EDUCATIONAL AND OCCUPATIONAL CHARACTERISTICS OF MANPOWER: AN INTERNATIONAL COMPARISON
| Date | 01 March 1966 |
| Author | J. C. Saigal,P. R. G. Layard |
| Published date | 01 March 1966 |
| DOI | http://doi.org/10.1111/j.1467-8543.1966.tb00929.x |
EDUCATIONAL AND OCCUPATIONAL CHARACTERISTICS
OF
MANPOWER:
AN
INTERNATIONAL COMPARISON
P.
R.
G.
LAYARD and
J.
C.
SAIGAL*
INTERNATIONAL
comparisons are often suggested as one piece of evidence to
be used in forecasting the requirements for educated manpower. It is
argued, for example, that if a country’s output per worker rises, the compo-
sition of its skill requirements will change, and that some guidance on
the pattern of change can be obtained by comparing skill patterns in
countries with different levels of output per worker. This belief has, in
fact, formed the basis of a number of actual manpower f0recasts.l
Against this it has been argued that there are no
a
prior;
grounds for
expecting any unique pattern of skill requirements for
a
given output per
worker.2 A unique pattern would only be found if certain strong assump-
tions which are discussed in Part
I,
Section
3,
were true. But the question
is whether, in the empirical world in which these assumptions are re-
laxed, there are sufficiently systematic relationships to be of service in
manpower planning; and if
so
what the parameters of these relationships
are.
It has only recently become possible to investigate this question, now
that the results of the
1960
round of censuses have been published for
most countries. A comprehensive study of the available data is now being
undertaken in this Unit, and this paper presents some preliminary
results.
The theoretical basis of the study is discussed briefly in Part
I
of the
paper. First wedescribe the typical framework within which manpowerfore-
casting operates and set out a simplified model which lends justification to
the procedures involved. Then we point out the limitations of this model
and, in the light of these, consider how results obtained from the model
might be used. Finally we discuss the relative significance of occupational
*
Mr Layard is Deputy Director, Unit for Economic and Statistical Studies
on
Higher Educa-
tion, London School of Economics and Political Science; Dr Saigal was formerly Research
Officer in the Unit and is now Reader in Economics, University
of
Rajasthan. The authors are
very grateful to Miss
S.
Hunt, Miss
J.
A.
Pinney and Miss
S.
Teper for their skilful and energetic
assistance in the study; to Miss B. Woods for the computations, and to Mr
J.
R. Crosslcy for his
constant advice.
See, for example, Commonwealth of Puerto Rico Planning Board, Bureau of Economics and
Statistics, in co-operation with the United States Department of Labor, Bureau
of
Employment
Security,
Pwrfo
Rico’s
Manoower Needs and
Suppry
(
1957).
a
For
a full discussion
of
the assumptions that would need to be satisfied for a unique pattern
to exist see R.
G.
Hollister, ‘The Economics of Manpower Forecasting’,
International
Labour
Reuiew,
Vol. LXXXIX,
No.
4,
April 1964. See also E. R. Rado and A. R. Jolly, ‘The Demand
for
Manpower:
An
East African Case Study’,
The
Journal
of
Development Studies,
Vol.
One,
No.
3,
April 1965, for a more optimistic assessment
of
the uses of international comparisons.
a
The full results will include more data and additional hypotheses.
222
EDUCATIONAL
AND
OCCUPATIONAL CHARACTERISTICS
OF
MANPOWER
223
and educational data in forecasting. Part
I1
of
the paper is concerned
with the analysis of results and ends with some tentative conclusions.
PART
I
THEORETICAL CONSIDERATIONS
I.
The framework for estimating manpower needs
The uses of international comparisons for manpower planning must
be judged in relation to the objectives of the planners. These differ between
countries. But there are now a large number of countries in which
a
major objective can be described in fairly standard terms. The purpose is
to forecast, for
a
given target of national output
(X),
the numbers of
workers required, in some optimal sense, with each type of education,
i.e.,
a
labour input vector which may be denoted
[Li]
where for example
the first element represents the number of graduates.4
It is usually considered that a more reliable estimate of the labour
input vector will be obtained by deriving
a
separate vector
[L,]
for each
economic sector
j
and then aggregating, than by making a direct estimate
of the aggregate.
For
this reason the manpower planner uses the estimated
breakdown of the target output by economic sectors
[Xj].
Corresponding
to this is a forecast of the expected growth of the total labour force
(L),
and an estimate of how this will be distributed between sectors. These
forecasts of the future breakdown of the labour force between sectors are
usually made by estimating the relative rates of change of output per
worker
(Xj/Lj)
in different sectors. This task is notoriously difficult,
since productivity trends in different sectors are not clearly related one to
an~ther.~ But the attempt is normally made.
These estimates of sectoral labour force
(Lj)
and sectoral output per
worker
(Xj/Lj)
are the basis from which the labour input vector
[Lij]
for each sector
is
predicted. There are a number of functions which could
formulate the dependence of
[Lij]
on
Xj
and
Lj,
but
a
common way of
viewing the relationship is to assume that the level of output per worker
specifies the proportionate structure of the labour force, that is,- the vector
[I,,/Lj].
6
'
The classic statement of these terms of reference is in H.
S.
Parnes,
Forecasting
Educational
Needs
for Economic and
Social
Dcuclofimnt
(Paris, O.E.C.D., 1962). We leave till Section
5
the
question of the relation between education, occupation and skill, and continue in this Section
to describe the case where educational needs are estimated directly from economic forecasts,
without use of the intervening variable of occupation.
See John W. Kendrick,
Productiuib Trends in
th
Unitcd
Skates,
A
Study
by
the National Bureau
of Economic Research (Princeton University Press, 1961)
pp.
133-64.
Alternative formulations of the relationship are mentioned in footnote 10.
224
BRITISH JOURNAL
OF
INDUSTRIAL RELATIONS
Thus manpower planners are interested in evidence concerning the
functional relation between
Xj/Lj
taken
as
the predictive variable and
[Lij/Lj]
taken as the variable to be predicted. If international comparisons
are to help manpower planners, one way is to provide evidence on this
relationship. The purpose of this paper is mainly to see if this is possible.
In Section
2,
we explain why it is reasonable to expect to find some
relationship. But there are two points to be made first. What one really
wants to know is how the pattern of educational requirements, rather
than of observed educational qualifications, changes with output per
worker. We may define educational requirements as the pattern of
education which would have been observed if past investment in educa-
tion had been optimal from
a
productivity standpoint. This raises
many questions which are discussed in Section
4,
among them the question
of the relation between education and skill. In order not to prejudge the
issue, Section
2
discusses the Iabour input vector in terms of
skill
rather
than education.
Second, no mention has been made of capital as an element in forecast-
ing manpower requirements. The reason is that, in the typical case which
has been discussed, manpower and capital requirements are estimated
at the same time (though not by a simultaneous optimizing procedure)
and estimates of capital are not given as data to the manpower planner.
This and, more important, the lack of comprehensive information are
the reasons why capital is not included
as
a
variable in this study.
2.
A
model
for
estimating
skill
requirements
We may now set up
a
simplified model for the study. This assumes
that the proportions of people needed with each type of skill depend
on
the
techniques of production chosen, which in turn depend on the production
possibilities available and the relative prices
of
the more important
factors of production. The production possibilities may be summarized
by the production function linking inputs and output:
X=
F
(K,
L,, L,,
. . . . . . L,)
(1)
where
X
is net output,
K
is capital and
L,
......
L,
are the quantities of
labour with each type of skill, and where the implicit function
F
represents
a
summary of techniques of production available to an economic unit.
This function could of course be studied by comparing different kinds of
unit, from establishments or firms up to countries. Our study compares
countries, but does this separately for each sector of the economy defined
as
the eight one-digit sectors of the International Standard Industrial
Classification
(I.S.I.C.).
'
International Standard Industrial ClassiJcation
of
All Economic Activities
(Statistical Papers, Series
hl,
No.
4,
Rev.
1)
(New
York,
United Nations,
1958).
To continue reading
Request your trial