т
Le
IUI
:tUI*i
G
err
s
Educat
n ber
'ee'&en
Unesco : Internationa! Institute for Educational Planning
IIEP/TM/38/69
INTERNATIONAL INSTITUTE FOR EDUCATIONAL PLANNING
7i rue Eugène-Delacroix
Paris l6e, Prance
EMPLOYMENT AND UNEMPLOYMENT
by
Paul Streeten
Warden, Queen Elizabeth House, Oxford
This lecture is part of Fundamentals of Educational Planning!
Lecture-Discussion Series' a controlled experiment undertaken by the
International Institute for Educational Planning in collaboration with
a limited number of organizations and individuals aiming at the
development of efficient teaching materials in the field of educational
planning. By their very nature these materials, which draw upon tape
recordings, transcriptions and summary notes of seminars, lectures and
discussions conducted by IIEP as part of its training and research programme, are informal and not subject to the type of editing customary
for published documents. They are therefore not to be considered as
"official publications'.
The opinions expressed in this lecture are those of the author
and do not necessarily represent the views of the Institute.
The use, adaptation or reproduction, in whole or in part of
these materials is limited to institutions and persons specifically
authorized by IIEP.
i
1ШР/ТМ/38/69 - page 1
EMPLOYMENT AND UNEMPLOYMENT
Most less developed countries have what is commonly described as a
serious and growing problem of surplus labour, unemployment, disguised
unemployment and underemployment, Whatever the precise interpretation of
these somewhat vague and ill-defined terms, there can be no doubt that
development policies must give a high priority to a fuller mobilization
and utilization of what is sometimes thought to be their most abundant
factor of production - unskilled labour. It is worth considering briefly
the notions of unemployment* disguised unemployment and underemployment.
Unemployment, underemployment and disguised unemployment are often
considered both a cause of poverty and a potential source of developmentApproaches in terms of 'employmentf, 'unemployment* and 'underemployment'
are misleading because they suggest that an increase in effective demand
and the provision of equipment are all that is needed to absorb labour
and raise production, while all other conditions are adapted or
easily and quickly adaptable to full labour utilization * In fact, a
number of other measures are necessary for a full mobilization and
utilization of manpower, which may be said to lie partly on the side of
demand for labour and partly on the side of supply, except that this neat
distinction sometimes breaks downs better feeding, improvements in health,
training and education, transport and housing and fundamental attacks on
prevailing attitudes to life and work (e.g. women's participation, a
contempt for certain kinds of work, the desire to minimize work, lack of
discipline) and on institutions (introduction of standard working week and
working day, creation of labour market, provision of information, readiness to move from one place to another or to change one's occupation etc.)«
As a first step, it is helpful to break down the multiplicity of
dimensions of Income (or Product) per Head of the Population into four
categories. These should aid the collection of data, the organization of
thought and the formulation of policies.(l)
Income
Production
Hours worked
Labour force
Population
Hours worked
Labour force
People of
working age
People of
working age
Population
(1) The identity was first used by Michael Lipton in his work for
Gunnar Myrdal. See also his Assessing Economic Performance, p* 40,
W.J. Barber 'Some questions about labour force analysis in
agrarian economies with particular reference to Kenya', East
African Economic Review, June 1966 and Gunnar Myrdal, Asian
IIEP/TM/38/69 - page 2
The identity brings out four distinct aspects of the Level of Living
(= Income per Head) on which more information would he useful for framing
policies for the multidimensional aspects of labour utilisation. Separation
of the four ratios does not imply that there is no interdependence between
them: there clearly is(1)
Production
Hours worked
or hourly productivity
depends, in any given activity in any given sector, on a large number of
factors, including other terms in the identity, such as: hours worked and
participation rate (see below (2) and (3)Ь equipment, fuel, raw materials
and other complementary productive factors ; education and training! health
affecting work such as intestinal parasites, amoebas, onchocerciasis or
schistosomiasis, levels of nutrition, intensity of application, itself a
function of moralej industrial relations j motivation! incentives, etc;
organization of work, management, etc*
This category covers numerous aspects, some of the most important of
which are difficult to measure. It should be analysed in greater detail,
For the country as a whole, it is an average of all sectors, each weighted
by its share in the total number of hours worked» If we denote the sectors
as 1, 2, 3 etc., and their shares in total working hours as h,, h p , etc.,
Output = Ц
Hours
^L
H,
- h2
^2
Hp
+
......
Hourly productivity can be raised if all other things remain constant,
either by transferring workers from low-productivity to high-productivity
sectors, or by raising productivity within sectors.
(2)
Hours
Labour force
or working time
depends on organizational and institutional factors ; whether there is a
standard working day and working week! whether overtime is workedj whether
multiple shifts exist! whether time is wasted in idleness, waiting for
materials and components, or spent on holidays, weddings, funerals and at
feasts. It also depends on natural factors such as the weather and the
requirements of harvest seasons. The ratio will depend both upon the level
of demand and on the availability of essential supplies. A shift of rural
labour to urban industry raises output not only by changing the weights
attached to low- and high-productivity sectors, but also by raising hours
per labour force, unemployment of people both willing and able to work
will show up as low hours/labour force. But the distinction between ability
to work and willingness to work in any occupation outside the home may not
IIEP/TM/38/69
- page 3
always be easy to draw or even logically legitimate (e.g. Moslem women).
Much time is spent in an underdeveloped country moving from one place to
another: peasants walk from one piece of their land to another; women
walk back and forth to draw water; migrant workers walk from one region
to another to collect the harvest, etc, In so far as these movements
are necessary to carry out specific tasks, given the prevailing institutions, transport facilities and co-operating factors, it is a factor
accounting for low hourly productivity. But if the movements are in
search of work, they come under low working time .
(3)
Labour force
People of working age
or participation rate
depends on attitudes to work and to gainful activities (their dignity
or ignominy), housing and transport facilities, legislation about
minimum working age, compulsory full-time education, etc. Removal of
the objections to certain kinds of work, increased incentives to earn
money, emancipation of women, improved mobility, etc. will raise participation rates.
Education can, in certain conditions, result in reduced labour
force participation. The educated unemployed, a widespread phenomenon in
South Asia, figure prominently in unemployment statistics. While their
geographical mobility between urban areas is high, their occupational
mobility is negligible. They are not prepared to accept manual work.
Prom a sample survey of unemployment in Calcutta in 1953 it appears that
only 10 per cent of the unemployed were illiterate and 27 per cent had
enjoyed higher education. Only 43 per cent of the total sample were
seeking work involving manual labour.
The attitude to work appropriate for one who has enjoyed education
is rooted in traditional attitudes and re-inforced by the colonial
heritage and possibly even by technical assistance. It is by no means
just a matter of the wrong curriculum, for there is large and growing
unemployment of engineers in India. If the existing engineering colleges
merely maintain their capacity, while national requirements expand at
the projected rate, there will be 50,000 fully qualified engineers unemployed in 1970. The shortage exists in the same occupation for less
qualified people, e.g. for semi-skilled technicians.
Both in Asia and in Africa education reflects and instils an
anti-rural bias; indeed the pressures for education arise from a desire
on the part of parents to free their children from the miseries and hardships of rural life.
Attitudes towards work among the educated - often ill-educated are deeply rooted in the social structure and cannot easily be eliminated
by restoring 'equilibrium' between supply and demand, by changing
curricula or by exhortation.
IIEP/TM/38/69 - page 4
There are parallels between the participation rate of the unemployed and the participation rate of women. Non-participation of women
is linked with status and prestige, particularly in the higher strata of
society.
(4)
People of working age
Population
is a demographic ratio and
will depend on the age structure of the population, It can be predicted
with a fair degree of accuracy for some time ahead.
Since each category is an average of sectoral ratios, the identity
can be rewritten as;
H,
H,
Y
h
P
h
i ÏT
2
Hi
2 L,
Л
L
P
where
Y
H
L
W
P
h
1
p
s
is
is
is
is
is
is
is
is
is
l
i
V^
+
P
2
W,
5
2 T,
total income' (output)
total hours worked
labour force
working age group
population
share in total hours worked
share in labour force
share in age group
share in population
and the suffices indicate the different sectors.
The conventional presentation suffers from the fact that intensity
of work, skill, organization, education, health, labour markets, transport,
information, etc. are assumed given. Thus the only variables are demand
and equipment. Furthermore, the assumption is usually made that unemployment and underemployment are 'involuntary'.(1) This implies that
(1)
',.. there is a great deal of idleness, voluntary, and involuntary'
(in Colombia), L. Currie, Accelerating Development, p. 168. In the
Ivory Coast, the essence of the primitive methods of producing coffee
is described by Professor Barna as minimizing the amount of work
necessary for obtaining a coffee crop of any sort.
IIEP/TM/38/69 - Page 5
willingness and ability to work are present. Unemployment and underemployment must also be defined with reference to some standard of working hours
per day and working days per week. But such standards do not exist in
large parts of developing societies and are therefore introduced, usually
implicitly, from outside. The whole set of questions relating to participation and organized work is thereby begged and a number of important
relationships are concealed. The distinction between voluntary and
involuntary unemployment, so crucial in our intellectual framework, does
not make sense in an environment without a labour exchange and hence without an objective test. Some men work with dysentry, others do not. And
how are we to describe the common situation that men do not seek work
because they know none is available?
Policies can then be classified according to whether they use compulsion, permission or persuasion. The following table provides illustrations. Further sub- and cross-division could classify measures according
to whether they are general or specific, and whether they are positive or
negative.
Measures
Compulsive
Permissive
Persuasive
(incentives)
Output/Hour
Make pay depend
on minimum output
Forbid trade union
restrictions
Piece rates
Hours/Lab.
force
Fix 8-hour day
Improve diet
Overtime
rates
Lab, force/
People of
working age
Lock up workless,
conscript, poll
tax
Raise demand,
provide equipment
Raise wages,
supply incentive goods
People of
working age/
Population
Draconian measures
against large
families, forced
late marriage
Birth control advice and contraceptives supplied
Birth control
campaigns, a
transistor
for a vasectomy, child
tax
CONCLUSIONS
The main conclusion of this brief discussion is that the utilization
of labour in developing countries has many dimensions and that it is not
warranted to assume that attitudes, aptitudes and institutions are adapted
11ЕР/Щ/38/69 - page 6
to full labour utilization. Measures which raise labour productivity may
reduce hours worked or participation rates(l) and measures which raise
participation rates may lead to work-spreading and less intensive or to
otherwise less productive work. Only a simultaneous attack on several of
the relevant variables can bring about fuller utilization of labour. The
discussion also explains the paradoxical co-existence of labour surplus
and labour shortage.
CREATION OF EMPLOYMENT OPPORTUNITIES
The analysis also bears on the argument that there is a surplus of
unskilled labour to draw on for any alternative activities and that labour
opportunity costs are therefore low or zero, If these alternatives
require attitudes, motivations, responses, work habits or institutions
different from those to be found in, say, coffee growing, the fact that
coffee growers dispose over spare time is irrelevant to the availability
of labour supply for these alternatives. It would be dangerous to argue
from the premise that coffee-growing does not take up all the potential
working hours of the farmers to the conclusion that alternative work
opportunities, either elsewhere or in the place of residence, would auto­
matically be taken up and result in larger production. Proposals for
alternatives must be accompanied by detailed specification as to what
measures of reform with respect to human attitudes to work and life and
social and commercial institutions, such as land reform, or reform of the
civil service, or the creation of a labour market, or of credit channels,
or of marketing outlets, have to accompany this shift in resources.
The main causes of the gross underutilization of labour are to be
found in rural underemployment, combined with an industrial sector which,
though often growing very rapidly in terms of production, is too small
and uses often techniques inappropriate to absorb even a fraction of the
rapidly growing potential labour force.
The small base of the industrial sector is an important reason
for its limited capacity to create employment opportunities and absorb
the rural surplus population. With 20 per cent of the labour force
employed in manufacturing and 80 per cent outside it and population
growth (assumed for simplicity to be identical with the growth of the
labour force) of 3 per cent, employment opportunities in industry would
have to grow at 15 per cent in order to absorb only the total additions
to the labour force, without reducing already existing unemployment and
underemployment. If only 10 per cent of the labour force are in industry*
(l) ' ... the relative high productivity of the machine, and the use of
better techniques in commercial farming lower the return the colonialtype farmer can gain and make it even less practical for him to do all
the costly things that would increase his productivity."
L. Currie, Accelerating Development, p- 156.
IXEP/TM/38/69
- page 7
employment growth would have to be 30 per cent. Although growth rates in
industry have, on occasion, been not much below these rates, the methods
used in manufacturing, often transferred from the advanced countries
where they were developed in conditions of labour scarcity, tend to be
labour-saving. Moreover, where the composition of manufacturing output
shifts from light consumer goods to heavier industrial products, the
weight of the more capital intensive activities in the total rises and
the demand for labour tends to" be less than it would otherwise be, even
if techniques in each sector are unchanged, Even if labour-using methods
are available, they often require management, supervision and other skills
in larger proportions than do some labour-saving methods.
The tendency to use labour-saving methods is reinforced by minimum
wage legislation and trade union pressure for higher wages which discourage employers from making use of the labour even where techniques
appropriate to the employment of plentiful cheap labour are available or
where the production of labour-intensive products would be profitable at
lower wages. Total industrial employment in East Africa has in fact
declined over the last ten years.
While there can be no doubt about the urgency of employment
creation in a well-designed development strategy, the discussion is sometimes obscured by two types of confusion: one between the choice' of
labour-intensive industries and products and the choice of labourintensive techniques within a given industry in producing a given product1
the other between labour-intensity in relation to capital and in relation
to output.
Foreign trade apart, there is relatively limited scope for permitting the composition of industries and the product-mix to be determined by factor availabilities. Hair-cutting is labour-intensive, but
this can be no reason for developing countries to establish large barbering
sectors. Some consumer goods are more labour-intensive than many capital
goods, but this can be no reason for encouraging consumption at the
expense of investment. Where there is scope for choosing labour-intensive
products that serve the objectives decided upon, clearly these should be
chosen. And having determined the sectoral structure and the product-mix,
it is sensible to employ techniques, where these are available, which
economize in scarce factors and make fuller use of abundant factors. In
some industries, there is a variety of processes between which one can
choose.
Next it is important to avoid saving capital in relation to labour
in a way which reduces output per unit of capital, In particular, the
advocates of labour-using techniques sometimes fail to pay adequate
attention to the capital requirements of inventories and work in progress«
IIEP/TM/38/69 - page 8
While it is desirable to use, ceteris paribus, methods of production
which save capital per worker, it is not desirable to do so if this means
using more capital per unit of output, i.e. reducing production by more
than in proportion to the capital saved. The ratio of labour to capital
is identically equal to the ratio of labour to product, divided by the
capital/output ratio. An increase in the labour/capital ratio should not
be achieved at the expense of an increase in the capital/output ratio, so
as to reduce the productivity of labour.
In the conditions prevailing in many underdeveloped countries, the
right result cannot always be taken for granted. It is likely that socalled labour-intensive methods will show less fixed capital per worker.
Visible structures and equipment may be smaller. But particularly in
rural and dispersed industries, requirements of working capital (inventories and work-in-progress) are likely to increase. Work-in-progress is
likely to be greater, the more dispersed the industry, the longer the distances from the centres, the higher the wastage of material and the
greater the losses from lack of standardization. Inventories are also
likely to be greater because economies of scale will not be available in
stock holding. Investigations in India have shown that labour-intensive
methods of production are, for these reasons, often capital-intensive in
relation to output.(1)
The argument so far has been that labour-intensive methods should
be efficient, i.e. they should not require more labour and more capital.
In many cases there will be little or no choice of techniques and capitalintensive methods only will be available. If the raw material is available locally, if other criteria are favourable and unit costs of production
are internationally competitive, capital intensity should not in itself
stand in the way of adopting the process.
In the long run, only accelerated successful development can solve
the problem of unemployment and for this production at low costs is
essential.
NOTE ON THE MEASUREMENT OF UNEMPLOYMENT AND UNDEREMPLOYMENT
In her well known essay 'Disguised Unemployment' Joan Robinson
coined this term for a situation widely observed in the Great Depression
in which men, thrown out of regular employment, crowd into occupations
like carrying bags, rendering small services or selling matches in the
Strand. The reasoning can best be brought out by a simple two-sector
model: in one sector money wages are rigid downwards, in the^other, where
self-employment is common, incomes are flexible. In competitive full
employment equilibrium, the marginal productivity of labour is the same in
both sectors. If then a fall in aggregate demand below the full employment
(1) P.N. Dhar and H.F. Lydall, The Role of Small Enterprises in Economic
Development (Asia Publishing House, 1961).
IIEP/TM/38/69 - page 9
level occurs, men will be thrown out of work in the rigid wage sector but,
rather than become unemployed, will move into the flexible income sector.
Money income per man in this sector will fall as more men are accommodated
to spread a smaller work load. Productivity differentials (measured in
terms of man years, man weeks or man days, but not in terms of man hours,
for productivity of hours not worked is not meaningful) will increase, but
no visible unemployment will appear. The difference between a situation
of general low labour productivity (say due to absence of skills) and a
situation of disguised unemployment in this sense is that a rise in the
level of effective demand will shift workers back into the highproductivity, rigid-wage sector and remove the disguised unemployment.
The workers are adapted to the requirements in this sector and, if the
time spent in the flexible sector has not been too long, so that they
have not forgotten their skills, have remained well fed and healthy and
have not been demoralized, a rise in effective demand is sufficient.
It is immediately obvious that the situation in the rural sector
of developing countries is quite different from that described by
Joan Robinson, It is true that the rural subsistence sector, in which
small holdings are cultivated by families, resembles the flexible income
sector in that it is capable of spreading a constant or slowly growing
work load and product over a rapidly growing number of people. But it is
not true that an increase of effective demand would, by itself, absorb
the excess population in industry. Clearly a series of additional
measures would be necessary. Machinery and equipment would have to be
provided, a work force would have to be trained, disciplined and educated
in co-operation, nutrition and health may have to be improved, public
services would have to be provided, objections to factory work would have
to be removed, etc.
But, looking at the problem simply from the point of view of rural
surplus population, it may make sense to ask the questions How many
people can be removed from agriculture without reducing output?
Rosenstein-Rodan, in a subtle analysis, distinguishes between two basic
concepts;(1) the static and the dynamic one, according to whether methods
of cultivation are assumed not to change or whether they are assumed to
change, when the surplus population is removed, while output remains
constant,
Rosenstein-Rodan claims that the static concept is clear, whereas
the dynamic concept requires a detailed specification of what changes in
methods of cultivation are envisaged. These changes might vary from minor
changes 'obtained merely through a rearrangement of work with but small
additions of circulating capital' to throughgoing and even revolutionary
(1) P.N. Rosenstein-Rodan, 'Disguised Unemployment and Underemployment in
Agriculture', Monthly Bulletin of Agricultural Economics and
Statistics, Vol. VI, Nos. 7/8, July/August,. pp. 1-7.
IIEP/TM/38/69 - page 10
changes, 'including additional use of both fixed and variable capital*. The
dynamic concept, carried to the extreme, becomes irrelevant for policy,
because it raises questions such as what would surplus population be if the
agricultural sector of an underdeveloped country were cultivated under Dutch
conditions.
Rosenstein-Rodan discusses two methods of measuring disguised unemployment in agriculture. The first and direct method is based on empirical sample
enquiry with questionnaires distinguishing between different types of
cultivation, different sizes and forms of property, the composition of the
labour force and the 'labour diagram' (number of labour hours required and
supplied). Such an enquiry would distinguish between permanent disguised unemployment and fractional or seasonal unemployment.
The indirect method may be used in three variants:
"(a) The number of labour hours required to produce a given
output is subtracted from the number of labour hours
available from the active agrarian population. The
difference represents the agrarian surplus population.
(b) The density of population deemed adequate for a given
type of cultivation is subtracted from the actual
density of population. In order to take into account
different grades of fertility of the soil, conversion
coefficients of arable equivalents are used, for example?
1 hectare of garden = 3 hectares of cultivated area;
1 hectare of meadow =0.4 hectare of cultivated area, etc.
(J. Poniatowski, quoted in 'Population in Agriculture',
League of Nations, 1939).
(c) The number of hectares required under a given type of
cultivation to provide one person with a 'standard income'
is contrasted with the number of hectares and the agrarian
population available. The difference represents people
for whom there is no land available and who are therefore
'surplus'. For income calculation 'crop units' are used
by H.E. Moore, instead of the arable equivalents (area
conversion coefficients) of J. Poniatowski,"
But even on the most stringent static assumptions, amounts of labour
required, adequate density and adequate income are vague concepts and
involve value judgments.
Rosenstein-Rodan employs the direct method to calculate disguised
unemployment in Southern Italy and spells out very lucidly the assumptions
which have to be made in such an enquiry:
IIEP/TM/W69 - page 11
1,
Only agricultural small holdings of 'direct cultivators'
(peasant owners and tenants) are considered. Employed workers, even
though they may spend time in idleness, are assumed not to be
underemployed•
2.
The agricultural area is divided into representative types of
cultivation and each of these types is grouped into holdings by
size,
3As labour force in each holding, he assumes active population
to comprise persons from 14 to 65 years of age- Fractions can be
attached to children below 14 and adults over 65. Those who work
outside their own holdings are excluded. Problems arise if outside
work is part-time and if certain jobs are traditionally done by
women and others by men and they object to changing this.
4.
It is assumed that one woman in a household of four is
occupied in household activities and not available for cultivation.
For larger-sized families greater numbers are assumed,
5.
It is assumed that those who are in surplus are involuntarily
unemployed. Where, owing to custom or religion, women do not
accept work outside the home, they should not be counted as disguised unemployed. But it is not entirely clear how men are to be
treated who would not wish to do more work than they are doing now
or who would object to different kinds of work from the one they
carry out now.
6.
Labour hours required for each type of cultivation over
the whole year* month by month, are counted and compared with
available labour hours, In the resulting seasonal underemployment,
two kinds are distinguished: first, 'seasonal underemployment of
the productive capacity', which depends on biological and
technical factors in growing crops and second, "seasonal underemployment proper', which takes into account labour not available
for climatic reasons, such as snow, and institutional reasons,
such as holidays. These reduce the number of working days available during the year*
7Next, allowance is made for the fact that the number of
labour hours available in different months varies: fewer in
winter and more in the summer.
IIEP/W38/69 - page 12
8.
After labour hours have been calculated in terms of labour
units (men and women) allowance must be made for the fact that not
all underemployed thus calculated could be removed without causing
output to decline. Only entire labour units (men and women) whose
removal would not cause such a decline can be considered as surplus.
Rosenstein-Rodan therefore distinguishes between (a) removable disguised underemployment; (b) disguised fractional underemployment,
i.e. labour hours not used through the year which do not add up to
an entire labour unit. These cannot be removed without a decline
in output but they can be provided with part-time work in rural
industries, rural public works, etcj (c) seasonal underemployment due to climatic factors. Even a brief seasonal peak,
together with a serious adherence to ceteris paribus, greatly
reduces the amount of disguised unemployment.
9Rosenstein-Rodan suggests a slight relaxation of the strict
ceteris paribus rule to permit re •»organization of peak loads of up
to two months and in this way the size of the surplus population
can be considerably increased - doubled in Southern Italy.