this kind of matching take place, if
indeed we are talking about a small amount of money?
And I think under $100,000 in this case is a small
amount of money.
I do not know enough, quite frankly, to give bacd
arguments to the bureaucracy that they do not want tc
make that information available. I am not saying that
I do not think this is not important. I do not knou
what we think that the Commission can do to highlight
Do we want to recommend it? Because it seems tc
me that if you have $100,000 laying around, you could
do it tomorrow. It is obviously not that simple, but
we do not need to change laws or concepts. Is that
DR. KALACHEK. That is right.
There are numerous economists, technicians, and
other trained personnel within the Bureau of Labor
Statistics who would be highly enthusiastic about the
recommendations which I have made and who would sort of
concur fully with them.
This Commission surely has an immense amount of
moral prestige in this area. A strong recommendation
from this Commission would greatly stimulate work on
Your predecessor, the Gordon Committee, made
recommendations on gross flows tables. The Department
of Labor and the Bureau of the Census accepted the
recommendations and said "We will proceed to do this."
Instead, over time they proceeded to enunciate a series
of technical defects requiring research and correction
as a prerequisite to making the data publicly avail-
The serious importance attached to these defects
boggles the mind. Don't misunderstand me, the defects
are there. They are quite troublesome if one is seek-
ing to create cross-tab tables. Matched tapes are
another matter. They will be used primarily to explain
and predict the behavior of individuals over time. For
example, the special Job Search Supplement indicated
differences in the techniques and intensity with which
unemployed workers searched for jobs. Matching would
allow us to follow these workers for a number of months
and to determine the impact of search time and tech-
nique on the length of unemployment.
In dealing with an analytic problem of this sort
an econometrician could control for gross flows biases
in such a fashion as to emerge wth reasonable results.
In summary, given the usefulness of matching and
the full awareness of this usefulness by many analysts
in the Department of Labor, a strong recommendation by
the Commission could have a quite beneficial effect.
MS. WILLS: You also mentioned--and I am not sure
that I caught what you were saying--that if we had a
simple agency, a statistical agency, that you think
that we would have or would not have a movement forward
in this area?
DR. KALACHEK: In response to Dr. Moskow's ques-
tion, I expressed my opinion that the development of
panels since the mid-1960s has shown verve and imagi-
nation. My suspicions are that if we had a central
bureau directing things back then, we would not be as
far along as we are today.
32-931 0 78 7
Having said that, I am going to have to agree that
there is need for more coordination between panels and
for an organized research program on the characteris-
tics of panel data.
CHAIRMAN LEVITAN: Do you think that longitudinal
studies, national surveys, should be taken over by BLS
or should they be kept independent?
Yes, no, or maybe?
DR. KALACHEK: Are we talking about the NLS?
CHAIRMAN LEVITAN: Well, I just wondered if it
pertains most directly to our work?
DR. KALACHEK: I would say the following: I would
see, at this stage of the game, no reason whatsoever
for changing the sponsorship or direction of NLS.
CHAIRMAN LEVITAN: Thank you very much. I hope we
will have a chance to continue talking about it in the
DR. KALACHEK: Thank you.
CHAIRMAN LEVITAN: Thank you for the excellent
paper for the Commission.
Professor Motley, welcome to the Commission.
Thank you very much for coming to testify before us.
BRIAN MOTLEY, PROFESSOR OF ECONOMICS,
UNIVERSITY OF KENTUCKY
DR. MOTLEY: Thank you for giving me this oppor-
tunity of appearing before you this afternoon to dis-
cuss the role of the unemployment rate as an indicator
in federal-state fiscal relations.
In recent years a number of federal programs have
been enacted under which federal grants to states and
localities are allocated on the basis of local rates of
unemployment. The Comprehensive Employment and Train-
ing Act (1974) and the Local Public Works Act (1976)
are prominent examples.
Presumably, Congress selected this criterion
because it assumed (i) that the unemployment rate is an
accurate indicator of local economic conditions and of
the need for federal assistance, (ii) that the concept
of "unemployment" is capable of precise definition and
that this definition is well-known and widely-accepted,
and (iii) that a statistical methodology exists for
consistently measuring the incidence of unemployment at
the local level. However, since the enactment of this
legislation, each of these assumptions has been
challenged and the use of the unemployment rate as a
guide for regional economic policy has come under
increasing criticism. The critics have argued that
(i) the concept of unemployment is less precise than is
commonly believed and the definition currently in use
may be inappropriate to present day conditions,
(ii) even if the present definition is accepted, the
measurement of the local unemployment rate is subject
to a substantial margin of error, and (iii) the local
unemployment rate is an inefficient indicator both of
the level of economic activity and of the need of
states and localities for federal assistance.
This paper seeks to examine these issues.
Section I discusses the concept of unemployment while
Section II briefly looks at the measurement methods
currently in use. Section III considers the appro-
priateness of using unemployment as a guide to the
allocation of federal grants. Section IV concludes.
I. Unemployment as an Economic Concept
To an economist, "unemployment" implies dis-
equilibrium in the labor market. When households wish
to supply more manhours of labor than employers are
willing to hire at the current wage rate, the excess
represents unemployed labor. For policy purposes,
however, this theoretical construct must be translated
into a measurable statistic. Thus, according to the
present Census definition, a person is unemployed if he
is not working, is able and willing to work, and either
is actively searching for a job, is on temporary layoff
or has accepted but not yet begun a new job. This
definition seeks to measure the extent to which the
amount of work being offered by employers falls short
of the amount households are willing to supply. How-
ever, for a variety of reasons, it fails to capture
precisely the theoretical concept.
Part-time work is one such reason. The theoret-
ical definition is in terms of "manhours," whereas the
empirical measure is in terms of "individuals." Thus a
person who is unemployed but seeking only part-time
work should be treated as partly unemployed and partly
not in the labor force, whilst one who is employed
part-time but would prefer to work full-time should be
treated as partly employed and partly unemployed. The
present definitions make no such distinctions: all
individuals are classified as employed, unemployed or
not in the labor force. Moreover, although data on
part-time employment and job-seeking are available at
the national level, no such information is collected
for states and local areas.
A second problem associated with the concept of
unemployment arises from the existence of "discouraged
workers." These are persons who are not employed, but
are not seeking work because they believe that no jobs
are available rather than because they do not want a
jot. This seems to imply that these persons would take
a job if one were offered and thus would be classified
as unemployed according to the economist's definition.
Hence, it is frequently argued that discouraged workers
should be added to the unemployment total in computing
the jobless rate.
On the other hand, the definition of discouraged
workers makes no mention of the wage rate they are
demanding. In some cases, the reason they were unable
to find a job may be because they were requesting a
wage which was higher than the market was willing to
pay. Moreover, many such workers probably do not fit
the usual picture of a discouraged worker as a person
who is suffering considerable hardship as a result of
joblessness, because the reason they have withdrawn
from the labor force is that their families have ade-
quate resources and hence they can be more "choosey"
about the type of work they will take and the wage they
will accept. It is of interest to note that although
more men than women remain unemployed for long periods
of time, the number of women who report that they are
not seeking employment because they could not find a
job or believe no jobs are available is much greater