INTERNATIONAL LABOUR OFFICE

STUDIES AND R E P O R T S
Series B (Economic Conditions) No. 30

THE WORKER'S
STANDARD OF LIVING

GENEVA
1938

Published in the United Kingdom
For the INTERNATIONAL LABOUR OFFICE (LEAGUE OF NATIONS)

By P. S. KING & SON, Ltd.
Orchard House, 14 Great Smith Street, Westminster, London, S.W. 1

PRINTED

BY ALBERT KUNDIG, GENEVA

CONTENTS

Page
INTRODUCTION

5

CHAPTER I: Determinants of the Worker's Standard of Living . . . .

7

General Well-being or " Standard of Life "

8

II.

I.

A " Sum of Economic Satisfactions "

9

III.

A Composite of Goods and Services

11

IV.

Levels, Norms and Standards of Living

12

V.
VI.
VII.
VIII.
IX.

Income, Prices and Consumption

14

Population Growth and Family Type

16

Social Services, Free Services, and Working Conditions . .

17

Group and National Differences

18

Summary and Definition

20

CHAPTER II : Description and Evaluation of Workers' Standards of Living
I.

21

Income
A. Distribution of Family Incomes
B. Average Family Income

23
23
30

Expenditures
A. Food and Nutrition
B. Housing
C. Clothing
D. Miscellaneous

31
32
39
47
49

III.

Social
A.
B.
C.

51
51
53
54

IV.

Problems of Evaluation
A. Intra-National Time and Place Comparisons . . .
B. International Comparisons
C. Other Methods
D. The Measurement of Deficiencies
E. Norms, Standards and " Freedom of Choice " . . .

II.

Elements in the Standard of Living
Health
Education
Social Services

54
55
56
57
57
59

— 4 —
Page
CHAPTER III : Some Aspects of Workers' Standards of Living in Certain
Countries . . :
I. The United States of America
A. Family Income per Unit
B. Food Consumption and Nutrition
C. Housing
D. Health
E. Literacy
II.

III.

IV.

63
64
64
66
71
73
74

Poland
A.
B.
C.
D.
E.
F.

Income
Expenditure
Food Consumption
Housing
Health
Literacy

'

75
75
77
78
80
82
82

India
A.
B.
C.
D.
E.

Income and Expenditure
Diet Adequacy and Malnutrition
Housing
Health
Literacy

82
83
85
87
88
89

Japan
A.
B.
C.
D.
E.
F.

Income and Expenditure per Family
Income and Expenditure per Consumption Unit
Food Consumption
Housing
Health
Literacy

CHAPTER IV: Summary

.-

.

89
89
90
92
94
96
97
98

INTRODUCTION

At its Session in October 1937 the Assembly of the League of
Nations adopted the following resolution:
" The Assembly, recognising that technical progress in industry,
agriculture and transport has made possible further advances in
human welfare and recognising also, that such advances depend
upon economic co-operation between the nations:
" Invites the Economic and Financial Organisation of the League,
in collaboration, when appropriate, with the International Labour
Office, to take whatever steps it may deem appropriate for the
examination of the following problems:
" Measures of a national or international character for raising
the standard of living, the first stage of this enquiry being
limited to a preliminary investigation, if necessary with the
help of experts, in order that the next Assembly may be
enabled to devote further attention to this question."
This Report has been prepared in accordance with the above
resolution. It is preliminary in character, and is intended mainly
to outline the problem and to indicate lines for further study.
The first task is to clarify what is meant by standards of living
and what are some of the determinant factors in the workers'
standard of living. This is the subject-matter of Chapter I. In
Chapter II a survey is made of available materials for the description of workers' living standards in different countries. ' In the
same chapter are considered some of the problems involved in
comparing living standards at different periods and between
countries. The idea of national minima is suggested as a basis
for evaluating workers' living standards and as a possible guide
to policy. This is followed in Chapter III by a brief summary
of available evidence on workers' living standards in four countries
—the United States of America, Poland, India and Japan.
This summary is presented, not for the purpose of comparing
conditions in these countries, but chiefly to illustrate methods
of describing workers' living conditions. The facts presented,
however, give some idea of the nature and extent of poverty and
of low standards to-day. Finally, in the summary chapter,

— 6—
some of the more important lines of further investigation of the
subject are indicated.
In the preparation of this preliminary Report the Office has
had the co-operation of a number of agencies in several countries.
In the United States, the Bureau of Labor Statistics of the Department of Labor and the Bureau of Home Economics of the Department of Agriculture have been particularly helpful. In Poland, the
Institute for Social Problems and the State Institute for Hygiene
have kindly supplied data on costs of adequate diets, on house
rents and on food consumption. With regard to food consumption
and nutrition in India, valuable assistance was given by the
Nutrition Research Laboratories of Coonoor. The Office takes
pleasure in acknowledging the assistance of these as well as of
other agencies in the preparation of this Report.

CHAPTER I
DETERMINANTS OF THE WORKER'S STANDARD OF LIVING

The term " standard of living " has had a long career in the history
of economic and social thought and a variety of definitions could be
cited apparently conflicting in character. The differences in these
definitions can, however, be traced largely to differences of emphasis
on different aspects of the problem. Some writers identify standards
of living with total income (money income plus income in kind),
others with the goods and services purchased with the income and
actually consumed or considered necessary by the individual or
family. Still others find the essence of the standard of living not in
the consumption of any particular set of goods or services, but
in the maintenance of a given level of well-being which may be
equally well provided by various combinations of goods and services.
The idea is sometimes expanded to include well-being derived not
only from goods and services but also from leisure, health, cultural
pursuits, and other non-economic sources.
Each of the aspects emphasised in the above definitions has a
certain relationship to all the others. Thus, money income sets a
limit to the possible total consumption of goods and services which
a given individual or family may purchase. The goods and
services purchased and consumed are dependent on tastes and
habits and on prices, and contribute in large measure to health,
efficiency and the feeling of well-being. Leisure maybe an alternative
to larger income and to more consumption. Lastly, general wellbeing, in the widest sense of the term, is correlated with the
attainment of desired health, comfort, adjustment to the community, which are in turn dependent, at least partly, on economic
consumption.
The question at issue is to draw from the foregoing series of
relationships those which may be considered as pertaining to the
worker's standard of living and which may be used as a basis

— 8—
for the present Report.
In the following pages a" more
detailed analysis is made of the relations between the different
aspects of the term " standard of living " leading up to a working
definition which may serve the purposes of practical policy.

I. — General Well-being or " Standard of Life "

The phrase " standard of life " has been used to express the
broadest concept discussed above, namely, general well-being in
the widest sense of individual happiness. The " standard of life "
or " way of life " would include all elements, economic and noneconomic, which go to make up the life-pattern of the individual
and which contribute to his well-being (or ill-being); that is, not
only food habits, housing arrangements, customary clothing and
consumption of other economic goods and services, but also
working conditions,,cultural enjoyments, the religious life, community activities, family relations, etc. In short, the phrase covers
the entire range of economic, social, political, intellectual, moral
and cultural considerations which affect a person's sense of
happiness.
There can be no question about the influence which the social
and cultural factors of life have on welfare in the narrower economic
sense of the term 1. Thus, religious beliefs may affect food habits,
and the ideas of general well-being may influence the valuations
placed on material and economic standards. In many countries
attempts to improve material standards of living would involve
policies affecting general social and cultural ideas and institutions.
Nevertheless, it is obvious that this large concept of a " standard
i " There is no precise line between economic and non-economic satisfactions . . . we can proceed from the undoubtedly economic at one end of the
scale to the undoubtedly non-economic at the other end without finding
anywhere a fence to climb or a ditch to cross." Edwin CANNAN: Wealth,
quoted by A. C. PIGOU: Economics of Welfare, London, 1922, p. 11.
The following quotation may further illustrate this trend of thought:
" . . . a moral judgment is implied in any operative standard of life. It is
worth noting also that the conception of a standard of life is the point in which
economics touch the frontier of other interests, for certainly no one supposes
that any economic standard (i.e. any standard calculable in exchange values)
can be discussed without reference to items in the general standard of life
which have no exchange value. For example, a standard of life morally
includes a reference to friendship or affection ; no goods and services can be
included in any standard which does not promote these non-economic goods,
and therefore all assessments of the standard of life in terms of exchangevalue should include a reference to non-economic goods." C. DELISLE BURNS :
Industry and Civilisation, 1925, pp. 193-194.

— 9 —
of life " in the sense of a " way of life " involving thè entire gamut
of the relationships of man to his neighbour, to his environment,
and to himself is too vague for t h e purpose of realistic studies of
existing economic and social conditions as well as for purposes of
immediate social policy. While this general concept is of value in
thinking about t h e problem, it is beyond t h e scope of the present
Report.

II. — A " Sum of Economic Satisfactions "
According to another view, the essence of the standard of living
lies in the subjective satisfactions arising from the consumption
of economic goods and services. According to this definition,
a standard of living does not hinge on any specific collection of
goods and services. An individual of a given economic or social
group would not mind giving up one commodity for another,
provided the satisfaction he derives from the consumption of the
two is equivalent or, in other words, provided he feels equally
well off in t h e two cases. This idea was incorporated in t h e definition of the standard of living quoted in a previous report of
the Office as follows:
The term " standard of living " or, as it has sometimes been called,
" standard of comfort " is intended to cover the sum of the economic
satisfactions or utilities which an individual (or a family) derives from
the consumption of the goods and services which he is able to obtain
with his income available during a certain period of time 1.
The above definition, being subjective, is also highly individualistic. The standard of living is regarded as a sum of satisfactions, a level of well-being to the maintenance of which the
individual attaches great importance. That the sum of satisfactions is not wholly arbitrary is due to the fact that the definition
assumes (1) that the set or sum of satisfactions is considered
essential by the individual concerned, and (2) t h a t it is customary
to him. The definition implies the possibility of dissatisfactions
caused by the necessity for work, by deprivations of things one
would like to have b u t cannot afford, etc. An individual's state
of well-being as judged by himself is thus a balance of satisfactions and dissatisfactions, utilities and disutilities which tends
1

INTERNATIONAL LABOUR OFFICE: A Contribution to the Study of Inter-

national Comparisons of Costs of Living. Studies and Reports, Series N, No. 17,
p. 7.

— 10to form a consistent whole and to remain in equilibrium despite
the changes in the specific things consumed.
Owing to its subjective character, the above definition of the
standard of living leads to a number of difficulties. As a psychological fact, the standard of living—at least in the present stage
of our knowledge—cannot be measured in absolute units. In the
case of a single individual, it is not possible to say how much
satisfaction he derives from a given object; all that it is possible
to say is that one can infer from the purchases he actually
makes that he derives more satisfaction (without saying how
much more) from one object than from another. Even more important is the fact that it is impossible to compare the satisfactions
of two or more persons. Since there is no way of measuring the
intensity of feeling, it becomes impossible not only to say by how
much one man's satisfaction is greater than the other's, but even
whether it is greater or less. There could logically also be no
such thing as a standard of living of a group of persons. Furthermore,
it would be impossible to form any measurement of standards
of living, the sole criterion being the subjective state of mind of
individuals. On the basis of this definition, the phrase " raising
the standard of living " means increasing subjective satisfactions,
which means that no proposed measure for raising living standards
can be judged good or bad by anyone other than the individual.
There would also be no way of judging whether certain actions
would result in an increase of subjective welfare of a group of persons. If the definition of the standard of living were left at this
point there would appear to be little occasion for writing a report
dealing with measures for raising standards of living.
To overcome the dilemma outlined above, those who proceed
on the basis of the subjective definition of standards of living
are forced to make two main assumptions. First, that satisfactions
are reflected in the way an individual spends his income, subject
to limitations of price and of certain conventional habits. Second,
that individuals whose economic position is the same and who
belong to the same social group derive identical amounts of satisfaction from their consumption provided there are no special limitations to their freedom of choice and they actually do choose the
same amounts and kinds of goods and services. These two assumptions are supposed to supply the bridge for the possibility
of using statistical and other materials for the description
and study of living standards of groups of individuals and
families.

— 11 —
The validity of these assumptions is open to question, but it is
not necessary to examine them here. What is important is that
once they are made the argument is turned towards a consideration
of actual goods and services consumed, income, and prices. In
other words, the ground is shifted from subjective satisfactions to
the objective elements in the standard of living, which is the basis
oí the next definition considered.

III. — A Composite of Goods and Services
In contrast to the subjective definition given in the preceding
section, most of the more recent definitions of the standard of
living run in objective terms. Despite the variations in their
wording, most of them are in terms of " a sum ", " a composite ",
" an amount ", or " a supply ", of goods and services enjoyed by
an individual or family within a given period of time on the basis
of a given income. For purposes of illustration, two definitions are
quoted below.
By the standard of living, then, I understand a composite of the goods
and services obtained in nearly the same quantities by normal families
whose general mode of life is similar 1 ; or:
In general, a standard of living may be defined as such a sum of
commodities and services of all sorts as is habitually consumed in a
given length of time—say a year—by a person or a family group of
given size and given social status. It is not itself a sum of money
but a sum of goods. Its money cost will necessarily vary at different
times and places and must therefore be recalculated periodically 2.
As may be readily seen, the above definition not only shifts the
emphasis from satisfactions derived from goods and services to
the goods and services themselves, but it also stresses the objective
and social basis of the standard of living. As inductive studies
of family budgets and consumer purchases show, families which
derive their income from similar sources, which customarily receive
about the same amount of income (within a given range), which
are in approximately the same social-economic group, and are
of similar "size and composition, usually display similarities
in the spending of their income, or, in other words, make approximately the same choices of goods and services. Such groups
thus may be said to have the same standard of living as is shown in
1

Paul and Dorothy DOUGLAS: What Can a Man Afford ?, p. 50.
BOWLEY : The Nature and Purpose of the Measurement of Social Phenomena, 2nd edition, London, 1923, p. 164.
2

— 12 —
the complex of goods and services on which they spend their
income.
IV. — Levels, Norms and Standards of Living

One of the difficulties in studying workers' living standards
has been the confusion of terms due to the fact that no clear
distinction was drawn between what the workers actually consumed,
what they individually thought necessary or desirable, and what
was recommended as necessary or desirable as a result of scientific
investigation. To obviate this difficulty, the effort has been made
in recent years to introduce more generally the use of three terms,
especially with regard to consumption, namely, levels, norms, and
standards.
The plane or level of living (or " content of living ") is defined
as the composite of goods and services actually consumed, which
may or may not be identical with what the individual or family
regards as necessary or desirable. In general, the goods and services
consumed by an individual or family are definitely limited in total,
at a given time, by the amount of income. The actual composition
of the items making up that total depends upon the tastes and
habits of the person or family in question and on the relative prices
of the various goods and services available in the markets to which
they have access. The level of living might be expressed in terms
of the money income which purchases it, or in terms of the degree
of health or comfort which it yields, depending upon the nature
of the problem in hand.
Norms of consumption, on the other hand, represent a combination of goods and services recommended by experts who base
their judgments on certain objective criteria. Thus, nutritionists
may define the kind and amount of foods necessary to achieve
a diet adequate to ward off certain diseases or to attain a positive
state of health (providing other factors are also favourable) ;
hygienists may specify housing requirements necessary to meet
minimum requirements adequate to avoid the occurrence of certain
diseases or sub-normal health conditions, excessive infant mortality,
and so forth. Experts may, to a less certainly objective degree,
specify the kinds and quantities of clothing necessary to meet the
hygienic requirements of the body. While health has been most
often used as the criterion for norms, it is apparent that experts
(for example, psychologists or social workers) might define certain
norms of consumption, work and leisure whose content would be

— 13 —
selected with a view to achieving other objectives, such as emotional
balance, comfort or cultural development. It is apparent, however,
that such criteria are less capable of objective measurement and
that there would be a necessarily greater element of arbitrariness
in the selections made by the experts if they tried to prescribe
the norms or means to attain such ends.
Norms of consumption, while based on objective criteria, are
not necessarily or usually divorced entirely from other factors.
Thus, experts in setting up norms as a rule take account of
tastes and habits at the points where the physiological or other
criteria permit a choice. Thus, diets recommended by nutritionists
as adequate from a nutritional standpoint are usually expressed,
as far as possible, in terms of the foodstuffs customarily consumed
by the group for which the diets are proposed. There is in fact
an indefinitely large variety of specific combinations of foods,
any one of which would be satisfactory from the physiologist's
standpoint (i.e. would contain the requisite numbers of calories,
grammes of protein, etc.) and among which the consumer might
freely choose on the basis of his income and tastes while at the same
time meeting the requirements of the norm. Furthermore, the
various possible combinations of goods and services required by
the norm may be expressed in terms of the money necessary to
buy thern at a given time and place ; in other words, the norms may
be combined with a recommended income necessary for attaining
them.
The third concept to be distinguished is the standard of living
defined as that consumption which the individual or family considers
to be necessary and proper. The consumption considered proper
may be either that which is habitual and customary to the individual
or family or that to which they aspire and for which they strive
as their legitimate and reasonably attainable goal. Actual consumption, i.e. the level or content of living, may deviate during short
periods from the standard defined in this sense for several reasons,
e.g. loss of income as a result of unemployment, etc. On the other
hand, a windfall in income may temporarily permit actual consumption to be in excess of standards formerly considered necessary.
In general, such divergence between actual consumption and the
standard considered proper, cannot be retained indefinitely: the
family with a permanently reduced income gradually and painfully
readjusts its standard to that level. On the other hand, the family
with an increased income, given time for adjustment, soon finds
that its standard exhausts the income.

— 14 —
The distinction of terms outlined above is important in so far
as it helps to state more clearly the problem of policy in relation
to the worker's standard of living. In the terms given above, the
deficiencies in the worker's consumption differ in character and
urgency. Such deficiencies may be due either (1) to a discrepancy
between what he is accustomed to or considers normal and proper
and the norms set up as an objective minimum or optimum; or (2)
to a discrepancy between his actual consumption and such norms;
or (3) to a discrepancy between what he can afford (actual consumption) and what he regards as normal and proper. But deficiencies
differ in character and urgency. Where the level of living is very
low, there is an absolute quantitative lack in simplest terms of
food, shelter and clothing. Here the problem of raising the level
more nearly in accordance with minimum norms or standards
considered proper is the problem of relieving dire poverty. There
are other cases when the group is at the subsistence level, or even
at the decency level. To raise the worker's standard of living under
these conditions might thus be interpreted as adopting measures
which would harmonise the level of living with the standard
considered proper or bring the latter into harmony as much as
feasible with recognised and established norms, or to improve
the content of living in general.
Such a statement of the problem comes within the boundaries
of the economic and social aspects of welfare and presents a useful
point of departure. It suggests also an answer to the question of
freedom of choice which lies at the root of some current discussions of the subject. The discussion of this problem is reserved
for Chapter II.
V. — Income, Prices and Consumption

In all considerations of the worker's standard of living, the
interrelation of income and prices is of key importance. The real
income of the worker depends on this interrelationship. In any
given situation, the goods and services entering into the worker's
standard of living are determined to some extent by habits, tastes
and ability to use income effectively, but to a much greater extent
by the size and stability of the income itself and its relation to
the prices of commodities and services.
The importance of income as a determinant of the standard of
living arises not only from the fact that it is the basis of the worker's
capacity to consume, but also from the fact that it tends to fix

— 15 —
habits of consumption. It has long been pointed out t h a t
" when any particular income has been received by a class for a
sufficient period to have become customary, the standard needs
the income ", and " the standard t h a t exhausts the income soon
becomes customary and conventional, and the goods bought
are regarded as necessaries of life " 1 . This works both ways.
On the one hand, it creates a feeling of distress when income is
reduced in such a way as to make the maintenance of the customary
level of well-being impossible. On the other hand, it forms at
times psychological obstacles in the way of changing habits of
consumption even after income is increased.
It was pointed out in a previous Office Report that changes in
money incomes on the one hand and in prices on the other may
have different effects on consumption, depending on changes in
the relative prices of different commodities and on elasticities of
demand. This subject is treated more fully in Chapter II. Here
it will suffice to quote from that Report as follows:
Purchasing power in general depends on the relation of money income
to the prices of commodities; purchasing power will be- increased if
income increases or if prices fall, and will be decreased if income decreases
or if prices rise. Thus, in regard to the consumer, prices and incomes
are in some respects only two sides of the same phenomenon, that of
purchasing power.
But there are important differences in the effects which changes in
prices and incomes may have on consumption. When the consumer's
purchasing power increases either through a rise in income or a fall in
prices, he tends to expand his consumption all along »the line. If all
prices increase or decrease proportionately, the effect on purchasing
power will be exactly the same as if incomes decreased or increased
in the same ratio. But as general price movements are accompanied
by changes in the relative prices of different commodities, the effect on
the distribution of the new purchasing power is likely to be different
from the effects which a corresponding change in money incomes would
have 2 .
Besides the size of money income, its distribution of incomes and
their stability have important effects on standards of living.
Inequalities of income affect in large measure the character and
direction of economic activities and thus employment opportunities
and workers' earnings. Fluctuations in economic activities and
therefore in workers' earnings have long been recognised as one

1

2

A.

L.

B O W L E Y , pp.

cit.

INTERNATIONAL LABOUR O F F I C E : Workers' Nutrition
Studies and R e p o r t s , Series B, No. 23, p p . 107-8.

and Social

Policy.

— 16 —
of the most serious factors of poverty and of low living standards
in most countries \

VI. — Population Growth and Family Type

Two other factors of importance in determining the worker's
standard of living which are interrelated are the rate of population increase and the composition of the family with regard to
size, sex and age, or family type. In terms of current analysis,
there is said to be for each country during a given period an
" optimum " population and an optimum rate of population
growth, that is, a rate which maximises economic welfare—roughly,
real income per head2. Any departure from this optimum in either
direction reduces the standard of living3.
1

An idea of the extent of fluctuations in labour income is suggested by the
figures given below:
LABOUR INCOME IN THE UNITED STATES AND IN GERMANY, 1 9 2 9 - 1 9 3 2
United States
(Millions of U . S . dollars)
Agriculture
and
mining

Year

1929 . . .
1932 . . .

2,952
1,200

Manufacturing,
building
and
electrical
industries
18,135
8,034

Transport,
communications and
commerce

Public
administration

17,138
11,229

4,984
5,277

Public and
private
service
and
miscellaneous
9,584
5,792

Germany
(Millions of RM.)
. Year

Wages

Salaries

1929 . . .
1932 . . .

23,339
11,403

11,044
7,379

Civil Service
salaries
6,902
5,342

Other
labour income
1,760
1,587

INDEX NUMBERS OP EMPLOYMENT IN VARIOUS COUNTRIES, 1 9 2 9 - 1 9 3 2
Year
1929
1932

. .
. .

Great
Britain

Germany

France

100
91

100
63

100
82

100
61

Czechoslovakia

Italy

Japan

Poland

100
'83

100
78

100
82

100
63

2
It should be noted that the idea of an " optimum population ", except in
a few cases, is still largely theoretical in character. The criteria on the basis
of which a country may be said to be under-populated or over-populated have
been studied recently, but further investigations will be necessary before it
will be possible to apply these criteria in concrete cases scientifically. See
Imre PERENCZI, The Synthetic Optimum of Population, International Institute
of Intellectual Co-operation, Paris, 1938.
3
Many parts of agricultural India, for instance, are even more densely
populated than industrial Britain; under the economic and social conditions
existing in India, this results in serious pressure on subsistence and accounts
in a large measure for low levels of living. See H. B. BUTLER: Problems of
Industry in the East, 1938.

— 17 —
Size and composition of family are important in so far as they
affect expenditure liability which has to be met out of family
income. Differences in the size and composition of the family
give rise also to important differences in the total income and in
the amount and kind of goods and services required. The effects
of family type on the kind and character of family expenditures
are obvious, e.g. on the need for and consumption of milk, clothing,
etc. But these effects are of comparatively small range. In
general, as various budgetary studies would seem to indicate, the
variations in the size and composition of the family affect more the
quantities than the kinds of goods and services consumed.
VII. — Social Services, Free Services, and Working Conditions

In most countries to-day, the worker's standard of living depends
not only on what he buys in the market but on goods and services
supplied partly or entirely, directly or indirectly by the community.
In other words, part of the worker's income comes in the form of
social services1 and of so-called " free " or " social income ".
Free schools, free hospitals and medical clinics, public parks and
concerts, low rent housing, good roads, etc., are only a few of
the goods and services which have entered increasingly into the
worker's standard of living as a result of changes in social organisation and social policy.
Among the free services which affect the worker's living, two
call for special emphasis, namely, those relating to health and
schooling. Though the subject has been but little investigated2, the
special studies so far made would indicate that there is a definite
correlation between ill-health and income, especially on low income
levels.
1
The term " social services " is understood to include only those services
which satisfy both the following criteria:
(a) that their object should be to cover one of the following risks : industrial
accidents, occupational diseases, sickness, maternity, invalidity, old
age, involuntary unemployment and family responsibilities;
(6) that they should be established on behalf of classes of the population
consisting mainly of wage earners and persons of small means working
on their own account.
See International Survey of Social Services, International Labour Office,
Studies and Reports, Series M, No. 11, 1933.
2
Interesting remarks on the fact that and the reasons why " writings
dealing with the history of man in its various aspects contain (for many
centuries) almost no reference to health or disease except occasionally in very
detailed biographies or when an epidemic carried off thousands of people "
may be found in J. H. S. BOSSARD: Problems of Social Well-being, 1927, p. 4.

S.L.

2

— 18 —
Elementary schooling and the provisions for vocational training
affect the worker's capacity to earn his livelihood. As is well
known, public authorities in recent years have not stopped at this
point, but have provided more and more free facilities for the
acquisition of general knowledge, recreation, and aesthetic enjoyment. These are definitely elements in the worker's standard of
living depending on social policy.
The worker's standard of living is further affected by the way
in which he and his family gain their income. Most economists
have insisted on the " pain " or " disutility " of work as a negative
element to be reckoned, as it were, against the " utility " of wagea
as a positive element of income. But the " disutility " of work
depends in large measure on the physical and moral conditionsunder which work is carried on (conditions of air and light, safety
provisions, methods of management, etc.), on the duration of
working periods, on the relative level of remuneration 1 , etc.
Particular importance attaches to two factors, namely, duration
of Work and regularity of employment. The former is important
in so far as it determines the range of time within which consumption
and leisure can be enjoyed; the latter affects the stability of the
worker's earnings and his capacity to maintain and plan consumption on the customary level.
The close interrelation between the various elements indicated
above suggests the necessity of extending the term " standard of
living " so as to cover working conditions, social services and
" free income ". In other words, the worker's standard of living
includes the consumption of goods and services purchased, free
and social services and conditions of work. The worker's living
standard is thus affected, first, by the size of the money income
or wages, its relation to prices and the way in which it is spent;
second, by the conditions under which the income is earned; and
third, by those social provisions which result in protection against
risks or in " free income " which affect most closely health and
physical well-being.
VIII. — Group and National Differences

Apart from small variations, individuals and families within thesame social-economic group or class tend to have, in a given period
1
This is important independently of the absolute amount of remuneration.
It is obvious that a worker may feel dissatisfied if he earns less than another
who does the same kind and quality of work.

— 19 —
of time, a certain average set of commodities which is fairly representative of the consumption of the group. There are, of course,
articles or services the expenditure on which may be different
in families which otherwise are much alike, but the spread
or dispersion of the consumption (or expenditures) around the
average is not very wide ; and the spread will be smaller, the smaller
the income. It is not necessary to examine the various influences
which make for similarity of tastes, habits and consumption within
groups that are economically and socially homogeneous 1. What is
important here is to emphasise the fact that standards of living are
shaped by group influences and that this group character affects
the problem of their evaluation and of their improvement.
It is thus inaccurate to speak of." the standard of living " of a
country as a whole. There is a number of living standards of
different groups in a community at anytime which cannot be fully
represented by an average. On the other hand, it is undesirable,
for purposes of analysis, to classify a population into a large
number of groups on the basis of small variations in income and
expenditures. In general, as the study of family budgets indicates,
there are several large social-economic groups in most countries
each of which has a more or less common character with regard
to income and its standard of living. Within each of these
groups there are, of course, variations in income and in ways
of spending it. But these differences are small, are due to chance
variations within the group, and do not, therefore, destroy the
essential homogeneity of the group. As to groups within the
working population itself, recent family budget studies have distinguished between skilled and unskilled workers, clerical and
salaried professional workers, urban and rural workers, etc.
While standards of living differ between groups of one country,
they also show considerable variations between countries. The
income and expenditures of similar social-economic groups in difi " It would, at any rate, in many classes be found that there was no great
divergence among the majority of the class. Not only is originality rare and
the influence of custom universal, but convention is very strong. In some cases
any great divergence from the mode in residence or of living and clothing
would make a man unacceptable to his employer, in others to his fellowworkmen. In many cases there is no alternative as to housing, and little
variety in the goods sold at accessible shops. . . . In many classes expenditure
on house-rent, food and clothes is practically determined by the custom of
the class; but expenditure on insurance, amusement, tobacco, drink, subscriptions, books, and other things not of primary necessity may vary greatly,
when the income leaves a margin for such things. It is to a great extent in
the variation of free expenditure and in the employment of leisure time that
individuality has play." BOWLEY, op. cit., pp. 151-152.

— 20 —

-

-

ferent countries show wide differences. While these differences can
be explained by many factors such as climate, racial qualities, etc.,
they are also in some measure the result of psychological and
cultural influences. In these national differences of consumption
habits one may see the interconnection of the ideas of " general
well-being " and economic well-being referred to earlier in this
chapter.
The group and national differences in standards of living are
significant in so far as there is a tendency to compare standards
and to use such comparisons for purposes of policy. Whether
such comparisons can properly be made, and by what methods
is considered in Chapter II.

IX. — Summary and Definition

The discussion of the previous sections may now be briefly summarised in a definition which may serve as a basis for considering
the relation of social and economic policy to standards of living.
Briefly, it was seen that while standards of living contain a subjective element—namely, the attitude towards economic goods
and life in general—it is not within the scope of this Report to
consider this element.
In view of this, the standard of living is here described in objective
terms as consisting of three main elements: (1) the level of consumption, or the composite of goods and services of a specific
quantity and quality consumed by an individual, family or group
within a given period; (2) social services and free services, particularly those which relate to health, education and recreation;
and (3) working conditions which affect not only the worker's
health and earning capacity but also the size and regularity of his
income.
On the basis of this definition and of the distinction between
levels, norms and standards, the problem of raising living standards
may be more precisely described. It is in the first place a problem
of doing away with poverty, and secondly of improving the content
of living of all workers with regard to consumption, social and
free services, and working conditions.

CHAPTER II .
DESCRIPTION AND EVALUATION OF WORKERSSTANDARDS OF LIVING

In the preceding chapter, the main factors which determine the
workers' standard of living have been examined and their interrelations discussed. In the present chapter attention is directed
to the materials available concerning these factors and to the
methods of describing and evaluating them.
As pointed out in Chapter I, the number of elements which may
be studied as part of the standard of living is very large 1. To cover
this aspect of the subject fully would prove an enormous task;
some information concerning the various determinants of workers'
standards exists for all countries, though much of this information is
incomplete, out of date or otherwise inadequate. In this preliminary
Report, the discussion is, therefore, limited to a consideration only
of the most important elements in the workers' standard of living,
and all that is attempted here is to indicate the chief sources of
information and how they can be used for appraising the standards
of living of workers and their families.
In the preceding chapter it was shown that the most significant
element of the worker's standard of living is his income or rather
the income available for the persons dependent on a common
income. The family income is perhaps the simplest, though not
the only determinant, of its standard of living. The material
available on incomes of workers and their families is therefore
1
The tendency in recent years has been to enlarge the scope of family
living studies so as to get as clear a picture as possible of the economic and
social life of the mass of the people. Some investigators have aimed at
including in their studies such elements as per capila income, the extent of
employment, the amount of illness and the mortality rates, the extent of
savings, loans and capital changes; the conditions of work, the use of leisure,
general cultural conditions, etc. The problems arising in the making of family
budget enquiries have been studied by the Office in a Report (Studies and
Reports, Series N, No. 7) and a further Report on the subject is in preparation.

— 22 —

treated first and the various sources of information are reviewed.
The next topic treated is that of how the income is spent. Satisfactory standards of living necessitate as a minimum adequate food,
shelter and clothing. Family budget enquiries are the principal
source of information on these matters.
There remain, however, various other elements which cannot
be overlooked in any discussion of worker's standards. The worker's
health, his education and amount and use of leisure are all factors
which determine in some measure his standard of living. These
are, of course, to a large extent dependent on income, but they are
fields in which social policy and community effort, rather than
individual choice, play a large part and they cannot always be
measured statistically.
These matters are therefore treated
separately.
A considerable amount of information on the statistical and other
aspects of the various elements in the worker's standard of living
has already been published by the International Labour Office in
recent years *. In the following pages it has been considered
sufficient to call attention to this material, and to give illustrative
examples of the way it can be used to throw light on the subject.
In addition, however, it has been thought desirable to make a more
detailed study of certain countries, in order to survey the data
available and to illustrate the methods of describing the living
standards of the workers in these countries. The four countries
selected are the United States of America, a large and a highly
industrialised country; India, a large area of over 360 million
inhabitants where special problems and conditions arise; Japan,
a modern industrial country of the Far East; and Poland, a
large area in Eastern Europe predominantly agricultural. In
presenting the available data on these countries, it is not intended
to make any comparisons between them. Thè only purpose is
to indicate how living standards of various countries may be
described so as to provide a basis for more precise judgments in
matters of policy.
This chapter will thus summarise existing materials on income,
expenditure, nutrition, housing, etc., and will discuss the problem
of methods of evaluation. A description of living standards in the
four countries mentioned will be presented in Chapter III.
1
See, e.g., the Y ear-Book of Labour Statistics, the International Survey of
Social Services, the statistical part of the International Labour Review, the
Report on Workers' Nutrition and Social Policy, etc.

— 23 —

I. — Income
Income, as pointed out in Chapter I, is the most important
single determinant of the worker's standard of living x. It is the
source from which in the long run all the purchased elements of the
standard of living must be paid for 2. The amounts of incomes
actually received thus furnish an upper limit to the standard of
living in so far as it is dependent upon purchased goods and services.
Also, the higher the income, the greater the freedom for choosing the
goods and services which make up the particular standard of living.
For the purpose of describing and measuring standards of living,
family income rather than individual income is the significant item.
A standard of living is a family matter 3. In order to obtain the
requisite information for determining the standard of living of the
family, not only must data on family incomes be available but also
data on the composition of the families, since, other things being
equal, large families require larger incomes to maintain equal
standards. The income per person or per unit of the family,
therefore, is the desideratum.
A.

DISTRIBUTION OF FAMILY INCOMES

(1) Census Data
No country has published statistical data showing for the entire
population the distribution of families according to size of income
and size of family, or a distribution of the population according
to size of income of the families.
1
Income is here defined broadly to include not only money receipts, but
also the imputed value of receipts in kind. Even the most advanced countries
include large sections of the population whose income to an important degree
consists in the consumption of their own products. Furthermore, payments
in kind are a more or less important form of remuneration in many occupations—e.g. agricultural labourers, domestic servants, etc. However, in some
cases, owing to the statistical difficulties involved in attempting to estimate
the money value of receipts in kind, the latter are disregarded and the amount
of money receipts taken as the size of the income.
2
In a particular year the goods and services purchased may be paid for by
changes in capital, by borrowing or other adjustment, and similarly in some
years savings may be made out of income; in these respects, the income of a
particular year may not correspond exactly to the standard of living as
indicated by the goods and services purchased in that year. So far as goods
and services furnished free by the community or by the State are concerned,
the cost of these may be considered to be defrayed by taxation, but it remains
true that so far as individual families are concerned, the amounts paid in
taxes do not correspond with the value of the free goods and services which
may be enjoyed by the particular family.
3
Except in cases where an individual lives alone and has no dependants.
Such cases are a small part of the population of a country.

— 24 —

A close approach to the latter is found in a study based upon the
Swedish Census of 19301. The Swedish Occupation Census of 1930
gives in addition to the number of persons gainfully occupied in each
occupation the total number of persons dependent upon each
occupation, and the total income received. From these figures the
average income per person gainfully occupied in each occupation
and the average income per adult unit (children under 15 counted
as half an adult) of all persons dependent upon the occupation,
including both the gainfully occupied and their dependants, has
been calculated. The entire population is then distributed according
to the size of income, not of income received by each individual
family, but of the average income per adult unit received by all
families in each occupation—that is, all persons dependent upon the
particular occupation are classified in the income group in which
the average income per adult unit for that occupation falls. The
result is a rough approximation to a distribution of population
according to family income 2.
(2) Estimates
An estimate of the distribution of the population according to
per capita family income has been made for Great Britain by Sir
John Boyd Orr in his study on Food Health, and Income 3, using
income-tax data, data on earnings according to occupation and
data on the number of wage earners and dependants per family.
The technique of such an estimate is discussed in detail by Colin
Clark 4, who describes the results obtained in a sample of 23,000
families enumerated in the 1931 census, with special reference
to the number of wage earners and dependants per family.
An estimate of the distribution of the population according to
1
SWEDEN: FINANSDEPARTMENTET. Statens Offenlliga Utredningar 1936:
IS. Undersökningar rörande Del Samlade Skaltetrycket i Sverige och Ullandet
av Inom Finansdepartmentel Tillkallade Sakkuniga. Stockholm, 1936,
pp. 34-41. Reference to the Census: SWEDEN: Sveriges ofßciella Statistik.
Folkmängden och dess Förändringar. Folkräkningen den 31 December 1930, av
Statistika Cenlralbyran. 111. Folkmängden efter Yrke, Inkomst och Förmögenhel: 1 Avd.
2
For two other countries—Australia and New Zealand—the census of occupations includes data on incomes or income class in conjunction with occupation
and industrial status, i.e. whether wage earners, employers, or persons working
independently. Since in the census lists the data for each family are usually
placed together, tabulation of the material by families should be possible,
giving the distribution of families according to income and size and the distribution of the population according to size of family income, separately for the
wage earners and other parts of the population. Such tabulations, however,
have not been published.
3
Food, Health and Income, Macmillan, London, 1936, p. 21.
4
Colin CLARK: National Income and Outlay, London, 1937, p. 11 ff.

— 25 —

family income in the United States is contained in a publication of
the Brookings Institution 1.
(3)

Materials for Estimates

(a) Partial Data on Family Incomes
Partial data on family incomes, or family incomes for samples
of a population, are to be found in most family budget enquiries2
and in certain other types of study. So far as the former are
concerned, family budget enquiries are for the most part limited
to relatively small samples, usually less than 1,000 or 2,000 families.
They cover wage earners' families in practically all cases, but in
many cases the families of salaried employees are also included.
In the United States a study of wage earners' family budgets
covers a sample of approximately 16,000 families 3 ; also for a
preliminary sample for a general study of consumer purchases
covering approximately 315,000 families drawn from six general
regions of the country, from all income ranges, from large, middlesized and small cities, villages and farms, data were obtained on
earnings and income in conjunction with family composition 4.
A special sample census in Sweden covering one-fiftieth of
the entire population of the country gives family incomes, size of
household, number of children under 15, together with certain
data on housing for nearly 32,000 households, with over 120,000
persons in 100 rural communes 5.
(b) Data on Personal Incomes
A large mass of data is available on the distribution of personal
incomes, that is, on the number of persons in a country or region
whose income per year (or other time unit) is of a specified amount
or comprised between specified limits. One source is the censuses
1
Maurice L E V E N , Harold G. MOULTON and Clark W A R B U R T O N : America's
Capacity to Consume, Washington, D.C., 1934; see especially Chapter V and
Appendices.
2
F o r bibliography, see W I L L I A M S , F , M., and ZIMMERMAN, Carle C. : Studies
of Family Living in the United States and other Countries : An Analysis of
Material and Method, U.S. Dept. Agrie. Miscellaneous Publications. No. 223,
W a s h i n g t o n , D.C., 1935.
3
" S t u d y of Money Disbursements of Families of W a g e E a r n e r s a n d
Clerical W o r k e r s " , Monthly Labor Review.
March, 1936, et seq.
4
Study of Consumer Purchases, in progress by the U . S . Bureau of L a b o r
Statistics in co-operation with the U.S. Bureau of Home Economics.
6
S W E D E N : Sveriges officiella Statistik.
Folkmängden och dess
Förandringar.
Särskilde Folkräkningen
1935-36, av Statistika Centralbyrän.
III.
Partiella
Folkräkningen
i mars 1936 : Specialundersökning
av Bosladsforhàllandena
i
100
Landskommuner.

— 26 —

of personal incomes, another the tax returns, a third the statistics
of wages and earnings, etc. The value of these sources for throwing
light upon distribution of family incomes is difficult to appraise:
• they must be used in conjunction with estimates of the frequency
of more than one income to a family, and for purposes of determining
the distribution of incomes per capita or per consumption unit in
conjunction also with data on size of family.
(i) Censuses of personal incomes
Censuses of incomes received by individuals are few. Such
censuses have been conducted by the Commonwealth of Australia,
1915 and 1933; New Zealand, 1926 and 1936; Denmark, 1916;
State of Pennsylvania, 1934; State of Wisconsin, 1934. Without
going into details, it may be noted that each of these censuses
presents important peculiarities as regards the definition of income,
the reliability of its results, the classification in which they are
presented, the time unit for which income is measured, and so on.
Tabulations of data from the Australian Census of 1933 give the
distributions of breadwinners and of others (" includes persons
engaged in home duties . . . scholars, and other dependants ")
by income class and sex 1, and distributions of employees, classified
according to whether they were full-time or part-time, employed
or unemployed, by income classes and conjugal condition. These
same persons are also classified according to conjugal condition
and the number of their dependent children, but this latter information is not related to their income categories 2. Distributions of
breadwinners' incomes are also given separately according to
whether or not the breadwinners were working on their own
account, etc. 3.
In the New Zealand population census of 1926 data were collected
on the income category into which the personal incomes of all
persons over 16 years of age fell. The data are presented by
geographical district, occupation, and other factors 4.
1

AUSTRALIA: MINISTRY

OF S T A T E FOR H O M E

AND T E R R I T O R I E S , B U R E A U

OF CENSUS AND STATISTICS: Census of the Commonwealth
1933.
Bulletin No. 15.
2

A U S T R A L I A : M I N I S T R Y OF S T A T E

FOR H O M E AND

OF CENSUS AND STATISTICS. Census of the Commonwealth
1933.
P a r t X X V , Employee Group.
3

AUSTRALIA:

COMMONWEALTH

BUREAU

OF

of Australia,

30 J u n e

TERRITORIES,

BUREAU

of Australia,

CENSUS

AND

30 J u n e

STATISTICS,

CANBERRA: Official Y ear-Book of the Commonwealth of Australia.
No. 30,
1937, p. 353.
4
DOMINION OF N E W ZEALAND : Population Census 1926. Vol. X I , Incomes.
Wellington, 1930.

— 27 —

Mention may also be made here of a special census of population
and unemployment in the State of Michigan, U.S.A., in 1935, which
gives the distribution of 60,000 employed and unemployed workers
according to income, classified by industry and size of community 1.
(ii)

Income-tax returns

Income-tax returns give a comprehensive picture of personal
incomes above the tax exemption limits which vary from country
to country. They are, however, of relatively little value for
purposes of estimating the distribution of family incomes of
wage earners, and of other low-income groups, since such incomes
fall in large part in the lowest brackets of the income-tax schedules
or are below the limit of exemption; the statistics, as published,
generally do not show the number of persons in the family. The
problem of utilising the income-tax returns depends therefore upon
the possibility of obtaining data from which to estimate incomes
below the tax limits. The various types of deductions for children
and dependants, and the fact that also in some cases figures show,
not the incomes, but the tax paid (from which incomes have to be
calculated) are further difficulties.
Among countries which publish data on incomes from income-tax
returns may be mentioned the United States, Great Britain 2 ,
France 3, Poland, Finland, Estonia 4, etc., no attempt being made
here to give a complete list or even to mention all the countries for
which data presenting some features of special interest are available.
In some countries, for example Germany, data are available, not
only for the incomes of persons subject to tax, but also for those of
persons exempt from tax (owing to deductions), and the two
1

U.S.A.:

STATE

OF MICHIGAN:

STATE

EMERGENCY

WELFARE

RELIEF

Co MMISSION: Michigan Census of Population and Unemployment, 1935. Total
Income during 1934 of Gainful Workers. Lansing, Michigan, 1937.
2
For Great Britain, Clark prepared a distribution of personal incomes in
1929, adjusting income-tax and other data to this purpose. CLARK, Colin: op.
cji., p. 109.
3
For France data are presented on the income distribution of persons with
incomes subject to the general income tax, with the number in each income
class who are single or divorced without children and the number married
without children. The income distributions of persons with taxable wages,
etc., are also given.

FRANCE: PRÉSIDENCE DU CONSEIL, DIRECTION DE LA

STATISTIQUE GÉNÉRALE ET DE LA DOCUMENTATION: Annuaire
m

Statistique.

52 e Volume, 1936. Paris, 1937, p. 243.
4
The Estonian Year-Book gives a distribution of persons paying income tax
according to the amount of tax paid, from which the incomes can be re-computed.
These data are shown separately for independents and wage earners and
separately for agriculture and for industry, commerce and transportation.
i ESTONIA: RIIGI STATISTIKA

KESKBÜROO.

rétrospectif. Tallinn, 1931, p. 285.

Estonie de 1920-1930, Résumé

— 28 —
groups together give a general picture of the distribution of personal
incomes \
(iii) Statistics of wage rates and earnings
A vast mass of data is available on wage rates and earnings.
For the most part these data are in the form of averages per
person and are not helpful in determining distributions. Much
of this material on wages is in the form of hourly, daily, or weekly
wage rates or earnings and offers therefore merely one element in
the data needed to obtain annual incomes. For that purpose
estimates of the amount and distribution of unemployment,
sickness, etc., must be available.
The principal sources include the Year-Book of Labour Statistics
published by the International Labour Office, which gives a
general summary presentation of the principal statistics available
in the several countries. For each principal country there is a
large list of sources to be consulted, from which it is difficult to
make a selection for mention here.
(iv) Other sources
Social insurance statistics often contain data of the incomes or
earnings of the insured.
For Czechoslovakia the distribution of workers insured against
sickness according to their basic wages is available. A distribution of non-manual workers in private industry classified
according to their annual salary is also available from the statistics
of their obligatory retirement insurance 2. For Germany, data
are obtained as a result of the contributions to the national systems
of invalidity and old-age insurance which are compulsory practically
for all wage earners, including agricultural labourers and domestic
servants. Data for France are derived from the returns of the
General Social Insurance scheme, which covers all non-agricultural
wage earners.
For Norway data of the distribution by income classes in 1934
and 1935 is available for members of the official sickness funds 3.
The distribution of non-agricultural workpeople insured for
1
G E R M A N Y : Statistisches Jahrbuch für das Deutsche Reich,
S W I T Z E R L A N D : F o r incomes in Basle canton, see Statistisches
Kantons Basel-Stadt, 1935, p p . 286-287.
2

CZECHOSLOVAKIA: S T A T N I U R A D S T A T I S T I C K Y .

République
3

tchécoslovaque.

NORWAY:

Argang, 1937.

DEPT.

Annuaire

1937, p . 535.
Jahrbuch des
Statistique

de la

P r a g u e , 1937, p p . 218, 222.

FOR SOCIALE

Oslo, 1937, p . 143.

SAKER:

Sta.tistisk

Ârbog

for Norge

56.

— 29 —
retiring pensions, according to weekly earnings, is given by sex
and district for Poland in 1934 1.
Special censuses or studies are sometimes a source of valuable
information- on incomes and the distribution of incomes. In
this connection may be mentioned the special census of earnings of
male wage earners in cities of 10,000 population and over in three
provinces in Western Canada. This report gives the distribution
of male wage-earning heads of families according to earnings groups,
showing for each earnings group the total number of persons in
the family and the number of children by age groups 2.
(c) Data on Number of Income Receivers per Family and Size of
Family
For the sake of completeness brief mention should be made of
statistics from which the number of income receivers per family
and the size of family may be obtained or estimated.
So far as size of family is concerned, statistics of family size are
available for the entire population in a number of countries. For
example, censuses of the United States of America, Great Britain,
and many other countries, give data showing the distribution of
families (in the sense usually of the household of persons living
together under a common roof and eating at a common table) by
number of persons in the family. For the most part, however,
these data are not segregated so as to show separately the size of
the family of the wage-earning population or the number of earners
and dependants by family. In this respect, therefore, data on
size of the family from special studies, e.g. family budget studies,
may be utilised to indicate how far the distribution of families in
the wage-earning class agrees with that in the general population.
The question of the number of income receivers in the family is
one on which it is difficult to throw light except by special studies.
No comprehensive material on this subject is available. One
special study is the sample of 23,000 schedules of wage earners'
families selected from the British Census of 1931 analysed to show
1

POLAND: CHIEF BUREAU OF STATISTICS. Concise Statistical Year-Book of

Poland, 1937.
2

Warsaw, 1937, p. 235.

CANADA: DOMINION BUREAU OF STATISTICS: Earnings

of

Wage-Earner

Heads by Tenure and Size of Family for Cities of 10,000 Population and over.
Census of Prairie Provinces, 1936. Bulletin No. XXXII, Ottawa, 1938.
Though these data partake of the advantage of a census, it scarcely needs to
be pointed out that for the purposes of estimating the distribution of family
incomes the restriction to cities of three provinces, the fact that only earnings
and not incomes are taken into account and that the earnings of the head only
and not those of other members of the family are included, constitutes serious
limitations.

— 30 —
distribution of wage earners per family 1. Material on this point
will be available for the United States in a sample of some 315,000
families in the Study of Consumer Purchases as well as for the sample
of 16,000 wage earners' families in the Study of Money Disbursements of Wage Earners and Clerical Workers.
Practically all family budget studies contain data on sources of
family income, including earnings of different members of the
family, and thus throwing light upon the number of wage earners
per family and upon the relative importance of their contributions.
The studies in which the distribution of family incomes has been
estimated generally throw some light upon the sources of data
utilised for estimating the number of wage earners per family as
well as on the methods by which this information is used in the
calculation of family incomes.

B.

AVERAGE FAMILY INCOME

If data on the distribution of family income is not available, a
more modest task is to determine the average family income of the
population or of the wage-earning part of the population. An
approximation to the average income per capita of the population
or of the wage-earning part of the population can be found b y
dividing the total national income by the population, or by dividing
the income received by the wage-earning part of the population
by the number of persons in wage-earning families. With regard
to the average per capita income of the population, attempts have
been made to estimate this for a number of countries. One of
the most recent of these presents estimates of average per capita
income for 28 countries 2.
Estimates of the total income received by wage earners are
difficult to make, especially the income from other sources than
wages. The division of the total wages received by wage earners
by the total number of persons or, if possible, the number of
consumption units, in wage earners' families, would fall below
the true figure. The proportion of such income other than wages
received by wage earners from such other sources as shown by
numerous family budget studies is not great.
Such estimates of average incomes per capita, etc., while available
for a considerable number of countries, are not sufficiently accurate
1

2

CLARK, Colin: Op. cit.

CLARK, Colin: Internationaler Vergleich der Volkseinkommen,
schaftliches Archiv, Vol. 47, January 1938, pp. 51-76.

Wellwirt-

— 31 —
and are likely to be misinterpreted in any attempt to compare
standards of living in different countries.

II. — Expenditures

While income may be deemed the most important determinant
of the standard of living, a study of how the income is spent shows
the concrete elements making up the standard so far as it depends
on income. Data on family expenditures are available in family
budget enquiries which give information on goods and services
purchased by selected families of varying size and composition and
of varying income levels.
During the last ten years family living studies have been
undertaken by some 25 countries, including all important industrial
countries. These studies are summarised in the Y ear-Book of
Labour Statistics and in the International Labour Review published
by the International Labour Office. Most of these enquiries relate
to families of industrial workers living in urban and industrial
centres. In some cases the studies cover also families of agricultural
workers. In addition, a number of studies include families of
salaried employees and civil servants.
These studies for the most part cover a range of income levels, so
that a comparison of the expenditures of families at different
income levels affords a clue to dependence of living standards upon
income. The full analysis of proportions spent for different major
items—food, shelter, clothing and miscellaneous items—throws
light upon variations in standards of living in these respects. The
various studies include, in addition to the direction of expenditure,
details on the composition of the family, the occupation of the
head of the household, sources of income, quantities of goods and
services consumed, etc. So far as expenditure is concerned, data
are analysed according to income or, more usually, expenditure
classes, either expenditure per family or expenditure per consumption unit 1 . With particular reference to food and housing for which
standards or norms of adequacy can be more definitely established,
a more detailed analysis of the available data is presented
below.

1

Por a discussion of " consumption unit " see INTERNATIONAL LABOUR
OFFICE: Workers' Nutrition and Social Policy, 1936, Chapter III.

— 32 —

A.

FOOD AND NUTRITION

The physiological requirements which food must satisfy include
a sufficient provision of energy or calories depending upon the needs
of each individual with reference to his age, sex and physical
activity, appropriate amounts of proteins, carbohydrates and fats
and necessary minerals and vitamins. If the allowance for food
in a family budget is insufficient, one or more of the necessary
elements for adequate nutrition may be entirely lacking or be
present in insufficient quantities. Apart from satisfying physiological
requirements, food expenditures may provide for food in excess of
the minimum necessary for optimum health l, or may include
expenditures for special foods which gratify taste and provide
variety to the diet, or may include margins for waste. The only
points which will be discussed in detail here are those which relate
to the provision of food to fulfil the optimum requirements, that is,
whether the food budget is adequate to satisfy established norms
for protein requirements, energy and for the protective foods.
Data on food consumption are of two types: (1) general statistics
of food production, imports and exports, and of food consumption
for entire areas, and (2) specific studies of food consumption of
individual families. These two sources may be utilised in combination with data on income distribution to form a basis for estimates
of food consumption in different classes of the population throughout
a country. This combination method is used, for example, in a study
made by Sir John Orr 2 of nutrition in Great Britain, in which the
conclusion is reached that the food consumption of about half the
families of Great Britain falls short of the optimum, either because
of insufficiency in amount of energy or insufficiency of one or more
of the necessary minerals or- vitamins.
(1)

General Statistics of Food Consumption

Food available per capita. — Data on national consumption of
various foodstuffs, per capita, in several European countries,
the British Dominions and the U.S.A. are contained in the League
1
Optimum health—" the standard of health adopted is the physiological
or ideal, viz., a state of well-being such that no improvement can be effected
by a change in the diet ". ORR, Sir John Boyd: Food, Health and Income,
London, 1936, p. 12.
2
Ibid.

— 33 —

of Nations' Nutrition reports \ Various estimates for a larger
number of countries are also summarised in the study on Workers''
Nutrition and Social Policy published by the International Labour
Office 2. Data of national consumption, even for foodstuffs for
which accurate estimates can be made, are satisfactory only
for indicating very wide divergencies in diets, since no means is
provided for differentiating between the consumption of the various
groups of the population. Furthermore, there are often considerable
gaps in the data necessary to estimate the amount of food available
for human consumption and in particular to estimate the proportion
of this amount available for consumption by the working classes.
Only where the assumption can reasonably be made that all classes
of the population are on a fairly uniform basis so far as food supply
is concerned (as was the case in many countries during the world
war, when the entire population was rationed with respectto certain foodstuffs), or where figures are available showing the probable
differences in consumption by different groups of the population
(as used by Sir John Orr in his study cited above) can figures
derived from this source be applied to the problem of determining
the level of food consumption of the working classes.
(2) Family Budget Studies
Food consumption per family or per consumption unit. — Studies
of actual food consumption in families of wage earners are available
(though in many cases out of date) as part of family budget studies
made in many of the principal countries of the world. These studies
are in most cases directed to an analysis of incomes and
expenditures, including expenditures for food, but in the majority
of cases, especially in studies recently made, information is available
not only with regard to money expenditures on food, but also with
reference to the physical quantities of specific foodstuffs consumed.
Most, if not all, of these studies include data on size of family, and
for the most part the data on food consumption are related to the
number of consumption units in the family, the latter being based
usually upon the relative requirements for energy of adults and
children of different age and sex. A summary statement of the more
important of these studies is contained in the Year-Book of Labour
Statistics, covering recent studies of family budgets of some twenty
1

LEAGUE OF NATIONS: The Problem, of Nutrition. Vol. IV. Statistics of
food Production, Consumption and Prices. See L. of N. Pubi. II. Economic
and Financial, 1936. II. B.6.
2
Op. cit., pp. 225-233.
S. L.

3

— 34 —

countries, and in an article on " Recent Family Budget Enquiries "
published in the International Labour Review for November 1933.
Since these figures were published, a number of important enquiries have been made, the largest of which is that made in the
United States including a large sample of many thousands of
wage earners' families 1. In Great Britain a study is now in
progress covering a large sample of families of wage earners.
In general, however, these family budget studies are based upon
a relatively small number of cases, owing to the difficulty of obtaining the full information necessary for a large number of families.
Hence it becomes difficult to draw conclusions applicable to the
working class as a whole, on account of the difficulty of proving
that the sample is representative. However, if the evidence
indicates that the food consumed in the selected families is inadequate, it may be inferred that the food supply for the working
classes from which the sample was drawn is also inadequate. Such
generalisation's usually justifiable, since the procedure of limiting
the data to families which kept complete records of food consumption over a long period, such as a year, is such as to ensure that they
in fact represent the more intelligent and more well-to-do portion
of the group sampled. In general, figures based upon family
budgets give two types of data: (1) those based upon the percentages of expenditures which are spent for food and (2) data showing
the quantities of food consumed per family or per consumption
unit.
(a) Percentage of Food Expenditure
Practically all family budget enquiries show the percentage of
expenditure which goes for food. In practice it has been found
that the percentage of expenditure for food can be used as a rough
index of the worker's level of living. It appears that, with increasing
income in each area studied, smaller percentages are spent for food2..
Hence, the smaller the proportion spent for food, the higher the
standard, or, conversely, the higher the proportion spent for food,
the lower the standard of living. However, attention has been
drawn to the fact that in certain cases (as shown, for example, by
certain Far-Eastern enquiries or by studies relating to the unemployed), when income is below a certain point, the percentage spent
for food tends to increase with income up to a certain point. After
1
2

Study of Consumer Purchases, cited above.
See percentages given in the Year-Book of Labour Statistics.

— 35 —
income reaches this point, the percentage spent on food tends to
decrease as income increases further. This suggests that below a
certain critical income level, other needs, e.g., shelter and clothing,
are more imperative, once hunger has been allayed, than the demand
for full satisfaction of food requirements.
(b)

Expenditures on Chief Groups of Food

Information on the distribution of the food budget, according to
specific groups such as bread and cereals, meat, fish, milk and milk
products, vegetables and fruits, miscellaneous foods, etc., is available
as a result of most of these family budget studies. The results
for some fifteen countries are shown in the 1937 edition of the
Year-Book of Labour Statistics. An analysis of these percentages
in conjunction with variations in income indicates that at low levels
of income a large proportion is spent for the energy-giving foods,
with the result of a relative deficiency in the more expensive protective foods. As income increases, the protective foods form a
larger proportion of the food budget.
(c)

Quantities of Food Consumed

A considerable proportion of these family budget studies give not
only the percentages of the budget spent for food but also the
specific quantities and show not only quantities per family but
also the quantities consumed per consumption unit. The data
available from the chief recent enquiries on family consumption,
the quantities of each of the principal foodstuffs consumed by
workers and salaried employees' households, have been brought
together with a rough analysis of their nutritive content, by the
International Labour Office (Workers'1 Nutrition and Social Policy).
Such data are available for many European countries, the U.S.A.,
Brazil, Mexico and Palestine. The Y ear-Book of Labour Statistics
for 1935-1936 summarises the data available on the quantities of
each of the principal types of foodstuff consumed by workers'
households. These references also consider the differences in food
consumption at different income levels.
Summaries and references to the scattered studies made in the
Far East are contained in the Preparatory Papers of the Intergovernmental Conference of Far-Eastern
Countries on Rural
1
Hygiene ; for instance, the results of several Chinese studies
covering a range of social classes are compared in the Report
on China.
1

Published by the Health Section of the League of Nations, Geneva, 1937.

— 36 —

(d) Appraisal in Terms of Adequacy
For purposes of judging the standards of living as shown by food
consumption the food consumed can be appraised in terms of
adequate nutrition. Optimum standards or norms of adequate
nutrition have been set up by means of physiological research in
terms of calories, proteins, necessary minerals and vitamins 1.
These standards take into account the age, sex and physical activity
of the individual 2 .
Comparatively few of the family budget studies made thus far
have included any attempt to appraise the actual food consumed by
workers' families in terms of adequacy on the basis of these norms
or standards. Among those in which some attempt at appraisal
is made may be cited the recent study in the Union of South Africa
in 1936 and certain studies of diets in the United States, including
a series of studies made in connection with the Study of Consumer Purchases 3. In one or two other cases appraisals are made
of the probable sufficiency of calories furnished in the dietaries
but without a full analysis of other elements in the diets. Mention
has already been made of the appraisal in Sir John Orr's study of
the food consumed by the population of Great Britain.
An alternative method of appraisal should be mentioned. When
1

The London standards established by a Technical Commission of the
Health Committee of the League of Nations which met in London in November
1935, and since amended, are the most widely agreed-upon standards; see
INTERNATIONAL LABOUR OFFICE: Workers' Nutrition and Social Policy, pp. 5-7.

2
The extent to which the London standards, which are applicable for
Europe and North America, are valid for the Far East is not known accurately.
It is established that physiological requirements vary with climate, body size
and activity, for which adjustments may be made, and possibly that requirements also differ according to race and the adaptation of the digestive process
to customary dietaries. There are even differences in the chemical composition of the same foods produced in temperate and tropical countries.
The Intergovernmental Conference of Far-Eastern Countries on Rural
Hygiene in 1937 reported that: " It is at present difficult to define
standards which will be useful and applicable throughout the East ",
but adopted general dietary recommendations, such as that the inclusion of some animal protein in the diet is desirable. The Committee of
Technical Experts has recommended further research in the problem of
standard and actual dietaries, the state of nutrition, and the composition of
foodstuffs peculiar to the Far East, and allied problems (Nov. 1937). Working
standards for various Far-Eastern countries are contained in the Preparatory
Papers of the Conference on Rural Hygiene. Most of these reports contain
budgets of foodstuffs providing " sufficient " protein, fat, and carbohydrates,
sometimes minerals, and occasionally vitamins. More attention has been
paid in those countries to the carbohydrate, protein and mineral content
of foods than to their vitamin content. Cf. LEAGUE OF NATIONS, HEALTH
ORGANISATION: Preparatory Papers of the Intergovernmental Conference of
Far-Eastern Countries on Rural Hygiene.
3
See also STERNER, Richard: " The Standard of Living in Sweden ". In
Annals of the American Academy of Political and Social Science, May 1938,
pp. 7-19, esp. p. 15.

— 37 —

information on the expenditure for food but not quantities consumed
is available, the food expenditures can be compared with the
estimated minimum cost of a standard diet which satisfies normal
requirements of various foods. This procedure in general underestimates the proportion of cases of deficient dietaries, since the
assumption is made that families with a food budget equal to, or
in excess of, the minimum actually obtain a diet equal to the standard—in other words, the assumption is that the most economical
selection of foods is made by the housewives. Standard diets,
with their costs, have been prepared in many countries for purposes
of setting minimum wages, relief allowances, etc.
(3)

General Statistics on Nutrition

Another approach to the problem of the adequacy of food or
nutrition among the working classes is through general statistics
of malnutrition or of the prevalence of deaths or cases of specific
diseases associated with diet deficiencies,
Surveys of malnutrition. — A great deal of work has been done
on the problem of surveying the extent of malnutrition in the
general population or in specific groups and particularly among
school children. Methods have been developed with reference to
the best way of selecting cases where malnutrition may be inferred
from the relation of weight to height or other physical measurements,
the selected children or selected persons being then given a detailed
medical examination to ascertain the presence or absence of malnutrition. Such tests based upon measurements usually have to
be supplemented by physical examinations in which clinical evidence
or clinical tests are relied upon to demonstrate the condition of
malnutrition 1.
With regard to the prevalence of malnutrition among working
classes and the families of wage earners, it is difficult to obtain
general evidence. For the most part evidence of prevalence of
malnutrition is not related specifically to wage earners or to wage
earners' families. However, since in general wage earners and
wage earners' families form a large proportion of the total population
and since, further, other things being equal, the more well-to-do
1
See BURNET, E., and AYKROYD, W. R.: " Nutrition and Public Health ",
Quarterly Bulletin, Health Organisation, Vol. 4, June 1935. League of Nations,
Geneva, p. 37. See also the recommendations on methods of assessing the
nutritional state of children and young people contained in the Bulletin of the
Health Organisation, Vol. VI, No. 2, pp. 129-204, April 1937, which contains
also a valuable bibliography of methods and studies.

— 38 —
classes tend to have a smaller proportion of malnutrition than
wage earners, the proportion of malnutrition found for a population
as a whole usually understates the proportion found among wage
earners and wage earners' families. On this general basis, the
evidence of the extent of malnutrition in the population may be
used as indicative of the minimum prevalence in the group with
which the present Report is particularly concerned *.
Mortality and morbidity statistics. — A second source of evidence
on the prevalence of malnutrition is the statistics of death rates or
specific diseases associated with diet deficiencies. These diseases
include for example beriberi, rickets, pellagra, scurvy, osteomalacia,
etc. Especially in the Far East studies have been made of the
incidence of these diseases. Beriberi is commonly related to the
consumption of polished rice. Pellagra is found in the southern
part of the United States, in parts of Italy, though much less than
formerly, Rumania, etc., and in the Far East and is usually to be
associated with a limited diet with the absence of a specific vitamin
or protective substance. For the most part these deficiency
diseases, where they have been identified, can be corrected by
relatively simple means. For example, beriberi can be prevented
by the substitution of unpolished for polished rice, rickets
by the addition of cod liver oil (or certain other substances)
t o the diet or by exposure to sunlight. To a large extent,
therefore, the prevalence of these deficiency diseases may be
attributed to ignorance as well as to the economic factor which
plays a part.
Besides the death rates and sickness rates from specific diseases
associated with deficiencies in the diet, high death rates and high
prevalence of sickness from pther diseases may be due to or favoured
by malnutrition, the resistance of the individual being so reduced
that he falls victim to diseases which in good health he might have
been able to throw off. Evidence of the extent to which this is generally true is difficult to bring. One disease, especially common during
the productive ages, where this is an important factor, is tuberculosis. Interesting evidence of the connection between tuberculosis and under-nutrition is shown by the increased prevalence of
tuberculosis in Germany during the period of the world war, when
food consumption was restricted.
1
For a summary of enquiries of the several types, see " Report on the Work
of a Group of Experts appointed to Study Methods of Assessing the State of
Nutrition in Infants and Adolescents ", in League of Nations Bulletin of the
Health Organisation, Vol. VI, No. 2, April 1937, pp. 188-199.

— 39 —
B.

HOUSING

A second major element in the family budget is housing or shelter.
From the point of view of any given family, housing must satisfy
three types of requirements. It must: (a) provide protection
against weather, etc.; (b) be adequate for the family, regard
being had to its size, age and sex composition; and (c) be at a
rent which the family can afford.
(1) Standards of Housing
Standards of housing might be set up for each of these three
requirements: standards governing the physical aspects of the
house, standards governing overcrowding and segregation of the
sexes, and standards of rental.
In respect of the condition of the house, standards may be established on three main bases—on a basis of safety as against fire
or collapse; on a basis of health, either of protection against a
specific disease or diseases or protection or promotion of general
health; and on a basis of comfort, where provision of running
water, hot water, central heating'facilities, etc. (so far as these
are not included within the scope of health), may form part of the
standard.- Standards of the first two classes, those for safety and
health, are commonly established by legal regulations, such as
those governing demolition of below-standard and condemned
houses, or those governing the construction of new houses. In
cases of the third type, affecting comfort, they are usually in the
form of norms agreed upon by experts for purposes of appraisal.
In practice, standards as set up in legal regulations or recognised
by official bodies vary greatly from place to place and from country
to country according to conditions 1 . As an example, housing
standards relating to health conditions may be considered. Regulations for the control of specific diseases are important only where
the specific diseases are prevalent; as examples may be mentioned
rat-proofing of the house as a measure against the spread of bubonic
plague, the abolition of packed earth floors—a regulation important
1
Most countries have laws providing detailed regulations in regard to the
construction and use of buildings. In general, the more specific rulings are
left to the local authorities. Statutory provisions of this type deal with zoning,
the size of buildings, construction for fire prevention, sanitation, light and air,
disposal of sewage, etc. More general standards may be drawn up for the
preparation of local housing regulations; for instance, in the Netherlands there
have been published under the auspices of the Public Health authorities guides
for the preparation of housing by-laws, containing quite specific provisions,
such as the ratio of window space to floor area.

— 40 —
for the prevention of yaws, and various measures relating to the
sanitary disposal of sewage for prevention of typhoid, dysentery,
hookworm, etc. Standards for the promotion of general health
cover the sanitary disposal of sewage, a safe water supply, and
such conditions as will promote cleanliness and a sufficiency of
light and air. All modern cities have faced the problems of sewage
disposal and the provision of safe water supplies ; though in practice,
because of the lag in requiring installations in old houses in some
cities with up-to-date sewage plants and waterworks, individual
houses may be found without sewer or water connections. In
rural areas, on the other hand, the problem of sewage disposal
may be met in practice in a great variety of ways, conforming
either to no standards at all or to standards of varying degrees
of excellence.
The Housing Commission of the Health Organisation of the
League of Nations is studying not only the problems of standards
for hygiene of housing but for comfort of environmental conditions
as well (such as temperature, moisture content, and movement
of the air) 1 . The results of research in light and heat and air change
requirements, noise, and similar subjects carried on by national
committees have been brought together by the Committee. This
work is intended to lead to " the definition, for various climates,
customs and regions, of standards of what may be termed healthy
urban and rural housing and healthy town and country areas " 2 .
In the reports of the national committees valuable materials
relating to housing standards, sometimes of wide application,
sometimes relating only to local conditions, often with extensive
bibliographies, are found. A comprehensive report on the " Basic
Principles of Healthful Housing " prepared by the Committee
on the Hygiene of Housing of the American Public Health Association attempts to formulate " the basic health needs which
housing should subserve ", together with the specific requirements
and methods of attainment. The thirty basic principles " are
believed to be fundamental minima required for the promotion
of physical, mental, and social health " and include such principles
as : " Maintenance of a thermal environment which will avoid undue
heat loss from the human body ", and " Provision of adequate
privacy for the individual ".
1

LEAGUE

OF NATIONS:

HEALTH

ORGANISATION: Housing

Commission.

Series of mimeographed documents numbered from C.H. Comm. Hab./l
(18 January 1936) to C.H. Comm. Hab./46 (2 February 1938). Geneva.
2
LEAGUE OF NATIONS: The Housing Problem and its Study onthe International
Plane. Note by the Secretary-General. A.26.1937.III. Geneva, 1937.

— 41 —
Housing standards as thus formulated vary widely, partly
because they are evolved with the idea of amelioration of conditions,
and partly because needs differ with climates, the prevalent types
of disease, and other factors.
Standards for various regions in the Far East are found in the
papers of the League of Nations Far-Eastern Conference on Rural
Hygiene, in the judgments expressed in the descriptions of existing
housing, in the discussion of the improvements being made by the
authorities, and in the recommendations regarding the control
of specific diseases. These standards usually relate primarily
to sewage disposal, sometimes also to light, air and crowding,
and to construction such as to make it easy to keep dwellings
clean. The standards in these reports are expressed in rather general
terms, such as " a satisfactory development . . . in the increased
installation of bored-hole latrines " 1. But in the Far East standards
are largely a matter to be evolved by local health and other authorities. The Far-Eastern Conference recommended specifically
in regard to housing that kitchens should not be inside the living
part of the house and that cattle-sheds should be detached ;
beyond that, however, it recommended that the Eastern Governments constitute committees for small-village planning, to set up
local standards dealing with details regarding siting, types of
houses, building material, ventilation, lighting and heating, waste
disposal and water supply.
Standards of a sort for housing in Europe and North America
are numerous and frequently detailed. Between the extremes of
the two types of standards, one expressed in.terms of objectives
(such as sufficient light and air) and the other expressed in engineering specifications, there is a wide range.
Other standards of the physical aspect of housing may be based
upon ideas of comfort or social well-being. For example, the
International Labour Office, for purposes of its rent investigations
in different towns, has set up a scale of qualitative standards by
which workers' dwellings may be rated, taking into consideration
the type of building in which the dwelling is situated, and the
installations of water, toilets and bathrooms 2.
Standards of occupancy include those relating to overcrowding
1

L E A G U E OF N A T I O N S : H E A L T H O R G A N I S A T I O N : Intergovernmental

Confer-

ence of Far-Eastern Countries on Rural Hygiene. 1937.
2
G U Y E , R o b e r t : " A F u r t h e r Contribution t o the I n t e r n a t i o n a l Comparison
of R e n t s ". International
Labour Review, Vol. X X X I V , N o . 5. November
1936.

— 42 —
in general and in particular those relating to the provision of
adequate bedrooms. The most common type of standard for
overcrowding is measured by t h e number of persons t o a room.
The various definitions used in different countries are presented in
a report of a sub-committee of the Committee of Statistical Experts
of the League of Nations; t h e most usual definition is that
families are considered overcrowded when there are more than
two persons to a room. This standard is adopted by t h e International Labour Office for its comparative study on overcrowding
in different countries made in its report on housing policy in
Europe 1 . A discussion of measures of density of occupation, such
as number of persons to a dwelling, or persons to a room, or rooms
to a person, and more detailed scales, is found in the International
Labour Office's Methods of Compiling Housing Statistics, published
in 1928.
Here again, standards which bulk large in appraising the quality
of housing in Europe may not be relatively so important in other
countries on account of different conditions. The standards for
housing in terms of space, the most common for Europe, are of only
secondary importance in many tropical countries. A report for
Siam states that " there is no question of overcrowding in rural
Siam " 2 ; a report of the Netherlands Indies Public Health Service
states : " On account of the climate and manner of living in this
country the factors of floor space and floor area per inhabitant are
less important, so far as housing problems are concerned, than in
Europe " 3.
After measures of simple density of occupation come those for
the separation of the sexes. In the United Kingdom the standard
adopted by the Ministry of Health considers a dwelling overcrowded if the number of persons sleeping in the house is such t h a t
any two persons of different sex, ten years old or more and not
living together as husband and wife must sleep in the same room,
and by combining the sex separation and density standards based
on number and area of rooms, schedules of the standard of capacity
of dwellings of different sizes are prepared 4 . Similarly, in a
> INTERNATIONAL LABOUR O F F I C E : Housing
Policy in Europe.
and R e p o r t s , Series A (Housing), N o . 3. Geneva, 1930, p . 22.
2

L E A G U E O F N A T I O N S : H E A L T H O R G A N I S A T I O N : Intergovernmental

Studies
Confer-

ence of Far-Eastern
Countries on Rural Hygiene. Publications: I I I . Health,
1937,
I I I . 13. Geneva, 1937, p . 34.
3
Ibid., I l l , p . 104.
4
Report on the Overcrowding Survey, op. cit., p p . iv-v, See also: A Social
Survey of Plymouth.
London, 1935, p p . 14-17.

-

43 —

report on overcrowding in the Netherlands, it is considered that
children of different sexes over 12 years of age should have separate
bedrooms 1.
Standards of rentals for housing are of limited extent, and in
general are to be found only in connection with public works projects for slum clearance, for construction of workers' dwellings,
and similar works of social policy.
In summary, these housing standards may be characterised
as of varying levels and of varying degrees of urgency. In general
a standard applicable to a particular area or community, if adopted
as a matter of legal regulation, sets a minimum which all dwellings
within the community should satisfy, and the presence of dwellings
which do not conform represents a deficiency in housing which is
more or less serious, depending upon the character of the standard
violated. While the type of minimum standards in force in a
community affords insight into its relative position with regard to
housing, it is far from affording a sufficient basis for a general
appraisal of housing facilities; this must be done in the light of all
the evidence available.
(2)

Statistics

of

Housing

In general there is a considerable amount of statistical data on
housing for most countries in Europe (except the Balkans), for the
British Empire, and for the United States. In most cases general
censuses of dwellings are taken at the time of the census of the
population and cover either the entire population or, as a minimum,
the urban centres. In many countries these are supplemented by
special enquiries made in selected cities.
The statistical data available for appraising housing in the
different countries will be considered briefly under three topics,
(1) occupancy and overcrowding, (2) the health, safety, and comfort
aspects of housing, and (3) rents, etc.
(a)

Occupancy Statistics

In many countries the census statistics of housing are available
for the whole population (Canada, Italy, Netherlands, Norway,
Poland, United Kingdom and Switzerland), showing the size of
families in conjunction with the number of rooms occupied, thus
1
KAA, H. VAN DER: The Housing Policy in the Netherlands. League of
Nations Health Organisation. Official No. C.H. 1165(a). Series of League
of Nations Publications: III. Health, 1935, III. 4. Geneva, 1935, pp. 19, 43.

— 44 —

giving the basic data needed for a classification of population
according to density of occupancy of dwellings.
These data permit classifying the population (but not always the
wage-earning classes) according to the average number of persons
per room, and with this information the proportion of the population
living under conditions of occupancy exceeding any given standard
of density, such as two persons per room, can be determined. Even
where the detail by size of family is not available, a simple classifition of the number of dwellings by the number of rooms shows
as a minimum the number of families living in one room, a point
which is itself of considerable interest.
Besides these, figures based on census data for urban areas only
are available for certain other countries (e.g. Czechoslovakia,
Denmark, France), or as a result of special enquiries as in Germany,
Sweden and the United States.
The precise significance of all these figures on density of
occupancy, however, depends in part upon other points. The
density of occupancy has been studied, especially in certain countries, for example, Germany, Great Britain, Norway, and Sweden,
by means of more detailed criteria in which age, sex, and marital
condition are taken into account 1 . In certain cases in special
investigations the size of rooms (area or volume) is utilised. This
is done, for example, in a first enquiry into " Workers' Dwellings
in Different Cities ", the results of which were published in the
International Labour Review, Vol. XXXIV, No. 5, November 1936.
A general survey of international statistics on these points is
contained in reports of the International Labour Office on housing
policy in Europe and in the reports on the results of studies in the
field of housing published in the International Labour Review.
In close connection with these statistics of occupancy, the question
of sharing the dwelling with one or more persons outside the family
or with other families is important from various points of view.
The cases of sharing a dwelling include many different kinds of
cases: for example, the family may have chosen to let rooms as a
method of meeting their particular requirements; in other cases
the family is forced to do so by economic necessity or because of the
absence of dwellings corresponding to its needs. Likewise, the
cases of sharing rooms correspond to a great variety of conditions,
1
In some countries the number of supplementary rooms not counted as
" habitable rooms " is taken into account. In certain cases, figures showing
the number of occupants of each sex, age, etc., are available, though not
utilised in calculating the figures for density of occupancy.

— 45

-

from the boarder not related to the family but included in the household as other members of the family to the case of two or more
families arranging their lives separately and as comfortably as
possible in the same dwelling. In the countries where the statistical
analysis takes these points into-account—for example, Germany,
Canada, Great Britain, Poland, Czechoslovakia, Sweden, etc.—
there are great differences in the methods and scope of the
data.
(b) Physical Characteristics and Equipment of Dwellings
The censuses and enquiries of certain countries, especially those of
Denmark, the United States, Italy, Norway, Poland, Sweden,
Switzerland and Czechoslovakia, give various details on the type
of dwelling. These data on what one may call in general terms the
quality of dwelling cover a wide variety of specific points—which
can be classified in two main groups :
(a) those relating to the character of the building in which
the dwellings are found, that is, the number of storeys
or the number of dwellings in the building, or distinguishing
between single family dwellings and apartment houses, or
according to the material used in construction, the date of
construction, or the state of their upkeep;
(b) those concerning the comfort of the dwellings as judged
by presence or absence of bathrooms, toilets, with and
without running water, hot and cold, the heating system,
electricity or gas for lighting or cooking, the availability
of a small garden, etc.
Among the countries with information on these points may
be mentioned: Czechoslovakia, Denmark, Italy, Poland, Sweden
and the United States.
The data on housing for the working classes are available in
certain countries, especially Denmark, Italy, Norway, Poland,
Sweden, Switzerland and Czechoslovakia, where the censuses
and enquiries furnish analyses distinguishing the social groups
to which the head of the family belongs. The criteria by which
these groups are distinguished may vary from country to country,
for they depend, in general, on the industrial or occupational
classification adopted in the general census. In these countries,
also, housing conditions are analysed separately as between urban
centres and rural areas.

— 46 —
(c) Rents, Tenancy and Home Ownership
(i) Rents
General data on rents, which are particularly interesting for
comparison with the incomes of the tenants, are available only
in a very small number of censuses and enquiries, principally
Canada, Denmark, the United States, Sweden, Switzerland and
Czechoslovakia. Family budget studies, however, available for
most important countries but based upon small samples of wageearning or other groups, usually contain data showing the proportion of the expenditure that is spent on rent or housing, classified
by income groups. It is true that because of the delay between
the taking of the censuses and the publication of the results,
the data on rents collected by this means are often somewhat
out of date at the time they are published. Index numbers of
the cost of living set up in the majority of countries offer an occasion
for obtaining returns periodically. Such data are obtained under
conditions which do not permit publishing the absolute figures,
with the result that, in general, only index numbers are available.
(ii) Home ownership
In close connection with rents, the question of home ownership
is pertinent. At present such data are available in censuses and
enquiries of certain countries, especially Canada, Denmark, the
United States, Italy, Sweden, Switzerland, and Czechoslovakia.
In certain of these countries, a classification is adopted that permits
of distinguishing homes owned from those being purchased on
time payments or on rental purchase plans. Special data are also
available on the number of cases of tenancy on the basis of services
rendered—for example, door-keepers, concierges, etc.
Besides the documentation already discussed, there are, in
certain countries, various official reports or private studies which
throw light upon existing housing conditions. It would perhaps
be going too far afield to make even a summary of these documents
in this connection. For the countries of the Far East mention may
be made of the Preparatory Papers of the Intergovernmental Conference on Far-Eastern Conditions of Rural Hygiene, already cited,
which furnish for most countries information on the general
character of dwellings, the way in which they are inhabited, the
materials used in their construction, and the type of roof. These
data are evidently much more general than those available for

— 47 —
Western countries, but they nevertheless allow a first view of the
riiost important problems in these countries.
C.

CLOTHING

A third essential component of family expenditures is clothing.
The requirements which clothing must satisfy include protection
against the elements, etc., and conformity with standards of decent
covering common to the community. In addition, clothing should
provide aesthetic satisfaction, etc.
(1)

Standards

Standards for clothing for purposes of appraising levels of living
of wage earners' families in different countries and in different
communities are difficult to establish. They must vary obviously
with external conditions to take account of differences in climate,
etc. Community attitudes also are an important element in such"
standards, both in regard to what is required by standards of
decency and in relation to habit and fashion. The customs and
traditions of the group are usually decisive in regard to certain
minimum requirements. In addition to these minimum requirements social standards of appearance may determine what
should be worn to work. The clothing requirements for a manual
worker are different from those for a clerical occupation. Social
standards are important also in determining clothing requirements
for children attending school. Apart from these, fashion, social
custom and tradition are potent influences in determining what
shall be worn and under what conditions. To a large extent these
influences determine all the so-called " upper levels " of clothing
standards, but, as just mentioned, occupation and school attendance also have an influence when they affect the clothing requirements of wage earners' families.
Finally, health plays a part in certain aspects. The wearing
of shoes, for example, in regions where hookworm prevails is
an important preventive measure. Sufficiency of clothing for
protection against cold may be regarded as essential in the interests
of good health. For present purposes, that is, for a study of the
clothing requirements or standards of wage earners' families,
certain types of standards set by fashion or social custom may
be regarded as outside the scope of the present discussion.
An approach to a series of specific standards which may be
utilised for purposes of appraisal may be found in specific quantity

— 48 —
budgets of clothing established for arbitrarily determined or
designated levels of living. These studies indicate the quantity
of each type of garment required for families of each different
size and composition at the so-called " relief " level (as set up
for purposes of determining relief payments), or at a level appropriate for determining minimum wages, or at a minimum comfort
level \
These standards represent not so much standards generally
agreed upon as the views of individual experts on what a family
will require in the way of clothing at a particular economic and
social level. Furthermore, the clothing actually purchased (or
regarded as necessary) by the highest economic group of wage
earners may be considered as a standard which wage earners at
lower levels of living would like to attain.
With special reference to certain possibilities of improving
standards, consideration should be given to physical and economic
qualities of clothing, for example, expense, durability and the
capacity for retaining warmth. Reductions in cost for identical
wearing qualities or improvements in quality without increased
expense may make possible, other things being equal, a higher level
of living in respect of clothing.
(2) Statistics
The principal source of data on clothing is family budget studies.
These show in practically all cases the proportion of family expenditures which is spent for clothing. The significance of these data
is limited in two respects; (1) in that the exact definition of clothing
differs in the different studies; for example, as to whether or not
laundry is included; and (2) the absence in some cases of data on
the number, age and sex of the persons for whom clothing is purchased 2. Obviously no attempt can be made to determine whether
the clothing allowance is adequate without reference to the composition of the family. In some cases expenditures are shown per
consumption unit, in which the composition of the family is given
in terms of the unit, which, however, is usually based on food
1
See, for example, ROWNTREE: Poverty. A Study of Town Life. See also,
Factors to be Considered in Preparing Minimum Wage Budgets for Women,
prepared by Economics Division, U.S. Bureau of Home Economics, Department of Agriculture, in co-operation with the Women's Bureau, Department
of Labor; also STECKER, Margaret Loomis: Intercity Differences in Costs of
Living in March, 1935, 59 Cities; preliminary report, Division of Social Research,
Works Progress Administration, Washington, 1937.
2
See note in section on " Food " for references to summaries of reports on
family budget studies.

— 49 —
requirements, though, in a few instances, on relative clothing costs.
In a survey of wage earners' and clerical workers' expenditures made
recently by the United States Bureau of Labor Statistics, the
relative expenditures in each group were analysed for persons of
different age and sex. Detailed data on the types and materials
of garments, prices, and the season of the year when purchased,
were obtained for each member of the family. In certain other
enquiries, certain details of clothing according to the age and sex
of the family members are available.
When data exist showing the actual garments purchased for the
different members of the family, the basis is given for an appraisal
of the level of living in respect of clothing. A substitute for these
data is evidence showing the per capita expenditure for clothing
in relation to the cost of a standard budget for clothing in the
particular locality, since with such a basis for comparison the
conclusion can be drawn whether the clothing budget for the
individual family equals or exceeds or fails to equal the amount
adjudged necessary. Such analyses where available are usually
in the form mainly of data showing the actual amount of money
per unit available for clothing; and the number and proportion of
families with classified amounts of clothing budgets. Such
material is naturally limited to the materials collected in family
budget studies. No comprehensive data are available on the
general adequacy of clothing budgets for working-class families.
D.

MISCELLANEOUS

Finally, some consideration should be given to the group of
miscellaneous items in the family budgets in which may be included,
for present purposes, all items except those for food, housing and
clothing. For present purposes special importance attaches to
this group, since many students of family living have considered
the percentage of expenditure allocated to miscellaneous items as
a rough index to the level of living. The larger the expenditure
on miscellaneous items, the higher the standard. There is, of
course, this justification, that with larger incomes there exists
normally a greater freedom and not only can more and better food
be purchased, more and better housing, more and better clothing,
but various other minor wants can be met, leading to a larger
proportion of the income being devoted to non-essentials and
luxuries, thus indicating a " higher " standard of living. This
conclusion is not true, however, without certain reservations;
s. L.

4

— 50 —

in particular, the reservation that the food purchased (for example)
is not skimped in order to make possible larger proportionate
expenditures for some one or other more miscellaneous items that
are deemed, in the particular case, to be more important than
food.
Standards of miscellaneous expenditure are even more difficult
to establish than standards of clothing. Two types of standards
may be considered : (1 ) those based upon proportionate expenditures ;
these standards are supported by family budget studies indicating
the increasing proportion of income devoted to miscellaneous
expenditure as income increases 1; and (2) those based upon an
attempt by experts in family budget studies to set up standards
in respect of the amount or presence of certain more or less typical
items of the miscellaneous group in the actual wage earner's
budget. That budget includes both necessaries as well as luxuries;
in the former may be cited soap, car fare to and from work, etc. ;
in the latter would be included such items as " amusements",
a wireless set, telephone, or even a motor-car. Obviously both the
items covered and the amounts provided in the budget would
bear relation to the particular level of living concerned. Such
standards may be found in quantity budgets set up by experts for
specific levels of living 2.
Percentages spent on miscellaneous items are given in the various
family budget studies 3. In this case also, variations in the average
of the group affects the significance of the differences in percentages.
Data on the presence of specific items, together with the amounts
or proportions expended on them, are also available in certain family
budget studies and throw light therefore, so far as the character
of the item can be utilised as an index of the level of living, upon
the standard of living in the families covered. Family budget
studies must be regarded as the principal source of information on
expenditures of this type.
General data, however, on the prevalence of specific items are in
some cases also available, but these are for the most part
fragmentary. However, information is usually available on the
number of wireless sets and the number of telephones in most
countries, though not usually indicating the type of family,.
whether wage earners or others, which possesses them.
1
Without necessarily attempting to label or fix exact percentages for different levels designated as poverty level, minimum comfort level, etc. Sea
below, p. 58.
2
See note in section C on " Clothing ", p. 47.
8
See note in section A on " Food and Nutrition ", p. 33.

— 51 —

III. — Social Elements in the Standard of Living

Important elements in the standard of living are dependent not
only upon the expenditures of individual families but also largely
upon community action. Examples of these are health, education
and social services. True, in many families expenditures are made
on health (doctors' bills, medicines, hospital bills, etc.), on education, such as private school fees, tuition, etc., and in some cases
these expenditures form an important proportion of the total
family budget. In the main, however, and among the working
classes especially, the level of the standard of living in respect of
health, education and social services is dependent upon community
or State services supplied free of cost. In regard to health, the
burden of the provision of medical services is for the most part
taken care of through various forms of social insurance; for the
rest, free community services such as hospitals, etc., may be provided. A minimum of education free of cost is provided for in
most countries by the community. This minimum varies from
place to place or from country to country, depending upon the
traditions of the various countries and how much of the public
funds can be devoted to these purposes. But the degree of public
health protection enjoyed by a community and the degree of
educational facilities provided free form such an important part of
the standard of living enjoyed by the working classes that some
discussion of these topics is essential to an adequate picture of
living standards. In this section, therefore, three topics will be
discussed briefly: health, education and social services.
A.

HEALTH

With regard to the wage earner, it is particularly true that good
health is a condition of maximum earning power; The dependence
of health, in part at least, upon economic conditions has long been
known. Infant mortality has often been described as one of the
most sensitive indices of social and economic conditions, though
perhaps more so in the past in areas where conditions were especially
bad, than now when many connections between disease and low
economic conditions have been severed. The close relation
between economic status and mortality has also long been recognised. The idea that " public health is purchasable " has been
effective in promoting increased appropriations in many commu-

— 52 —

nities for protection against disease. Many examples can be cited
of the effectiveness of the introduction of specific measures in
terms of decreasing death rates, for example, the relation between
the diminishing typhoid fever rate and protection of water and
milk supplies; the decrease in prevalence of diphtheria with the
measures taken for immunisation of susceptible persons; the fall
in tuberculosis mortality, etc.
In so far as the evidence available on the state of health among
the working classes and on the possibilities for its improvement are
concerned, two lines of enquiry may be referred to: first, the
expenditures for health and medical services by working-class
families or on behalf of the working classes and, secondly, disease
and death rates among workers and their families.
Expenditures for health and medical services. — Family budget
studies of workers' families show small amounts expended for
medicine, doctors' bills and other health aids. These expenditures
are for the most part small in proportion to the total expenditures.
Further, they bear no relation to health, but rather to ill-health,
nor are they specially significant of the general level of health.
They show merely that in certain families expenditures must be
made for these purposes. In regard to health, the principle of
insurance is particularly applicable. The incidence of sickness in
a particular year is not uniform, and hence the actual expenditures
in an individual family or for an individual case are not indicative
of the general level of living. The fact that regular provision for
sickness insurance exists is important. But since social insurance
plans provide for only a small part of the health services which
are to the advantage of the worker, even such expenditures are
not specially significant for the present purpose.
In some countries general surveys, or surveys based upon large
samples, have been undertaken to determine the extent of sickness
in different groups of the population and to show the amounts
expended for medical services and other health items; in particular
the extensive studies in the United States made for the Committee
on the Costs of Medical Care may be cited, as well as the studies now
under way of the prevalence of sickness in a number of cities.
Disease and death rates among workers and their families. —
The second line of enquiry is into the death and disease rates
among workers and their families. In practice, reliance must
be placed for the most part on death and disease rates for the general
population. Since the workers form the largest proportion of the

— 53 —

population and since in general death rates and disease rates are
higher among workers than among the rest of the population,
the figures for the general population can be used as an indication
of the rates among workers.
This is not the place to attempt to survey in detail all the sources
of information on this subject. It is enough to say that in all
modern industrial countries adequate statistics of deaths and death
rates are available. Besides the general death rate as an index of
healthfulness, which certainly has to be corrected so as to make
allowance for the variations in the age and sex composition of different populations, special attention should be paid to the infant mortality rate and the death rates from certain diseases associated with
poverty or malnutrition—for example, tuberculosis. So far as special
rates for occupational diseases are concerned, comparatively
few countries have any adequate material on this subject. Great
Britain, which publishes analyses of occupational mortality every
ten years, is practically the only country with data showing the
relation between occupation and mortality, though a first study
of occupational mortality is available for the United States.
B.

EDUCATION

Granted that at least an adequate elementary education is an
important determinant of the standard of living of the working
population, a variety of data is available to throw light on the
question of the degree of such education enjoyed, if not by
working-class families, then by the population as a whole 1.
In the first place, the degree of education received can be tested
by the results. A minimum education should permit of reading
and writing in some language. Statistics of illiteracy are collected
less frequently than data on school attendance, but are available
in the censuses of a number of countries and show the extent
to which the population of those countries is able to read and
write. All the countries of Western Europe have small percentages
of illiteracy. In the United States in certain groups—Indians,
foreign-born, Negroes—substantial proportions of illiterates are
1
For a bibliography, see TUROSIENSKI, Severin K. : Foreign and Comparative
Education. A List of References. U.S. Dept. of Interior, Office of Education,
Bulletin 1934, No. 10, Washington, 1934. The Educational Yearbook of
Columbia University Teachers' College (United States of America) contains
accounts of the educational systems of various countries from time to time.
See also the publications of the International Bureau of Education, with
headquarters in Geneva.

— 54 —
found. In the U.S.S.R. a considerable proportion of the older
age groups especially are illiterate. In certain countries, such
as India or China, extraordinarily large proportions of the population are unable to read and write.
Other sources of information on the amount of education available
are data on the number of schools, the number of qualified teachers,
the number and proportion of children of each age attending school,
and the number of persons attending high schools, universities
or other institutions of higher learning. Finally, statistics of the
amounts of money spent by Governments on schools furnish
another source of information.
C.

SOCIAL SERVICES

Apart from health and education, perhaps the most important
aspects of living standards which are dependent upon group
activities are the social services, including arrangements for social
insurance of all kinds—accident, sickness, unemployment, oldage pensions or insurance, together with relief and all related
activities. All these are in the nature of services which are
necessary if the individual falls into the group of those in need,
and are of great value in any case since he knows that if need
arises the services are there to take care of him and his family.
In the absence of such social services the individual either has
no provision at all or else he is forced to make provision by his
individual efforts, which in most cases means that little or no provision is made.
So far as the extent of these services is concerned, reference
may be made to the comprehensive survey contained in the International Labour Office publication, International Survey of Social
Services, 1933 1.
TV. — Problems of Evaluation

The account given above of the material available on standards
of living has shown that a large amount of information exists on
this subject, but that considerable gaps remain to be filled if an
adequate picture of workers' living standards is to be drawn.
It has also shown that the material available is heterogeneous;
1

Studies and Reports, Series M, No. 13, 2 vols.

Geneva, 1936.

— 55 —
any attempt therefore at evaluation of standards depends on the
elaboration of methods for the use of the existing material. In
view of the common tendency to discuss living standards in general
terms, and to compare them both as between different dates and
between different groups or different countries, brief consideration
is given in this section to the problems involved in such evaluations.
A.

INTRA-NATIONAL TIME AND PLACE COMPARISONS

There are, as indicated above, two ways of comparing standards
of living within a country, which may be briefly referred to as time
and place comparisons. In the former, the standard of living of the
population of a country or of certain groups of the population are
compared at one period with that in another. These are perhaps
the most frequent, and also the easiest, comparisons to make. Thus
it is often said that the standard of living has risen by so much in
the last 50 years, or since 1914, or has fallen by a certain amount
since 1929.
Many national enquiries have been made into the changes in
the standard of living of the workers, and it is not possible to
give an enumeration of all those which have been published l .
The method usually adopted in these enquiries is to measure the
changes in the incomes of the people and, at the same time, the
changes in the cost of living. The cost-of-living changes are
usually measured by computing the changes in the cost of an
unchanged set of goods and services, as shown by family budget
enquiries. If money wages have increased more than the cost of
living, it is held that real income and hence the standard of living
have risen: and the same is true if wages have decreased less than
the cost of living has decreased. In the contrary case, real income
and the standard of living are said to have fallen.
Such methods, as will be clear from the arguments of Chapter I,
do not necessarily give an exact picture of the changes in standards
of living: they ignore various factors, especially those relating to
free services, working conditions, etc., and changes in the size of the
family. However, as these factors change slowly within the same
country, this method of comparing changes in money incomes
with changes in the cost of living and thus compiling what are
1
Reference may be made however to Dr. A. BOWLEY'S study on Wages,
Prices and Income since 1860 (London, 1937) and to the monumental New
Survey of London Life and Labour (9 vols., 1932-1938, London).

— 56 —

known as index numbers of real wages and real incomes, has been
widely adopted.
For " place comparisons ", i.e. comparisons of standards of
living at the same period among different localities or occupational
groups, the methods are not so simple. Within a country, comparisons are frequently made between workers in different occupations
or in different towns, or between different categories of workers
(according to skill, sex, etc.) by methods similar to those used for
making comparisons in time. Provided the milieu or habits of the
different groups are similar, such methods are justified. In the
case of comparisons within the same country, the advantages of
free services, etc., may be assumed to be the same for different
categories, and the comparison of real incomes would then give an
approximate measure of the relative standards of living for families
of similar type.
B. INTERNATIONAL COMPARISONS

The other form of place comparisons is that of comparing
standards of living between two different countries within the same
deriod of time. In this case considerable difficulties arise, which can
only be commented on briefly in this preliminary Report. It has
already been pointed out that it is not quite accurate to speak of the
standard of living of a country; there are many standards, appropriate to different groups and categories. International comparisons can therefore be made more accurately for particular groups
of workers and classes of population.
Even if the question is thus restricted, there are further
difficulties. In in tra-national comparisons it is assumed that
tastes, habits, climate, etc., are the same, or approximately so, for
the different categories or periods covered. This cannot be
assumed for international comparisons, nor can it be assumed
that what are described as " free services ", etc., have the same
determining influence on the standard of living. Where habits
of life approach each other in two countries, such comparisons
may be attempted, but the greater the difference between the
milieu, the more difficult it becomes to give an answer. The
elaboration of methods for such comparisons is an important task
to be pursued 1.
1

For a further discussion of this problem and its application to one partic-

ular category of worker, see INTERNATIONAL LABOUR OFFICE: A Contribution

to the Study of International Comparisons of Costs of Living.
Reports, Series N, No. 17, 1932.

Studies and

— 57 —
C.

OTHER METHODS

While continuing to regard income and expenditure as among
the chief determinants of the standard of living, attempts have
been made to find criteria other than those of differences in
purchasing power, or real incomes, for evaluating standards of
living. The percentages spent on different items such as food,
clothing, rent and miscellaneous have sometimes been used. It has
been often pointed out that, as income rises, the percentage devoted
to food falls and t h a t spent on items other than the chief necessaries
of food, shelter, etc., increases. Other criteria have been occasionally used, such as the death rate, the infant mortality rate, the
expectation of life, or even the consumption of some particular
commodities. There is a number of difficulties in such procedures.
A series of such indexes could not be reduced to a common denominator, and the importance of each criterion would vary as between
different groups and different periods.
International comparisons of standards of living are of importance
in relation to questions of international competition and migration; in some cases the migration policy of a country is based on
the alleged lower standards of living of certain races or nationalities.
But it is not possible to go further into this matter in this preliminary
Report.
D.

T H E MEASUREMENT OF DEFICIENCIES

Apart from the problem of comparing actual standards and of
measuring changes in standards is the question of how deficiencies
in living standards can be measured. This problem is of greater
practical importance for social policy. Two questions are involved :
iirst, how can deficiencies in living standards be gauged; and
secondly, by what methods can it be determined whether standards
of a given group are being raised or lowered.
Norms
As was indicated in Chapter I, one approach to these questions
is by way of norms elaborated by physiologists, hygienists, housing
specialists and others. Such norms set either an optimum or a
minimum limit in the light of which existing conditions can be
judged and on the basis of which policy may be shaped. In other
words, if there is a gap between what a family or a group of the
population actually consumes (in the way of food, clothing, housing,
etc.) and what is set up as a " norm ", this may be regarded as a

— 58 —

deficiency which would need to be corrected if the standard of
living of the family or group is to be improved.
This method of evaluation— on the basis of norms—has been
gaining ground rapidly in recent years. It is certain to be more
and more widely applied in years to come as the various sciences
dealing with nutrition, health, and human needs in general continue
to make progress.
Nevertheless, for the time being, and in the near future, evaluation cannot be made on the basis of norms only. The main reason
is that the elaboration of norms for many elements of consumption
is difficult and slow. There are disagreements among specialists
which are not likely to disappear entirely. Besides, as shown
above, some norms can be expressed in definite quantitative terms
while others involve qualitative judgments, which are more
uncertain. The whole subject is still in its early stages of development and much more progress will have to be made in this field
before any attempt to set up a model standard on the basis of
norms can be made.
Standard Budgets
The other main method available is that which has been followed
by investigators for many years. That is the method of setting up
standard budgets for population groups at different levels of income.
Such standard budgets, expressed either in terms of quantities of
goods and services or of their money costs, have been used in the
fixing of minimum wages and for other economic and social purposes.
In practice, the number of such standard budgets recognised in any
one country has varied from time to time. It would seem that
there is a tendency to accept three levels of such budgets for the
working population, namely (1) a subsistence level; (2) a decency
level; and (3) a comfort level.
It is obvious that in the setting up of such standard budgets,
the differences in the economic and social conditions of various
countries cannot but play a large part. Also, the setting up of
such budgets may be affected by cultural and political considerations. But a certain degree of uniformity of method is achieved
owing to the fact that in most countries the tendency is to regard
the actual level of living of the common labourer as the subsistence
level and the level of living of the skilled worker as the decency
or comfort level. As a result, it is possible to draw a poverty line
which lies below or near the subsistence level, and thus to grade
the population of a country into four main groups of (1) those

— 59 —
below or at the poverty line; (2) those on a subsistence level; (3)
those on a level of decency, and (4) those at the comfort level.
National Minima
It seems logical that for purposes of social policy, the two methods
outlined above should be combined. As norms of nutrition, housing,
clothing, etc., are elaborated, they can be applied in setting up and
in computing the costs of the standard budgets at the four levels
indicated. Thus, there is a progressive tendency towards greater
objectivity in the whole process of judging family living standards.
The method outlined may be called that of setting up national
minima at different levels (a subsistence minimum, a decency
minimum and a comfort minimum).
The above method provides a basis both for the evaluation
of levels of living and for the making of policy. The living conditions of a single family or of a group of families in any country
can be compared with any one of the three national minima and
appraised on the basis of such a comparison. The main problems
of social policy to improve the workers' living standard can be
more precisely defined in terms of what is proposed, namely
whether it is a question of (1) lifting persons below the poverty
line to a subsistence level, or (2) of improving the content of living
at the subsistence, decency or comfort level.
E.

NORMS, STANDARDS AND " FREEDOM OF CHOICE "

In this connection there arises the question, already briefly
referred to 1, of the relation between norms, standard budgets,
tastes and habits. In general, this is the problem of freedom of
choice in determining the distribution of family expenditures.
The problem arises owing to the fact that tastes and habits may
not coincide with the requirements of norms or with the allocation
of the family income to the different items in a standard budget. In
deciding upon the way in which the family diet is to be made up,
a housewife may arrange her money expenditures, not with a
view to securing amounts of protein or calories or vitamins, but
with a view to satisfying specific tastes and habits. Such a divergence may persist even on the assumption that those who set up
norms make every reasonable effort to take into account known
habits and tastes. It is argued that any attempt to base consump1

See Chapter I, p. 12.

— 60 —

tion on objective norms would be in disregard of tastes and habits
and would be a limitation of " freedom of choice ". The importance
of the problem lies in the fact that freedom of choice and ability
to gratify tastes are prized in themselves and are regarded as an
element of welfare.
The answer to this question depends upon a consideration
of the several levels of living distinguished above. When the problem is one of lifting a family or whole group of the population
above the poverty line, there can hardly be a question of freedom
of choice. The deficiencies are of such an elementary character and
the needs are of such.a simple physiological character. On the
other hand, when planes of living are already at a fairly high
level, when for instance food is at least sufficient in quantity and
variety to meet minimum physiological norms, and housing and
clothing serve with some degree of adequacy their most obvious
physiological functions, the problem becomes more real. In other
words, freedom of choice gains in importance with increasing
family income and on higher living levels. .
One way of meeting this problem has already been suggested 1,
namely that a large number of composites of food items, satisfying widely divergent tastes, may furnish the same number of
calories, proteins, etc., which the physiologists might recommend
as minima or as optima. Similarly there can be many varieties
of housing arrangements and clothing wardrobes, any one of which
might meet the few requirements upon which a substantial degree
of scientific agreement exists.
Another way of meeting this problem is that of educating the
consumer to a better understanding of consumption values. With
regard to food, there is abundant evidence in many countries that
education, properly applied, can serve to reshape popular dietary
habits along the lines of more rational nutrition. The widespread
tendency in recent years towards increased consumption of milk,
fresh fruit, fresh vegetables and other protective foods is a case
in point 2 . The same is true, though in less degree, with regard to
housing and clothing.
As is well known, Governments have assumed more and more this
function of consumer education, and much has been made of the
dangers inherent in such State action. However, in democratic
communities the State is only one of the several agencies engaged
1

See Chapter I, p . 13.
INTERNATIONAL LABOUR O F F I C E : Workers'
pp. 119-120.
2

Nutrition

and Social

Policy,

— 61 —
in such work; aside from extensive private advertising, there are
many voluntary non-profit organisations, scientific bodies, cooperative societies, etc., which try to reach the mind and pocket of
the consumer. If the State merely tries to eliminate fraudulent
advertising and does not prevent freedom of discussion, its educational efforts in this field can be highly beneficial, as shown by the
experience of a number of countries.
Another point calls for emphasis here. It may be granted that
much of the worker's consumption to-day is determined by considerations extraneous to needs and health. But of greater importance
is the fact that the worker's " freedom of choice " is very much
limited by income and by economic and social conditions and that,
as all studies of family budgets show, there is a strong tendency
to change habits of consumption, in accord with norms, as income
increases and as education spreads *. This means that " freedom of
choice " need not be infringed if by educational and social
policies tastes and habits are changed so that choice is not only free
but also rational. The task of improving living standards involves
not merely supplying existing wants in a more satisfactory way but
also in stimulating and developing more and better wants 2.
Still, it is possible in setting up norms or standard budgets to
disregard the short-run view in favour of long-range purposes. This
has been illustrated, for instance, by housing policies in some countries where workers' families were given better housing facilities
but at such costs as to reduce seriously their consumption of protective foods. It is also important to keep in mind that education may
be a slow process and that attempts to force rapid changes in
budgetary expenditures in disregard of tastes and habits may
defeat the very purposes of social policy.
To meet this situation, social policy could further be guided by
studies showing the effect of increased income on the demand for
commodities and services. It is known that increased income will
usually not be spread evenly over all the commodities and services
purchased with the former income, and that increased prices of

1

See INTERNATIONAL

LABOUR OFFICE: Workers' Nutrition

and Social

Policy.
2
" The chief thing which the common-sense individual actually wants is not
satisfaction for the wants that he has, but more and better wants. The things
which he strives to get in the most immediate sense are far more what he
thinks he ought to want than what his untutored preferences prompt. . . .
However, it is of the greatest importance to realise that for the vast majority
of individuals happiness depends on the possession of certain simple things
which all men desire". F. H. KNIGHT: The Ethics of Competition, p. 22.

— 62 —

different commodities do not lead to a correspondingly decreased
consumption. Among the very poorest classes any increase of
income is almost certain to be spent on greater consumption of the
necessaries of life, but as income increases there is a tendency to
redistribute the family resources in favour of the less necessary
articles, the better qualities of food and clothing, and the various
miscellaneous items which do not fall within the group of
" necessaries ". Theoretical and a few practical studies have
been made of these questions by examining what is termed the
elasticity of demand, i.e. the responses which consumers make to
changes in relative prices and in incomes. By studying how
and in what proportion expenditure on different items increases
or decreases as the prices of various articles go up or down, or as the
income of the. family increases or decreases, it is possible to obtain
measurements of the relative importance attached to these articles1.
Conclusion
The questions discussed above have an important practical
bearing on the further study of workers' standards of living and
on problems of policy. They cannot, however, be further considered
in this preliminary Report. But in order to give concrete meaning
to the various questions raised in this chapter and to illustrate the
methods described, it has been thought desirable in this preliminary
Report to present in Chapter III a summary description of workers'
living standards in a few countries.

1
See Preliminary Investigation into Measures of a National or International
Character for raising the Standard of Living, Economic Committee, League of
Nations, Geneva, 1938.

CHAPTER III
SOME ASPECTS OF WORKERS' STANDARDS OF LIVING
IN CERTAIN COUNTRIES

To give an approximate picture of actual living conditions in
various parts of the world and to illustrate the use in specific
instances of the types of material outlined in Chapter II, the
present chapter presents data regarding income and levels of living
in four countries. These countries are the United States of America,
a large and highly developed country; Poland, a country in Eastern
Europe, predominantly agricultural ; India, a large area of 360 million inhabitants where special problems and conditions arise; and
Japan, a modern industrial country of the Far East.
It should be emphasised that it is not the purpose here to present
a comprehensive analysis of the conditions of living of the workers
in each country—a task which would be quite impossible in the
space available, even if the data were in all respects adequate—but
to show by illustrative data the methods and possibilities of using
the available materials (and revealing also the gaps in these
materials) in a survey of living standards. Where the data permit,
actual consumption is compared with norms of adequacy for food
and housing. Information respecting health and literacy has been
presented because of their basic relation to other determinants of
the worker's standard of living.
The norms of adequacy used for the analysis here are necessarily
provisional. In view of the factors discussed in Chapter II
making for variations in food and housing requirements, no attempt
has been made to apply the same norms to each of the four countries.
On the contrary, the conclusions of experts within the several
countries, taking into consideration the peculiar characteristics of
each country, have been taken as the basis for the norms or points
of reference. The food norms in general take the London standards
as a point of departure, but make allowance for climate, average
body size, activity and, to some extent, customs prevailing in the
country.

— 64 —
In the case of housing, where the necessity for variation with
climate, type of society, and custom is apparent, no internationally
accepted criteria are available as a basis for norms; as a result, the
tentative norms, where any have been used, are those put forward
within the several countries by public or semi-public bodies, or,
in default of these, are such as by other tests may be fairly regarded
as meeting the community standards of adequacy in housing.
The discussion begins with the United States of America and
takes up in succession Poland, India, and Japan.
I. — The United States of America
In describing or appraising living standards in a country of about
130,000,000 inhabitants, comprising native whites, foreign born, and
Negroes, with widely varying climates and with contrasting industrial
and agricultural conditions, the question must be faced whether to give
averages and give tables showing distributions, or to give materials for
particular selected areas. The national averages tend to obscure the
wide diversities which give the real clues to deficiencies in actual consumption; distributions for the entire country, while more satisfactory
from the point of view of the size and proportion of divergent groups,
cover up the varying regional differences; while the analysis of selected
communities suffers from the difficulty that they may not be representative of the whole.
As the available data do not give national averages or even distributions on many points, the procedure adopted here has been to study
the data for selected communities. The communities selected are:
New York City, the largest city in the country, with a heterogeneous
population of over 7,000,000 engaged in a wide variety of commercial
and industrial pursuits; Portland, Oregon, a city of approximately
300,000 on the Pacific Coast, predominantly native white, a shipping,
trading, and manufacturing centre for a relatively sparsely populated
hinterland of timber, wheat, and fruit lands; and Atlanta, Georgia, a
city in the south-east with a large Negro population, a commercial and
industrial centre for extensive cotton, tobacco, timber, and fruit-growing
areas. In addition, data are presented for " share croppers " in the
south-east, who till land not their own on a crop-sharing basis.
A.

FAMILY INCOME PER U N I T

An estimate for the United States as a whole of the distribution of
family incomes is being prepared by the National Resources Committee
on the basis of the extensive sample studies covering large, mediumsized and small cities, villages, and farm areas in six general regions',
made for the Study of Consumer Purchases in the United States during
1936-1937. This estimate is not yet available at the time of writing l .
But the basic materials for a few cities and regions have been made
.-1 The only other comprehensive estimates covering family incomes are
those contained in the study of America's Capacity to Consume, by LEVEN, M.,
MouLTON, H. G. and WARBURTON, e , published by the Brookings Institution,

— 65 —
available x a n d are utilised here as illustrative d a t a in t h e
of i n c o m e a n d e x p e n d i t u r e w h i c h follows.

discussion

TABLE I.
CUMULATIVE PERCENTAGES OF NON-RELIEF WHITE AND
NEGRO FAMILIES ACCORDING TO ANNUAL INCOME PER CONSUMPTION UNIT
IN SELECTED AREAS, UNITED STATES, 1936
White population

Annual income
per
consumption unit i

New
York
City

Portland

.
.
.
.
.
.
.
.
.
.
.
.
.
.

0.5
1.4
3.0
6.6
13.6
24.5
36.5
45.6
54.2
63.2
71.7
81.9
87.9

0.0
0.3
2.2
5.9
11.4
22.2
36.8
49.9
60.8
67.1
75.7
83.3
90.7
94.2

0.0
0.5
2.8
7.0
14.5
23.8
36.2
47.7
56.2
63.3
71.1
78.2
86.8
91.9

. . . .

100.0

100.0

100.0

(dollars)
Less t h a n

,
,
,

50
100
150
200
300
400
500
600
700
800
900
1,000
1,200
1,400

T otal

.
.
.
.
.
.
.
.
.
.
.
.
.
.

Negroes

Atlan- Southeast
ta
region

New
York
City

Atlan- Southeast
ta
region

0.3
5.4
32.7
58.9
80.4
92.5
97.6
99.6
99.7
99.9
99.9
100.0
100.0
100.0

0.1
1.8
5.5
14.5
26.6
48.6
64.9
77.2
80.9
88.5
93.3
97.5
98.4

0.3 • 1.4
3.9
17.4
18.3
55.4
38.6
87.7
56.8
95.0
75.4
99.3
88.0
99.8
92.6 100.0
95.2 100.0
95.9 100.0
97.6 100.0
98.6 100.0
99.2 100.0
99.7 100.0

100.0

100.0

100.0

100.0

i In calculating this table, the annual income per consumption unit for each family
was calculated as follows: the annual income corresponding to the mid-point of the income
group in which the family was classified (for the lowest income class ($0-250) the average
was estimated at $200) was divided by the number of consumption units in the family
as estimated from the family type in accordance with the following schedule (each adult
16 years of age and over counted as one " unit " and each child under 16 as half a " unit ") :
Nïimhpr
Number
Family type
ofpSsons of consumption
1. Husband and wife
2.0
2.0
2. Husband, wife, and one child
3.0
2.5
3. Husband, wife, and two children
4.0
3.0
4. Husband, wife, one person 16 or over, and one or
3.5
3.25
no other person of any age
5.5
4.25
5. Husband, wife, one person 16 or over, one child,
5.5
3.75
and one ortwo more persons of any age . . .
7.5
5.875
6. Husband, wife, and three or four children
. . . .
7.0
6.0
7. Husband, wife, one child, and.four or Ave other
persons of any age .
W a s h i n g t o n , 1934. However, for present purposes of showing the distribution
Others.according
All other family
types income per unit, these d a t a fail to meet the
of families
to available
r e q u i r e m e n t s , since no analysis is made according t o t h e size of families
receiving each income, except for a distinction between persons living alone
and other families. F u r t h e r m o r e , the d a t a are for 1929, a year when
n a t i o n a l income reached i t s highest point, a n d are therefore n o t applicable
to the years of more modest national income which followed; and finally the
picture does n o t separate the wage-earning group from those in other
occupational classes. Though the s t u d y utilised all t h e materials available
a t the time for its estimates, the d a t a on certain points were extremely meagre
as c o m p a r e d w i t h the more a b u n d a n t material available in the Study of
Consumer
Purchases.
1
T h e U . S . B u r e a u of Labor Statistics furnished d a t a on incomes, family
composition and expenditures for New York City, P o r t l a n d , and A t l a n t a ;
s. L.
5

— 66 —
Table I presents a distribution of family incomes per unit for selected
samples of the population (exclusive of relief groups) in New York City,
Portland (Oregon), Atlanta (Georgia), and the share croppers of the
south-eastern region.
It appears that in the sample the median family income per unit per
year for native whites in New York City fell between $700 and $800,.
in Portland (Oregon), and in Atlanta it was slightly above $600, while in
the south-eastern region it fell between $150 and $200 a year. For the
Negroes in New York it was just over $500, while it fell below $300in Atlanta and below $150 in the south-east region 1.
B.

FOOD CONSUMPTION

AND

NUTRITION

What do these incomes per unit mean in terms of food consumption;
and adequacy of nutrition ?
For this purpose data are available for the white families (non-relief)
in the same cities 2 ; and for the white farmers (also non-relief) in the
south-east, showing the analysis and appraisal of a sample of actual
families in relation to expenditure on food. This material shows the
ercentage of families which at each level of food expenditure per unit
ad diets which the Bureau of Home Economics characterised as " good "
or " very good " (corresponding approximately to the London standards).
On the basis of the data for families in New York City, the following
table has been set up :

E

TABLE I I . — RELATION B E T W E E N FOOD E X P E N D I T U R E P E R UNIT AND
ADEQUATE D I E T : BASED ON A SAMPLE OF W H I T E ( N O N - R E L I E F ) FAMILIESIN N E W YORK CITY

Food expenditure
Percentage of families
per unit
with good or very
per year
good diets
Up to $ 75
0
106
10
155
25
195
50
270 and over
100
This table indicates therefore that in the sample of families whose
diets were studied, about half of those spending at the rate of $195 per
unit per year were able to achieve a good or very good diet — in other
the U.S. Bureau of Home Economics furnished comparable data for farm
share croppers in the south-east. Analyses of nutritive adequacy of diets
were made by the Bureau of Home Economics on the basis of records of actual
weekly food consumption obtained partly by the Bureau of Labor Statistics.
and
partly by the Bureau of Home Economics.
1
These conclusions of course are advanced subject to reservations,
the chief of which perhaps is the condition derived from the fact that thefigures are based upon a sample of the population, and, though made with
due care and precaution, the values of a sample are always subject to fluctuations due to possible bias in method of selection or to chance. In regard
to these questions reference must be made to the sources in which the figures
will
be published.
2
For Portland, data for several Pacific Coast cities were used; for Atlanta,.
data for several south-east cities combined were used.

— 67 —
words adequate nutrition. While the families in the sample which were
willing to supply food records may have had superior managing ability,
it seems a reasonable assumption for present purposes t h a t on the average
all families with an equal amount to spend on food might be able to obtain
equally good results, at least with proper educational measures covering
the purchase and utilisation of foods \
This material, in conjunction with the data of table I suggests t h e
ossibility of an interesting generalisation. If food expenditure can
e inferred from the average income per unit, then the proportion of all
families having incomes per unit sufficient to purchase adequate
diets can be estimated, or the proportion of families with adequate or
inadequate diets can be estimated for each incorae-per-unit group. The
correlation between these two, food per unit and income per unit, is
also available from this basic material.
Without detailing all the steps in the procedure the general results
for New York City are presented in table I I I in the form of estimated
percentages of families (and of persons and of children) in each of the
five food adequacy groups. 2 These groups are defined in terms of the

E

1
In the analysis of diets the number of units in particular families was
calculated according to the food expenditure scale of the Bureau of Home
Economics, but in table I above, an approximate consumption unit has been
used; and the assumption is made throughout the present analysis that this
approximate unit gives substantially the same results as the more refined
unit.
2
For the families studied in each area, as described above under B (p. 66),
the percentage of families having good and very good diets was determined for
each level of food expenditure per " food-expenditure-unit ". These results
were plotted and a free-hand smooth curve drawn through the points to show
the relation between food expenditure per " food-expenditure-unit " and percentage of families with good or very good diets.
From this smooth curve points of food expenditure per " food-expenditureunit " were ascertained at which zero, 25, 50 and 100 per cent, of the families
had good or very good diets. These points were taken as boundaries of the
five classes described in table III, the lowest of which had no families with good
or very good diets, the highest had 100 per cent., and the intermediate classes
different proportions of good and very good diets. (Table II shows the results
for New York City.)
The other steps in the procedure were: for a sub-sample of the whole sample
on which table I is based the annual food expenditures per consumption unit
were determined for each income per consumption unit class; in calculating
the amounts per consumption unit the same assumptions were made as in
making the distributions according to income per consumption unit (see footnote to table I). The analysis actually was made for each family type. It was
found that within each population group the same amounts were spent per
consumption unit on food at any given income per consumption unit class,
independent of family type. Very appreciable differences, however, were
found between different population groups as regards the amount spent per
consumption unit on food at any given income per consumption unit. Thus,
at the income level of $300 per consumption unit per annum, white families
spent on food about $170 in New York City, $130 in Portland (Oregon),
and $120 in Atlanta (Georgia). Negro families at the same income per consumption unit of $300 had food expenditures per consumption unit of about $115 in
New York City, and $90 in Atlanta; while both white and Negro share croppers
spent about $140 (including the local retail value of home-produced foods).
The results were plotted in graph form and a free-hand smooth curve drawn
through the computed points to show the relation between food expenditure
per consumption unit and income per consumption unit.
To permit of applying the boundary points or class limits of the five classes,
which were in terms of food expenditures per food-expenditure-units, to the

— 68 —
estimated proportions of families within each group, with good or very
good diets, namely: I, None; II, 1 per cent, to 25 per cent.; I l l , 25 per
cent, to 50 per cent. ; IV, 50 per cent, to 99 per cent. ; V, 100 per cent.
A rough estimate on the basis of these probabilities, of t h e proportion of
the total of each population group which presumably had good or very
good diets is shown in the last column of the table.
The figures indicate t h a t in New York City over half of t h e native
white families, 46 per cent, of the persons and two-fifths of the children
fell in group V (were certainly well nourished), while 19 per cent, of
families and 29 per cent, of t h e children fell in one of the three lowest
groups, namely those in which the chances of achieving an adequate
diet were less t h a n one-half 1 . On the basis of the chances for an
adequate diet at each level, it has been roughly estimated, as shown
in the last column of table I I I , t h a t 79 per cent, of the families, 76 per
cent, of the persons and 73 per cent, of the children in the total native
material available in table I, which shows a distribution of families by income
per consumption unit, two further steps were necessary. First, the assumption
was made that the class limits for the five groups in terms of food expenditures
per food-expenditure-units yield for present purposes approximately the same
results as would class limits in terms of food expenditures per consumption
units (calculated as described in the footnote to table I). Secondly, an estimate
of the class limits in terms of income per consumption unit was made on the basis
of the relationship, shown in the smooth curve described above, between income
per consumption unit and food expenditure per consumption unit. The assumption was made that these class limits in terms of income per consumption unit,
though developed from part of the sample only, can be applied to the whole
of the sample used as the basis for table I.
The results are put in the form of incomes per consumption unit corresponding to points at the boundaries of the five classes noted above.
(In applying the points of food expenditure per consumption unit obtained
as above (third paragraph) to the population groups, an adjustment had to be
made for Portland (Oregon) and Atlanta to take account of the fact that the food
study was for the Pacific coast and the south-east region respectively. The
adjustment was made by means of price differences ascertained in a W.P.A.
study of inter-city differences in costs of living. In the three cities a slight
adjustment was made to take account of price changes between the time of
the food study and that of the budget study.)
Finally, the relative numbers of families, persons, and children at each food
adequacy level was determined by placing the distributions shown in table I
in relation to these income per consumption unit boundaries of the five classes.
A final estimate of the proportion of good and very good diets in each
population group was obtained by summing the estimated proportions of good
and very good diets in each of the five classes.
For the Negro population groups no estimates of diets are shown in the table.
Substantial differences are found in the white and Negro families in food
purchases—Negro families spend a much smaller proportion of their budgets
on food than do white families at the same income level. Since special data
on the adequacy of the diets resulting from a given food expenditure by Negro
families were not yet available at the time of the study, the proportions of
good and very good diets could be calculated only on the assumption that the
same proportions found for white could be applied to the Negro families,
i.e. that a given food expenditure by Negro families would give the same
proportion of adequate diets as in white families. On this assumption very
much smaller proportions of adequate diets were found for the Negro than for
the white families.
1
If an income equivalent to that at which half the families did achieve
good or very good diets may be regarded as an income sufficient to buy an
adequate diet, this means that the three lowest groups had inadequate incomes
from this standpoint.

— 69 —
white population sampled in New York City were well nourished.
The relatively greater predominance of persons and especially of children
in the lower food adequacy groups is explained by the greater proportion
of large families at the low income per unit levels and is particularly
significant for the future health of the population. The same situation
was found to prevail also in all the other regions studied.
When, however, separate figures were computed for families of white
wage earners (i.e. skilled, semi-skilled and unskilled manual workers),
the proportions with adequate diets were lower in all categories. Thus
only 35.9 per cent, of wage earners' families, as contrasted with 52.2 per
cent, for all occupations, were estimated to fall in group V, and the
proportion of children of wage-earning families in group V was only
24.4 per cent, as compared with 39.5 per cent, for all occupations.
TABLE III. — LEVELS OF FOOD ADEQUACY, WHITE FAMILIES,
NEW YOKK CITY, 1936-1937
Level of food adequacy

Item and population
group

Chances of being well
nourished

Total
population
studied

—

Approximate amounts
of a n n u a ] food expenditure per unit
a t which
chances
were as indicated
above
among the
W h i t e population .

I.

II.

III.

IV.

Certainly
ill
nourished

Probably
ill
nourished

Possibly
ill
nourished

Probably
well
nourished

BeBeBetween tween tween
None none
%
and Vt and y 2 a n d l / 1

Percent.
of total
estimCerated
tainly to be
well
well
nour- nourished ished
V.

1/1

—

U p t o $ 75 $ 1 5 0 $ 200 $ 2 6 0
$ 75 to 150 to 200 t o 260 and
over

Percentage falling in
each food adequacy
group1 :
All occupations:
Families . . .
Persons . . .
Children . . .

100.0
100.0
100.0

0.5
0.6
.0.7

5.2
6.5
7.3

13.3
16.2
19.2

28.8
30.4
33.3

52.2
46.3
39.5

79.4
76.0
72.6

W a g e earners only :
Families . . .
Persons . . .
Children . . .

100.0
100.0
100.0

0.2
0.3
0.3

7.4
9.5
10.7

19.6
22.8
26.3

36.9
37.0
38.3

35.9
30.4
24.4

71.9
67.9
64.3

i i.e. having total incomes per unit corresponding to the indicated range of food
expenditure per unit.

Following for other areas methods similar to those used in New York,
striking regional differences are found. The summary figures for.

— 70 —
Portland (Oregon), Atlanta (Georgia) and south-eastern share croppers
are presented in table IV. It will be seen even from these limited and
approximate figures that the proportions pi the populations in the lower
adequacy groups were much higher in the other regions, with the result
that the proportions of all families estimated to be well nourished
dropped to 54.6 in Portland and to 27.0 for share croppers in the
south-east. Percentages for persons and children were still less. Among
the share croppers, no families at all fell in group V. If the point at
which 50 per cent, of families achieved adequate diets be taken as the
income sufficient to buy such diets, it is seen from the lower part of
table IV that especially large proportions of share croppers fell into
this category.
TABLE

IV.

LEVELS OF FOOD

ADEQUACY, W H I T E

FAMILIES,

PORTLAND, ATLANTA, AND SOUTH-EASTERN FARMS, 1 9 3 6 - 1 9 3 7
All occupation groups
Item and population
group

Portland

s

Atlanta

Share croppers
South-eastern
farms

Percentage of total population estimated to be well
nourished 1 :
Families
Persons
Children

54.6
52.0
48.5

81.0
76.9
42.1

27.0
23.7
22.6

Percentage of total population falling in adequacy
groups I, II and I I I
combined 2 :
Families
Persons
Children

50.0
54.9
61.2

16.9
21.6
27.3

83.4
88.1
90.3

i Figures corresponding to the last column in table III.
» i.e. having total incomes per unit corresponding to ranges of food expenditure per
unit indicated above and with chances of being well nourished of less than ¡4.

It must be borne in mind that the figures shown in tables III and IV
are all rough approximations and are subject to a number of important
reservations. Chief of these is the fact that the food adequacy
groups do not show positively the proportion of good and very good
diets. They are based rather upon the probabilities for families in
each group of having adequate diets as explained above. Nevertheless,
a rough estimate may be made, on the basis of these probabilities, of
the proportons which in each group presumably had good or very good
diets. These estimates are shown in table III in the last column and
in table IV as the first item. The other reservations which should be
mentioned include:
(1) Except in the case of New York, the figures relate to the native
white population as a whole, without distinguishing wage earners.
(2) In establishing the percentages of families which actually achieved
adequate diets, no distinction was made between manual workers
and clerical workers.

— 71 —
(3) The analysis of the diets follows a classification according t o
food expenditures, whereas the analysis of families is made
according to income available rather t h a n expenditures made.
(4) The analysis covers only non-relief families; in other words in
this respect it tends to be too favourable.
In interpreting the results of table I I I the implications of the method
of classification must be borne in mind. The lowest level was defined
in terms of the absence of any families with good or very good diets;
on t h a t level inadequate nutrition may be said to be exclusively a
problem of income.
The importance of income as a cause of malnutrition diminishes as
the higher adequacy levels are successively considered, until, at the
highest, it entirely disappears. On the intermediate levels other factors,
such as lack of skill in management, etc., play an important part. On
t h e IVth level, for instance, where more t h a n half of the families had
adequate diets, failure to obtain an adequate diet may have been
due in part to other causes t h a n insufficient income. Evidently,
therefore, there is much room at this level for improvement by means
of appropriate advice and educational measures.
A p a r t from t h e general d a t a already presented on the proportion
and amount of income spent for food and inferences on the adequacy
of diets in each income class, limitations of space do not permit any
analysis of the actual foods consumed by workers in the United States,
or details of food consumption on different economic levels. Some
details on this point are given in Chapter I I I of Workers'
Nutrition
and Social Policy l .
Space is not available for any general discussion of malnutrition.
But a brief reference may be made to the statistics of deaths from
pellagra. In 1935 in the United States there were 3,543 deaths from
this " deficiency disease ". Pellagra is largely confined to ten States
of the eastern and central South, where the death rate from this disease
averaged 10.4 per 100,000 of estimated population 2.

C.

HOUSING

An appraisal of housing in the United States on the basis of four tests
gives the following results: six million non-farm homes and five million
farm homes in the United States, or over a third of all the homes of the
nation are estimated to be definitely sub-standard 3. Houses reported
1
Additional data permitting considerably more elaborate analyses will
shortly be published in the United States by the Bureau of Labor Statistics
and the Bureau of Home Economics. The data were obtained in connection
with the Study of Money Disbursements of Families of Wage Earners and
Clerical Workers conducted by the Bureau of Labor Statistics in 1934-1936 in
representative large cities and also from the Study of Consumer Purchases
conducted co-operatively by both Bureaux in 1936-1937 in representative urban
and rural areas in six regions of the country.
2
Data on deaths from Mortality Statistics, 1935. U.S. Bureau of the Census,
Washington, 1937. Estimated populations from Statistical Abstract of the
U.S., 1937, p. 9. The problem of pellagra is treated more fully in a forthcoming report of the National Resources Committee.
3
Housing is considered sub-standard " if each family is not furnished with
an ample and pure supply of running water, with an indoor flush toilet for its
exclusive use, with a bathtub or shower, and if, in a built-up community, it

— 72 —
as needing structural repairs to foundations, roofs, walls, etc., or as
wholly unfit for h u m a n habitation constituted 18 per cent, of all t h e
houses surveyed in 64 cities in 1934. The proportion ranged from
39.5 per cent, in Paducah (Kentucky), to 8.3 in Syracuse (N.Y.) 1 .
The following table indicates the situation with respect to crowding
and absence of certain facilities for 64 cities covered in the Real Property
Inventory.
TABLE V. —

CROWDING AND HOUSING FACILITIES

1

(64 representative cities in the United States, 1934)
Measure of crowding or type
of facility
Dwelling units with:
3 rooms or less
More than one person per room 2
No running water
Neither gas nor electric light
No private indoor water closet .
Neither bath tub nor shower

Percentage of dwelling
units in 64 cities
20.7
16.8
5.0
8.1
13.5
20.2

i S O U R C E : Ibid., p . 82.
2 Including occupied dwellings only.

The conditions summarised in t h e foregoing table varied considerably
from city to city. Certain cities showed high percentages of dwelling units
lacking the various facilities. Furthermore, the preponderance of overcrowding and of absence of various facilities in the given cities coincide
with the lower categories of monthly rental.
TABLE V I . —

OVERCROWDING AND HOUSING FACILITIES, SELECTED

CITIES, 1934

Providence, Portland, Charleston, Atlanta,
Oregon
R.I.
Ga.
S.C.

Item
Number of dwellings

1

. . .

Per cent, of dwellings with:
More than 1 person per
room 2
No private indoor toilet
No bath tub or shower
No gas or electric light
No running water . .

66,734

18.8
3.1
20.4
2.3
0.1

96,225

7.3
. 6.4
8.4
1.3
1.1

17,911

72,266

40.4
48.7
56.2
50.0
21.7

30.3
20.6
31.9
29.4
8.1

i S O U R C E : Real Property Inventory, U.S. Bureau of Foreign and Domestic Commerce.
Including occupied dwellings only.

2

is not connected with the sewer system ". The estimate was prepared by
counting houses for which the rental value shown in the 1930 Census would
not obtain adequate housing under average conditions of building costs, etc.
WOOD, Edith E.: Slums and Blighted Areas in the United Slates. Federal
Emergency Administration of Public Works, Housing Division Bulletin
No: 1. Washington, 1935, pp. 4-6.
1
Ibid. Data from the Real Property Inventory, a housing census taken in
64 representative cities, conducted by the U.S. Bureau of Foreign and
Domestic Commerce.

— 73 —
The situation is worse with respect to farm homes, where distances
between dwellings make it more costly t o provide sewer, electrical and
similar connections. Over half the sample of the farm homes in 36 out
of 46 States studied in the Farm Housing Survey were without piped
water 1. In certain remote mountainous regions in the United States, for
example the southern Appalachians, the conditions of living are much
worse. In a study 2 of family living in K n o t t County, Kentucky, in
1930, over 40 per cent, of the 228 farm homes visited were composed
of 3 rooms or less, and in 70 per cent, of those studied there was more
t h a n one person per room. Only 4 of the families had running water or
a p u m p in the house; all the others carried water into the house from
wells or springs. 51 per cent, of the families studied had no toilet facilities
of any kind, not even an outdoor privy, and only one family had an inside
flush toilet with a tile drain to a stream. The water supply was frequently
situated too close to the privy. In 84 cases families reported drainage
from t h e privy to t h e water supply and in 42 drainage from the manure
pile to the water supply.
Conditions of living among Negro families in t h e south-east are
distinctly worse than among the white families, their incomes being
lower 3 .
D.

HEALTH

The extent of illness is indicated by the estimate for the population
as a whole, t h a t on an average four million persons (out of a total of
about 130 million) are disabled by illness 4 .
Data for t h e country as a whole and from certain special studies serve
to suggest the close relationship between income and health. The
National Health Survey showed t h a t among the urban population illnesses
disabling for one week or longer in a twelve-month period occurred
among families on relief at a higher, rate t h a n among families with a
higher economic s t a t u s 6 .
The rate of infant mortality has also been found to be higher at the
lowest income levels. A study in seven cities (data for 1913-1915) showed
mortality rates of infants classified according to per capita income from
their fathers' earnings to rise from 60.5 for those in t h e " $400 and over "
1
The survey in 1934 by the U.S. Bureau of Home Economics covered farm
homes in 352 counties of 46 States.
2

WILLIAMS,

F. M.,

STIEBELING,

H.

K.,

SWISHER,

I. G.,

and

WEISS,

G. S. : " Family Living in Knott County, Ky.". U.S. Dept. of Agriculture,
Technical Bulletin No. 576, August 1937.
3
See data from Farm Housing Survey quoted by WOOFTER, T. J., and
others. Landlord and Tenant on the Cotton Plantation. Works Progress Administration, Division of Social Research. Research Monograph V, Washington,
1936.
4
Estimated in a report submitted by the Technical Committee on Medical
Care to the Interdepartmental Committee to Co-ordinate Health and Welfare.
Quoted in " Need for National Health Program ". U.S. Dept. of Labor,
Bureau of Labor Statistics, Monthly Labor Review, Vol. 46, No. 6, June 1938,
p. 1369.
6
The National Health Survey : 1935-36. " Illness and Medical Care in relation to Economic Status ". Preliminary Reports, Sickness and Medical Care
Series, Bulletin No. 2. U.S. Public Health Service, Washington, 1938. (Mimeographed) p. 1. See also SYDENSTRICKER, Edgar, Health and Environment,
New York, 1933, for other data regarding the relation of health to income.

— 74 —
group to 215.9 in the " less t h a n $50 " group 1 . In a Cleveland study in
1928 the rate was 86.6 deaths per thousand live births in the lowest
economic group as contrasted with 35.7 in the highest 2 .
W i t h regard to relative incidence of illness among Negroes and whites,
a study in New York City of Harlem Negroes and Lower East Side
whites, of roughly similar occupational status, employment and incomes,
showed an excess of disability illness rates (excluding epidemic diseases)
for adult Negroes classed as of " poor " and " moderate " economic
status. In the group as a whole the sickness rate for Negro females
15 years of age and over was 36 per cent, higher t h a n t h a t observed
among white women of comparable ages in the Lower East Side 3. Data
from the National Health Survey, however, classifying Negro and
white families according to whether or not they were on relief, show a
slightly lower disabling illness rate for Negroes on relief t h a n for whites.
The report of the study concludes : " Low economic status, rather t h a n
inherent racial characteristics in the reaction to disease, thus appears to
account in large measure for the higher disability rate observed among
Negroes " 4 (relief and non-relief together).

E.

LITERACY

5

The simplest measure of educational achievement, literacy statistics,
shows the United States to have improved in the period between the last
two Censuses. In 1930. 4.3 per cent, of the population ten years of age
and over were unable to write in any language, as contrasted with 6.0 per
cent, in 1920. Of the native white population of native parentage
aged from 10 to 20 years, 99.2 per cent, were literate in 1930; the percentage reached 99.8 per cent, in the Pacific and East North Central
regions.
The problem of illiteracy is more serious for rural t h a n for urban
areas (6.0 per cent, as compared with 3.2 per cent.), since the provision
of school facilities in the former is more difficult, more costly per capita,
and the school attendance laws more difficult to enforce. In particular,,
in certain rural mountain regions of the South-eastern States a considerable proportion of illiterates is found even among the native Whites.
Special problems are found also for the Negroes and for Indians and
Mexicans. In the Southern States, separate schools must be provided
for Negroes and whites: the money available for education has to be
divided, and the result has been t h a t the school facilities, especially
those for Negroes, have suffered. This is reflected in the relatively
high proportion of illiterates among Negroes of 16.3 per cent, for the
country as a whole, and similarly high rates for a number of States
1

WOODBURY, R. M.: Infant Mortality and Its Causes. Baltimore, 1926,
p. 135.
2

3

SYDENSTRICKER, op. cit.,

p.

100.

HOLLAND, D. F., and PERROTT, G. St. J.: Health of the Negro. Reprinted
from the Milbank Memorial Fund Quarterly. Vol. XVI, No. 1, January 1938.
Part I, " Disabling Illness Among Negroes and Low-Income White Families
in New York City — A Report of a Sickness Survey in the Spring of 1933."
4
HOLLAND and PERROTT, op. cit., Part II. " A Preliminary Report on a
Study of Disabling Illness in a Representative Sample of the Negro and White
Population of Four Cities .Canvassed in the National Health Survey, 1935-36."
6
Statistical Abstract of the U.S., 1937. Washington, 1938, p. 41.

— 75 —
of the South where large proportions of the Negroes are found.
Indians and for Mexicans, the illiteracy rates are also high.

For

II. — Poland
Standards of living in Poland must be considered with reference to
the background of economic and population structure, as well as of its
history and customs. With the territory reunited to form the Polish
Republic following the war, Poland is the largest country in Eastern
Europe except the U.S.S. R. Rapid population growth is indicated by an
18 per cent, increase during the decade 1921 to 1931, with the total
population reaching 32 millions in the latter year.
Despite a slight increase in the percentage of urban dwellers between
the two censuses Poland remained in 1931 essentially an agricultural
country with 73 per cent, of the population dwelling in rural areas.
The urban population was found principally in the western part of the
country in the important mining and industrial regions of Silesia,
Krakow and the Dombrowa Basin, and in Warsaw, Lodz and Poznan.
The southern and eastern parts of the country were essentially agricultural.

A.

INCOME

The family budget surveys l of 1927, 1928 and 1929, though based on
small samples, give some notion of the income levels among families of
employed workers in the principal industrial cities prior to the Great
Depression. In all three years the group spending 600 to 899 zlotys 2
per year per consumption unit contained the largest number of families,
while approximately 15 per cent, of the families spent less t h a n 600 zlotys
and around 13 per cent, spent 1,200 zlotys and over per consumption
unit. The families of industrial workers willing t o keep household
accounts for a year and included in the surveys were, however, somewhat above the average in income 3.
1
SOURCE: Budgets des Familles Ouvrières. Résultats de l'Enquête effectuée à
Varsovie, à Lodz, dans le Bassin de Dombrowa et en Haute Silésie, 1927, Office
Central de Statistique de la République Polonaise, Tome XL, Fascicule 1,
Varsovie 1930; Ibid. 1928, 1929, Tome XL, Fascicule 2, Warsaw, 1933.
2
The Polish monetary unit, the zloty, is valued in 1938 at a little more
than four-fifths of the Swiss franc. The exchange rate on the Warsaw bourse
from 1928 to 1935 was approximately 172 zlotys per 100 Swiss francs, or
0.58 franc per zloty.
3
This conclusion is confirmed by a study of wages made by Dr. Ludwik
LANDAU: The Evolution of Wages in Connection with the General Economic
Conditions in Poland. Institute for Social Problems, Labour Market and
Unemployment Series, No. 4, Warsaw, 1933. According to this study, the
industrial worker's average monthly earnings were estimated at 172 zlotys
per month. The earnings of over 25 per cent, of all industrial workers were
below 100 zlotys per month. On the other hand, 25 per cent, were
estimated to earn over 230 zlotys per month and 12 per cent, over
300 zlotys, these latter figures approaching the earnings of office workers.
It may be concluded on the basis of very rough calculations that the annual
family income per unit of 25 per cent, of all industrial workers in Poland in
1929 was 600 zlotys or less. This estimate has been made on the following
assumptions, whose approximate and tentative character must be stressed:

— 76 —
For the city of Warsaw more precise data on income distribution are
available in the form of a sample survey in 1935 of 300 representative
families in working-class districts of Warsaw 1. The d a t a regarding
incomes of these families are presented in table VII as an illustration of
the distribution of families and of consumption u n i t s 2 according to income
per unit. The median family income was 8.6 zlotys per unit per week
or approximately 400 per year. More t h a n two-thirds of the families had
less t h a n 12.50 zlotys per week or 650 per year. It will, of course, be
observed t h a t the year 1935 was one of marked depression as compared
with 1929. The index of production for Poland, which stood at 100 in
1929, had fallen to 66 in 1935. Some or all of the members of approximately one-third of the families studied were unemployed at the time of
the investigation.
TABLE V I I . — CUMULATIVE PERCENTAGES OF FAMILIES AND
ACCORDING TO INCOME P E R CONSUMPTION UNIT

Total family income in zlotys
per week per consumption unit i
Up to

„
„
„
„
„

2.50
3.75
5.00
7.50
10.00
12.50
15.00
20.00
25.00

Total

UNITS

Cumulative per
cent, of families2

Cumulative per
cent, of units s

4.0
15.7
28.7
48.7
61.0
69.3
78.0
90.7
95.0
100.0

5.3
21.0
37.6
58.4
71.2
77.5
84.4
94.5
97.8
100.0

1
If the income is expressed in zlotys per year per unit, with intervals comparable to
those used in the family budget investigationsof 1927-1929, the correspondingpercentages
are:

Total family income in zlotys
per year per unit
Up to
„
2
3

600
900
1,200
Total

Cumulative per
cent, of families

Cumulative per
cent, of units

66
85
94
100

76
89
96
100

Ibid. Table 17 (b).
Supplied in a special communication to the International Labour Ofllce by Dr. Landau.

A feature of the Warsaw income distribution which corresponds with
t h a t found also in the other countries is t h a t the families with
(1) that sources of income other than earnings were negligible, an assumption
borne out by the facts shown in the budget studies; (2) that annual earnings
were twelve times monthly earnings; (3) that there was a rough average of
1.7 earners and 4.1 persons or approximately 3 consumption units per family.
1
Ludwik LANDAU: Unemployment and Standard of Life in Working-class
Districts of Warsaw, Institute for Social Problems, Labour Market and Unemployment Series, No. 8, Warsaw, 1936.
2
The consumption unit scale used in the study was the same as

— 77 —
greater numbers of children are found consistently in the lower income
per unit groups. T h e result as shown in column 3 of table V I I is
t h a t there is even greater concentration of consumption units t h a n of
families at the lower income per unit level. In other words, the largest
number of children are in t h e lower income-per-unit groups which are
in general just those groups where food consumption and living conditions
are most unsatisfactory.
Data with respect t o t h e incomes of t h e rural population of Poland
are even more fragmentary 1 . Estimates of t h e size of rural holdings
serve to give a general idea of t h e situation. It has been calculated
t h a t in 1938 in Poland there was a total of some 4,200,000 rural holdings ;
of these more t h a n 60 per cent, consist of less t h a n 5 hectares, which
are " holdings too small, often unfit to supply a farming family with
food and work in sufficient measure " 2.
Data on average returns per hectare of small holdings for t h e years
1931 to 1935 3 suggest t h a t for the great bulk of Polish peasant proprietors
t h e level of living is largely limited b y what t h e y are able t o produce
for their own consumption, i.e. b y subsistence farming.
B.

EXPENDITURE

An estimate of the meaning of the foregoing money incomes in terms
of the goods and services bought b y urban families is available for 1927,
that used by the Chief Bureau of Statistics in the family budget studies,
namely :
Consumption unit scale
Age in years
Men
Under

1

1
1
2
3-7
8-11
12-13
14-17
18 and over

Women
0.25
0.30
0.40
0.50
0.60
0.7(1

0.85
1.00

0.75
0.85

Data concerning the income of the rural population may be found in

publications of the INSTYTUT BADANIA KONJUNKTUR GOSPODARCZYCH I CEN:

Badania nad dochodem spolecznym w Polsce (Institute of Trade-cycle Research;
Enquiries on National Income in Poland), Vols. 1 and 4 (by M. Kalecki and
L. Landau), Warsaw, 1934 and 1935.
2
Calculations made by Czerniewski on the basis of sample studies
conducted by the Institute of Social Economy at Warsaw (Instytut
Gospodarstwa Spolecznego w Warszawie) and cited in Unemployment and
Under-nutrition of the Rural Population, a report presented to the Fourteenth
General Assembly of the International Institute of Agriculture in the name
of the Polish Government by Professor W. Staniewicz, May 1938, p. 4. As
a result of the agrarian structure in Poland, there is said to be a real problem
of rural over-population, especially in the southern region. Estimates recently
made by M. Josef Poniatowski of Warsaw, quoted by Professor Staniewicz,
indicate that approximately 4,000,000 of the 11,000,000 active farm workers
could without prejudice to Polish agriculture be directed to other occupations
or towards emigration.
3

CHIEF BUREAU

OF STATISTICS OF THE REPUBLIC

Statistical Year-Book of Poland, 1937, p. 62.

OF POLAND: Concise

— 78 —
1928, and 1929 in the budget studies previously referred to. For
families in three important industrial centres in 1929 the average
spending followed the trends shown in table VIII.
TABLE V I I I .

DISBURSEMENTS OF POLISH W O R K E R S F A M I L I E S , WARSAW,
LODZ, AND THE DOMBROWA BASIN, 1 9 2 9

Families spending per year
per consumption u n i t in zlotys
All
families

Item

Up to
599

600899

9001,199

1,200
and over

3.87

4.62

4.01

3.77

3.15

Total net disbursements:
Zlotys per family
Zlotys per unit

3,490
902

2,472
535

3,017
752

3,887
1,032

4,802
1,526

Per cent, for:
Food
Rent
Heat and light
Other household operation . .
Clothing and shoes
Other disbursements 1 . . . .
Total disbursements . .

54.3
4.0
4.7
2.5
16.9
16.6
100.0

64.4
3.9
5.9
1.9
14.8
9.1

59.8
4.5
4.7
2.9
15.3
12.8
100.0

52.6
3.8
4.6
3.5
17.2
18.3
100.0

43.1
3.5
4.2
5.1
19.6
24.5
100.0

Average number of consumption
units per family

1

100.0

Greater detail is available in t h e original report.

S O U R C E : Budgets des Familles Ouvrières. Résultats de l'Enquête effectuée à Varsovie, à Lodz,
dans le Bassin de Dombrowa et en Haute Silésie, 1928-29. L'Office Central de Statistique
de la République Polonaise, Tome XL, Fascicule 2. Warsaw, 1933.

It will be seen from the foregoing table that there are important variations in spending habits at different economic levels. Thus, the percentage
for food showed a sharp tendency to decline at higher levels, while that
for rent, plus heat and light, declined only slightly. On the other
hand, the percentages devoted to other household operations and to
educational and cultural expenses increased markedly as incomes permitted. Though not shown in the table, there are important regional
differences in spending habits among the three areas covered.

C.

FOOD CONSUMPTION

Details regarding the average diets of urban Polish workers' families
(based on the family budget studies of 1927-1929) have been presented in
Chapter III of Workers'1 Nutrition and Social Policy. They show
the predominant role of rye bread and potatoes—consumed by all the
families, even those at the higher expenditure levels. They also show the
strong tendency towards an increase in the consumption of fresh vegetables and fruits, meat, fish, and dairy products, as incomes increase.
The data on the nutrition values of these average diets for three industrial

— 79 —
regions presented on page 68 of Workers' Nutrition and Social Polio/
further suggest t h a t particularly on t h e lower expenditure levels these
diets do not meet t h e optimum food requirements of t h e London
standards. The low quantities of dairy products, fruits, and vegetables
(other t h a n potatoes) consumed b y t h e urban Polish families suggest
the danger of insufficient protective foods.
From t h e data on income distribution for Warsaw, it is possible t o
make a somewhat more detailed estimate whether incomes are sufficient
to buy adequate diets. In 1935 the cost of a worker's diet tentatively
considered as adequate * in Warsaw was estimated a t 0.928 zloty per
day or 6.50 zlotys per consumption unit per week. If a very rough
estimate be made t h a t food expenditures would tend to constitute
50 per cent, of total expenditures, it will be seen from the income table
on page 76 t h a t about 71 per cent, of the families and 79 per cent, of the
consumption units in families in representative working-class districts
in Warsaw in 1935 had incomes insufficient t o be associated with a
food expenditure large enough to b u y such a diet. If it be assumed
t h a t an adequate diet could, by careful planning and management,
have been achieved in Warsaw in 1935 for almost half t h e figure
previously used, or 0.50 zloty per unit per d a y (see footnote)
or 3.50 per unit per week, and t h a t food constituted 65 per cent, of
total expenditure, it would still be found t h a t approximately 29 per
cent, of t h e families and 38 per cent, of t h e consumption units in
representative Warsaw working-class districts had incomes insufficient
to purchase such diets 2.
With respect to diet in rural Poland, there is, correspondingly,
evidence of malnutrition, especially among large families living on
small holdings of from 2 to 5.5 hectares. An enquiry regarding farm
family food consumption in four important agricultural regions in
densely populated areas was conducted for the period 1931-1933 by t h e
Economic Section of t h e Pulawy Scientific Institute 3 . The findings
1

STATISTICAL

MUNICIPAL

BUREAU

OF WARSAW:

Kronika

Warszawy,

No. 11,1930. The nutritive value of this diet is as follows:
Calories: total number
3,011
from food of animal origin
977
„ „ plant
„
2,034'
Proteins:
78 grammes
from food of animal origin
42 gr.
» » Plant
„
36 „
Fats
81 grammes
Carbohydrates
470 grammes
Attempts have been made by the Social Trade and Domestic School for
Women in Lodz to work out menus adapted to the purchasing power of
working-class families. Such menus, costing from 0.42 to 0.49 zloty per
unit per day, have been worked out on the basis of about 2,000 calories,
70 grammes of protein, and 40 grammes of fats.
2
The rough character of the foregoing calculations must be stressed in
view of the tentative nature of the assumptions enumerated above. For
analysis of actual Polish food consumption see various publications of Dr. Emit
Paludi; and A. Szczygiel i Z. Czerny, W sprawie akcji dozywiania bezrobolnych.
Warsaw, 1937. (Food Consumption of the Unemployed.)
3
Z. GRONOWSKA i W. OBREBSKI: Spozycie rodziny w gospodarstwach
karlowatych.
(Biblioteka Pulawska, Seria prac spoleczno-gospodarczych,
Nr. 82.) Warsaw, 1937. The findings of this study have been summarised

— 80 —
show failure to reach London standards for calorie requirements, though
protein consumption was adequate. Considerable regional variations
were found as well.
D.

HOUSING

One method of estimating housing deficiencies is to ascertain the cost
of housing which meets requirements set down by official or semipublic agencies on the basis of Polish living habits and reasonable
relation to the workers' ability to pay. The standards set down by the
Workers' Housing Association (T.O.R.) present a basis of comparison
for urban Poland. This Association, subsidised by the Government,
has undertaken the construction of modest workers' dwellings, with the
provision of the maximum facilities in the way of running water, central
heating, etc., which can be brought within range of the workers' ability
to pay. A typical dwelling 1 constructed by this association is called
a one and one-half room apartment, with central heating and sometimes
with an elevator, consisting of a hall or antechamber, a principal room,
a toilet and] a kitchen with gas, electricity, and running water. Such
a dwelling in the subsidised apartment houses in Warsaw in 1935 2
could be rented for approximately 25 zlotys per month. These model
workmen's flats, which provide central gardens and play space for
children, are, however, available only for a small portion of the workingclass population. Two-room flats, without bath, and generally without
toilets, in private buildings in 1935 could be rented for much less in
old buildings t h a n in new buildings constructed since the war and
subject to no rent restriction. The rent in the old buildings of a
two-room flat in Warsaw in 1935 was approximately 30 zlotys per
month in certain districts and 20 to 25 zlotys 2 in suburbs. If, then,
for purposes of illustration, it is assumed t h a t roughly 25 zlotys per
month per family 3 was necessary to rent an adequate dwelling in Warsaw
in 1935 and t h a t families spent as much as 10 per c e n t 4 of their income
for rent, it will be seen from the data on income distribution on page 76
t h a t 91 per cent, of the representative families in Warsaw workingclass districts had incomes insufficient to meet such expenditure. If it
is assumed t h a t rent constituted as much as 20 per cent, of total expenditure in 1935, then 61 per cent, of the families and 71 per cent, of the
consumption units would fall into the group with incomes inadequate
to meet this expense 6.
in English in a report on Unemployment and Under-nutrition of the Rural
Population, presented in the name of the Polish Government by Professor
W. STANIEWICZ to the International Institute of Agriculture, May 1938.
1
See Jan STRZELECKI: La question de l'habitation urbaine en Pologne,
Société des Nations, Organisation d'Hygiène, C.H.1165(c), Geneva, 1936.
2
According to a communication from Dr. Ludwik Landau.
3
With the average of 2.89 consumption units per family found by Dr.
Landau in the study quoted earlier, this would constitute a weekly rental of
approximately
2.00 zlotys per consumption unit.
4
In the budget study of 1929 the average expenditure for rent in Warsaw
was only 6 per cent. ; since incomes have decreased from 1929 to 1935, probably
more
proportionately than rents, this increase is a reasonable assumption.
6
The specific proportions of families or consumption units found to have
incomes inadequate to purchase minimum housing of course depend upon
the particular grade of house selected and whether its rental is taken at a
subsidised or a market rate, and on the relation between income and housing
expenditure found in fact to exist.

— 81 —
In view of the tentative character of the foregoing analysis, t h e problem of adequacy of housing of Polish workers m a y be examined on the
basis of data regarding overcrowding and housing facilities of all houses.
The situation with respect to crowding is seen from t h e fact t h a t according to the 1931 census 1 37 per cent, of urban dwellings and 51.5 per cent.
of rural dwellings were composed of 1 room. The average number of
persons per room in these dwellings was 3.9 in urban areas and 4.8 in
rural. An additional 32 per cent, of urban dwellings were composed
of 2 rooms, inhabited by an average of 2.3 persons, while two-room
dwellings constituted 35.5 per cent, of rural dwellings and contained
an average of 2.7 persons per room. Taking more t h a n two persons per
room as the definition of overcrowding, 45 per cent, of all urban dwellings
and 72 per cent, of one-room urban dwellings were overcrowded 2.
The 1931 census also showed t h a t for all urban dwelling houses,
only 13 per cent, had sewer connections, 16 per cent, had running
water, 38 per cent, had electricity, 7.5 per cent, had gas, while the combination of sewer connections and running water and gas or electricity
were present in 10 per cent, of the dwelling houses. This means t h a t
90 per cent, of the urban dwelling houses lacked the foregoing combination
of facilities. The foregoing discussion has been illustrative of the use
of data for the entire population. It m a y be legitimately inferred t h a t
most urban workers are found in the one and two-room dwellings
whose characteristics were described above.
Comparable data for rural areas are not available, but it may be
said in general t h a t very few farm houses are equipped with any of
these facilities. Legislation has recently been passed to improve the
housing conditions of agricultural labourers 3 .
Housing conditions
among small peasant proprietors are also to be improved in connection
with the programme of agrarian reform under way in Poland *.
In the urban areas certains groups of the population, particularly
the unemployed, live in dwellings where the overcrowding, absence of
furniture, and lack of facilities are particularly marked 5.

1
In this connection it must be recalled that there was a post-war housing
shortage in Poland and that in the decade 1921-1931 the population increase
was not matched by a corresponding increase in housing construction. Therefore
the figures for the 1931 census present a situation worse than that in 1921.
2
SOURCE : Deuxième recensement général de la population du 9 décembre 1931.
" Logements et ménages." Statistique de la Pologne, Série C, Fascicule 62,
Warsaw, 1937.
3
Thus a recent Decree requires that dwellings furnished to agricultural
labourers and foresters with families must consist of at least one room of
specific minimum dimensions and a smaller half-room. Single men and women
labourers must be lodged separately. The Decree also contains specifications
relative to the sanitary conditions of the dwellings.
4
Small proprietors on small holdings, with opportunities to pay for them
over a period of years, are given long-term credits for construction of houses,
according to a plan prepared by agrarian officials. Such houses consist of a
separate dwelling for the family, a stable for the cattle, and a barn for the
storage of crops.
6
Halina KRAHELSKA and Stefan PRUSS: Life of the Unemployed, Questionarti
Data, Institute for Social Problems, Labour Market and Unemployment
Series, No. 2, Warsaw, 1933. Also Anna ODERPELDOWNA: Jeunesse d'un
faubourg: Enquête faite dans le faubourg de Varsovie-Ochota. Institut des
Problèmes sociaux. Le marché du travail et le chômage, No. 11. Warsaw,
1937.

S.L.

6

•— 82 —
E.

HEALTH

In the large cities of Poland death rates from pulmonary tuberculosis
and pneumonia rank high, 13.4 and 10.4 deaths per 10,000 inhabitants
respectively in 1936, being exceeded only by heart disease as a cause
of death *.
A total of 28.2 deaths per 10,000 inhabitants in cities of 100,000population and over were due in 1936 to tuberculosis, pneumonia, or
other diseases of the respiratory system, diphtheria, or dysentery.
Infant mortality, one of the most sensitive indices of adequacy of
nourishment and general living conditions, was high. In 1936 the
rate of infant mortality was 141 deaths during the first year of life per
1,000 live births.
F.

LITERACY

Compulsory primary education was established in Poland immediately
upon the creation of the Republic after the war. The illiteracy figures
show a sharp drop since 1921. In 1921, 33.1 per cent, of the population aged 10 years and over were unable to read and write, whereas
in 1931 this figure had decreased to 23.1. The percentage was lower
among the males, 17.8 as compared with 27.9 for the females. There
were also sharp differences between literacy attainment in urban and
in rural areas and in different parts of the country. Thus, only 12.2 per
cent, of the urban population were illiterate, as compared with 27.6 per
cent, for rural inhabitants. Certain differences are further reflected in
the proportion of children enrolled in primary schools. The average
enrolment in 1934-1935 per 100 children aged 7 to 13 inclusive ranged
from 74.7 in eastern Poland to 99.3 in western districts; the central
and southern districts were intermediate in this regard, with 90.8 and
92.1 per cent, respectively of children of primary school age enrolled
as pupils.
III. — India 2
The problems of the standard of living in India are not comparable
to those of the Western world. India is both from the point of size
1
The striking coincidence of high death rates from these diseases and:
districts of overcrowded dwellings owned by low-income families has been
established in studies in such cities as Paris, Chicago, and New York, and
probably exists in Polish cities also.
2
Politically, India is composed of British India, having 11 separate provinces and a few minor administrative areas, and of the Indian Stateswhich contain more than one-fourth of the area of the country and
include about one-fifth of the population. Also, although India is one of the
important industrial nations of the world, over 90 per cent, of its people
live in rural areas. There are some 700,000 villages, with a population of lessthan 2,000 each and half of these are villages of less than 500 inhabitants.
Socially, the population is made up of peoples living in every stage of civilisation—from the most primitive village to the modern industrialised city.
There are many separate and distinct religions, castes, languages (there are-

— 83 —
and the heterogeneous make-up of its people more truly a continent
than a country. Included within the enormous extent of its area of
1,575,187 square miles, excluding Burma, are approximately 360 millions
of people, or almost one-fifth of the human race.
As a result of deeply rooted social and mental attitudes the average
Indian peasant is content to exert a minimum of effort and asks for little
in return. In short, his scale of values is such that he has scarcely come
to recognise the idea of raising standards of living as an objective of.
policy, thus making it impossible to judge his level of living by European
or American standards.
The descriptive and statistical information which is available in India
is of a scattered and incomplete character. Although an attempt has
been made here to present as representative a picture as possible, the
limitations of the data must be recognised.

A.

INCOME AND

EXPENDITURE

Wage rates and earnings vary widely between the different provinces
and industrial areas of India. In 1929, wage statistics for several groups
of semi-skilled workers showed that, whereas in Bombay Presidency
only 13 per cent, of the workers earned less than Rs.17% per month %
the corresponding percentages were 31 per cent, in Bengal, 47 per cent.
in Madras and 53 per cent, in the United Provinces, thus showing that
wage levels are considerably higher in the Bombay Presidency than in
other areas in India a. This fact must be kept in mind when considering
the figures which follow, which have been taken from family budget
studies carried out in the Bombay Presidency.
The most recent of these studies 3 show that the average monthly
income per family surveyed in Bombay City was slightly above Rs.50
in 1932-1933, in Ahmedabad about Rs.46 in 1933-1935 and Rs.40 in 1925
in Sholapur City. In these cities the average family size is about 4 persons, of which 1 y2 to 2 are wage earners. As to the distribution of the
families by income groups, the data show that the monthly income of
over 60 per cent, of the families in Bombay and Ahmedabad ranged from
Rs.30 to Rs.60, and in Sholapur from Rs.20 to Rs.50. The average
monthly income per familv in most of the other provinces varies between
Rs.25 and Rs.35.
The following table shows by income groups the average monthly
income and expenditure of families in Bombay City and the percentage
of families in each group:

nine major recognised languages but, in the Census of 1931, 225 languages
were reported, without accounting for hundreds of dialects) and social customs
which at present clash on al] sides with the industrialisation of the country
and the introduction of Western methods.
I
A rupee equals about 1.60 Swiss francs at present exchange rates.
II
Report of the Boyal Commission on Labour in India. Govt. Pub., 1931,
p. 204.
3
Report on an Enquiry into Working-Class Family Budgets in Bombay City.
Bombay Labour Office, 1935.
Ahmedabad, 1937.
Cotton Mill Workers in Sholapur, 1928.

— 84 —
TABLE IX.

INCOME AND E X P E N D I T U R E BY INCOME

G R O U P S , BOMBAY,

1932-1933
Monthly income groups

Percentage
of
families

Average
monthly
income

%

Rs. a. p.

Rs.40
Rs.50
Rs.60
Rs.70
Rs.80
Rs.90
. . . .

12.25
20.22
26.21
16.27
10.01
5.24
4.08
5.72

All incomes . .

100.00

Below R s . 30
Rs.30 and below
Rs.40 „
„
Rs.50 „
„
Rs.60 „
„
Rs.70 „
„
Rs.80 „
„
Rs.90 and over

25' 3
34 3
43 4
53 3
63 1
73 3
82 15
114 9
50

1

Average
monthly
expenditure
Rs. a. p.

8
3
3
5
8
4
8
4

27 9 7
34 11 7
41 9 1
48 6 9
55 15 5
62 4 3
70 4 11
88 12 0

7

45 15

9

The monthly expenditures of the Bombay families were distributed on
the average as follows: 46.6 per cent, on food excluding liquor, 7.1 per
cent, on fuel and lighting, 7.9 per cent, on clothing and household necessities, 12.8 per cent, on house rent, and 25.6 per cent, on miscellaneous
items. With the exception of house rent and miscellaneous, the percentage expenditure on the different items showed little or no variation
between income groups. However, as income increased, the percentage
expenditure on house rent declined and that for miscellaneous items
steadily increased.
The miscellaneous expenditures include items of both luxury and
non-luxury goods and services not included in the other groups, but
excludes interest on debt and expenditures for marriages, festivals, etc.
The main items are tobacco, pansupari (a leaf and root chewed together),
liquor and toddy, travel to and From native place, and payments to
dependants. Expenditure on education and health is negligible and
only about 25 per cent, of the families spend anything on amusements.
The expenditure on clothing remains more or less constant between
income groups, as the demands of the workers in this respect are small
and quite simple. Clothing requirements and customs vary according
to climate, occupation and caste. No statistical information is available
as to the amount of clothing owned by the worker, but general observations indicate that he often experiences considerable discomfort during
the cold months because of insufficient clothing. As regards bedding,
the majority of the workers possess little or none. However, most
of them possess one or more string cots which are often kept outside
the tenements for want of enough space in their overcrowded rooms.
From the above table it is seen that in the income groups above Rs.40
there is a surplus of income over the current expenditures. It is from
this increment that interest payments on debt are made or extraordinary
expenditure, such as marriages, festivals, etc., are met. In India the
majority of the workers—both industrial and agricultural—are in debt
for most of their lives. The Bombay budget study showed that 75 per
cent, of the workers were in debt and that the average amount of indebtedness was about 3y2 times their monthly income. Marriages and
festivals account for the greater proportion of debt, as they are, by
tradition and without reference to what the family can afford, usually

— 85 —
very elaborate and expensive. In Bombay, unemployment and sickness
are reported t o be t h e next most important causes of debt. The burden
of interest payments throughout India is heavy and interest rates vary
from 18.75 per cent, to 150 per cent., t h e most common rate reported
being 75 per cent, per year.

B.

DIET ADEQUACY

AND M A L N U T R I T I O N

There is evidence t h a t , under present agricultural conditions and
methods of cultivation, India is barely able t o supply enough food t o
ensure an adequate diet for the total population. During 1933-1934, t h e
total acreage under food crops amounted t o only 0.72 acres per head
of population in British India and, according to agricultural experts,
it is impossible to provide a satisfactory diet for t h e present population
from an acreage of this dimension. In Madras Presidency, an attempt was
made in 1933-1934 t o calculate t h e food supply available and to assess
its value in calories, etc., per consumption u n i t 1 . This study suggested
t h a t t h e total calculated food supply was inadequate, both in quantity
and quality, to cover the food requirements of t h e population of t h a t
Presidency. The diet of t h e Madras population is largely composed
of cereals and shows a great lack of protective foods. The supply of
protein is low and t h a t of animal protein almost negligible. Milk
production figures confirm t h e observation t h a t milk is consumed in
very small quantities or not at all by t h e mass of t h e population.
In 1936, diet surveys 2 were carried out in several South Indian
villages which confirm t h e conclusions of t h e above-mentioned study.
These diet surveys covered both agricultural and non-agricultural
families having varying incomes from Bs.50 t o over Rs.500 per year.
Judged by t h e standards of minimum requirements 3 , it appears t h a t
all t h e families, except for a few in t h e higher income groups (Rs.350500 per year), had an insufficient diet. The investigators stated t h a t
it was probable t h a t from one-third to one-half of all t h e families did not
consume enough food during the period of the investigation. The diets
were expressed in t h e various dietary constituents which, when compared
to either t h e recommended diet or to t h e minimum requirements, show
t h a t practically all the diets were inadequate or ill-balanced. The
1

AYKROYD, W. R. : "Nutrition". Intergovernmental Conference of FarEastern Countries on Rural Hygiene.
L. of N. Health Org., C.H. 1253(c),
No. 2, 1937, p. 6.
2
AYKROYD and KRISHNAN: "Diet Surveys in South Indian Villages",
Ind. Jour. Med. Res., 24, 3 Jan. 1937. Also AYKROYD and KRISHNAN: " A

Diet Survey Repeated at Another Season ", Ind. Jour. Med. Res., 25, 1 July
1937.
3
The results of these surveys have been analysed for comparative purposes,
on the basis of the minimum diet requirements as laid down by Dr. Aykroyd
and his colleagues in the Nutrition Research Laboratories in Coonoor, South
India. Very briefly, these recommendations are that, for an adult male in
South India, doing moderate work, 2,600 calories, 65 grammes of-protein,
60 grammes of fat, 0.60 grammes of calcium and 20 mgs. of iron are desirable.
A recommended well-balanced diet is also given which, although falling far
below dietary standards set for Western countries, is considered still too high
for practical nutrition work, as Dr. Aykroyd states that in many parts of
India it would be hard to find a worker whose diet reaches it.
See: " The Nutritive Value of Indian Foods and the Planning of Satisfactory
Diets ", Health Bulletin, No. 23, Manager of Publications, Delhi, 1937.

— 86 —
preponderance of cereals and deficiencies in the protective foods were
apparent. The protein intake was low and was almost wholly of vegetable
origin. In general the diets showed serious deficiencies in vitamin
content.
No expenditure figures were given for these diets, but food expenditure
figures are available for industrial workers in Bombay City 1. Thus,
inasmuch as nutrition studies in South India serve to indicate that in
general the poor in the villages are often better fed and suffer less severely
from malnutrition than do the urban workers, the food expenditure
figures of the latter may serve to indicate the relation between food
expenditure, diet and income levels in general. Food expenditures per
month fall below the cost of the recommended well-balanced diet in
Bombay with the exception of the higher income groups.
Diet surveys taken in Northern India indicate a relatively greater food
consumption2 and, in general, more adequate diets. In a survey taken
in Calcutta , the average calorie intake per adult male ranged from 2,400
to 3,800 per day. This compares favourably with the minimum requirements of 2,800 calories per adult male estimated to cover the daily
energy needs of a North Indian worker. However, the diets investigated
showed marked deficiencies in the same direction as the South Indian
diets, namely in animal protein, animal fat, dairy products, calcium, and
an excess of cereals. This study did not include the poorest classes, so it is
quite likely that the calorie intake is much lower among a large proportion of the population as well. Similar deficiencies in animal fats and
protein and protective foods were found in the diets of jute mill workers
in Bengal in a survey carried out in 1930 *.
In India it is generally accepted that diet deficiency far outweighs
other factors as a cause of malnutrition, so that the term " malnutrition " is used to cover defective development and pathological states
resulting primarily from diet deficiency. Studies of the nutritional state
of the poorer classes of India, especially of children, show the link
between diet and health. Clinical examinations made, for instance,
among the Madras Corporation Schools showed that during 1930 to
1934 some 20 per cent, of the boys were suffering from malnutrition.
An investigation was made in 1936 of the state of nutrition of 2,003 day
school children in three South Indian towns and it was found
that over
21 per cent, showed definite clinical signs of malnutrition 4. The fact
that malnutrition is definitely related to economic conditions was clearly
evident in that symptoms were observed in a larger percentage of children
of the poorer classes than of children of ,the more prosperous classes 6.
In the areas of North India where whole wheat is the staple cereal
and relatively large quantities of milk are consumed, the state of nutrition is considerably better and symptoms of deficiency disease are
1

Report on an Enquiry into Working-Class Family Budgets in Bombay City.
Bombay
Labour Office, 1935.
2
WILSON, AHMAD and MULLICK: "A Diet Survey of Some Families and
Institutions
in Calcutta", Ind. Jour. Med. Res., 24, 1 July 1936, p. 161.
3
Dr. A. C. Roy CHOUDHURY: Report on an Enquiry into the Standard of
Living of Jute Mill Workers in Bengal. Commerce Dept., Govt, of Bengal,
1930.
4
AYKROYD, W. R., and ROJAGOPAL, K. : " The State of Nutrition of School
Children
in South India ". Part I. Ind. Jour. Med. Res., 24, 2 Oct. 1936.
6
It is interesting to note also that investigations of height and weight
carried out in Ceylon and certain areas in India tend to indicate that height and
weight are correlated with economic status and income (and hence with diet)
and not with race. See AYKROYD, W. R., "Nutrition", referred to above.

— 87 —
infrequent in school children. However, diseases associated with vita
min D deficiency—rickets and ostoemalacia—are met with 1.
C.

HOUSING

Among the major causes of the urban housing problems in India have
been the lack of control over the selection of sites intended for industrial
development and the consequent additional overcrowding caused by the
increasing immigration of workers seeking accommodation in the centre
of cities already suffering from a shortage of houses. In many of the
large cities, the limitation of space and high land values brought about
by the pressure of an ever-increasing population add further to the
congestion.
In the chief industrial areas of the Bombay Presidency, the dwellings
in which the working classes are housed are of three main types : (1) sheds
built of corrugated iron, empty kerosene tins, wood, etc., which are of a
semi-permanent character and. for the most part, present very inadequate
housing conditions; (2) " Zavli " sheds —huts usually constructed by the
workers themselves of dry leaves from the date or coconut palms; and
(3) " chawls " or buildings let in separate tenements. They are usually
built in single or two-storied blocks, although they frequently run as
high as five or six storeys, and consist largely of single rooms, sometimes,
and increasingly so, of double rooms, but very seldom of more than two
rooms. Inasmuch as the majority live in tenement buildings or chawls,
more detailed consideration will be given to this type of housing.
Housing in India is marked by overcrowding. In the busiest industrial
centres the houses are built close together, eave touching eave, and
frequently back to back in order to make use of all available space.
Space is so valuable that in place of streets and roads, narrow winding
lanes provide the only approach to the houses. In Bombay City, the
figures as reported in the Census of 1931 show that 33 per cent, of the
population lives in rooms occupied by more than 5 persons at a time
and 1 per cent., or 15,000 persons, in rooms occupied by 20 or more persons at a time. About 86 per cent, of Bombay's population is housed
so inadequately that the streets have to be used to supplement the
sleeping accommodations which the houses themselves cannot provide. Only 4 per cent, of the population live in conditions which ensure
reasonable privacy and domestic seclusion 2. In addition to these conditions of overcrowding, goats and other domestic animals are often kept
in a corner of the room or on the verandah 3.
Apart from considerations of space and overcrowding, conditions of
sanitation, ventilation, light and water supply leave much to be desired
in the housing of the working classes.
During an enquiry covering
5,363 families in Bombay City in 1930 4 , it was found that 59.26 per cent.
of all the buildings surveyed provided only one window per tenement,
25.99 per cent, provided two, and 4.5 per cent, of the buildings had no
windows. In regard to latrine accommodation, it was found that one
latrine was provided for the use of 1 to 8 tenements in 89.33 per cent.
1
WILSON, AHMAD and MITRA: "Nutritional Survey of Schoolboys in
Calcutta and the Punjab ", Ind. Jour. Med. Res., 24, 3 Jan. 1937.
2
Census of India, 1931, Vol. IX, p. 89.
3
" Working-Class House Rents in Bombay City ", Bombay Labour Gazette,
Vol. X, Mav 1931, p. 890.
1
Ibid., pp. 891-893.

— 88 —
of the buildings, for 9 to 15 tenements in 8.35 per cent, of the buildings,
for 16 tenements and over in 1.59 per cent., and no latrines at all were
provided in 0.69 per cent, of the buildings. Similarly for water-taps,
the report shows that one water tap was provided for the use of 1 to
8 tenements in 25.71 per cent, of the buildings, for 9 to 15 tenements
in 40.41 per cent, of the buildings, for 16 tenements and over in 32.97 per
cent., and in 0.87 per cent, of the buildings no water-taps are provided.
Many cases were also reported in which water-pipes were out of order
or the water did not reach the upper floors owing to insufficient pressure
in the municipal water-mains. It will be noted that these figures for both
the number of latrines and the number of water-taps are per tenement
and not per family or per person. Thus, when the conditions of overcrowding are considered in conjunction with these figures the availability
of such conveniences for the individual appears to be very small.
In regard to general housing conditions in other industrial sections
of India, available information indicates that conditions vary little
from those in the Bombay Presidency.
The housing of the workers in the urban industrial areas is said to be
no worse than that in the villages. The houses in the villages are dark,
unventilated and insanitary, but the rural worker has certain material
benefits such as light, fresh air, and a certain degree of privacy.
D.

HEALTH

One of the basic indications of health conditions is presented by the
variations of the mortality rates. In 1933 the reported birth rate
was
36 per thousand and the death rate was 22.4 per thousand 1. The
rates for India are among the highest in the world. Over 45 per
cent, of the total deaths were among children below 5 years of age
and, although the general death rate has been decreasing, the infant
death rate has been increasing and in 1933 was 171 per thousand live
births. The fact that the urban rate is much higher—210 per thousand
—is most likely a reflection of the overcrowding and poor housing
conditions.
The chief causes of mortality in India are fevers, respiratory diseases,
and dysentery and diarrhoea combined, the respective rates per thousand
in 1933 being 12.9, 1.6, and 0.9. Owing to poor diagnosis, the " fevers "
group is undoubtedly overstated, but malaria accounts for the majority
of deaths under this head and presents a serious public health problem in
India to-day. Tuberculosis ranks next to malaria in this respect and
shows a steady increase in incidence; as urbanisation progresses there
is a steady diffusion of the disease throughout the country. Other
diseases which showed a high incidence were cholera, plague, smallpox
and influenza. Plague has lost much of its former importance and the
cholera death rate in 1933 was unusually low.
The close connection between health, housing and sanitary conditions is clearly shown by the high rates for respiratory disease and
dysentery and diarrhoea in the urban areas. It is clear that in India
the congestion and overcrowded housing conditions, and the unhygienic
habits and customs of the people all go towards creating conditions
which spread infection.
1
Annual Report of the Public Health Commissioner with the Government of
India for 1933. Manager of Publications, Delhi, 1935.

— 89 —
E.

LITERACY

Over 90 per cent, of all t h e people in British India aged 5 years and over
are illiterate—i.e. unable t o read and write an intelligible letter in any
language 1. Literacy conditions are better in t h e cities, as in Calcutta
only 57 per cent, were illiterate and in Bombay 73 per cent. In 1936,
of t h e total boys of school-going age (6 to 11 years) who should be in
primary schools, 51 per cent, were enrolled, while the percentage of girls
was only 17 per c e n t 2 . Of every 100 boys t h a t were enrolled only
27 reached the end of the primary course and of the girls only 14 per
cent 3 .
IV. — Japan 4
Problems of living conditions in Japan have been t h e subject of considerable interest both within and outside of Japan in recent years. I t
is to be -recognised, as in t h e case of t h e other countries studied, t h a t
these problems must not be extracted from their setting and judged
in relation to those of Western countries. The fact t h a t modern J a p a n
has risen within t h e past 70 years from an ancient civilisation involving
customs and habits of life totally different from those of t h e Western
world creates special characteristics which must be kept in mind when
considering present-day standards of living. The examination of living
conditions in Japan is made here in accordance with Japanese standards
of comparison, on the basis of physiological and social norms set up
in accordance with t h e simple tastes and frugal mode of living of t h e
Japanese.
A.

INCOME

AND E X P E N D I T U R E

PER FAMILY

With respect t o t h e frequency and continuity of its family budget
enquiries, Japan takes a high place not only in t h e East, b u t in t h e world
as a whole. Figures from t h e latest of these enquiries 6 show t h a t
1
Census of India, 1931. Vol. I, Part II, Manager of Publications, Delhi,
1933,
pp. 424-425.
2
PARKINSON, J. E. : Educational Commissioner with the Government of
India: Education in India in 1935-36. Manager of Publications, Delhi, 1938.
3
For a fuller discussion of conditions in India and of the problem of raising
standards in that country, see BUTLER, Harold: Problems of Industry in the
East. (Studies and Reports, Series B, No. 20.) International Labour Office,
Geneva, 1938.
4
Japan proper covers 382,545 square kilometres, which are divided among
several small and mountainous islands. In the past 65 years the population of
these islands has trebled and in 1935 was 69,254,148 or 181 persons per square
kilometre—forming one of the densest population areas in the world. With the
introduction of industrialisation and its very rapid growth in recent years, the
pace of urbanisation has been increased and the rate of growth of cities has
been over twice that of Japan proper. In 1935, there were 34 cities having
a population of 100,000 or over, four of which had over 1,000,000 inhabitants.
Over 48 per cent, of the gainfully occupied are reported to be engaged in
agriculture.

, 5 Cf. BUREAU OF STATISTICS OF THE CABINET: Kakei chosahokoku (Report

of Family Budget Enquiries), Tokyo, 1929, 1933, 1934, 1935, 1936, 1937.
For summaries, see International Labour Review, Vol. XXXII, No. 5, Nov. 1935,
for the enquiries 1926 to 1934, and Vol. XXVII, No. 6, June 1938, for the
enquiries 1934-1935 and 1935-1936.

— 90 — '
in 1935-1936 the average monthly income of about 1,100 urban wage
earners' households was about 87 yen 1 and that about 40 per cent, of
the households investigated had a monthly income of less than 80 yen.
On the average, the earnings of the head of the household represented
about 89 per cent, of the total and about 4 per cent, was derived from
the earnings of women and children in the family and about 7 per cent.
from other sources.
With regard to family expenditure, the enquiry shows that the average
monthly expenditure for all families was 81.14 yen, thus allowing a
surplus of 5.85 yen. This excess of income over expenditure varied
directly with the amount of income; for the lowest family income
group (less than 50 yen) a deficit of 5.34 yen was shown. The relative expenditure for food was on the average 37.34 per cent., for miscellaneous expenditures 32.15 per cent., housing 15.32 per cent., for fuel
and light 4.62 per cent., and for clothing 10.57 per cent. For the family
income group of less than 50 yen, in which a deficit was shown, food
expenditure amounted to 45.68 per cent, of the total expenditure and
miscellaneous expenditures amounted to over 22 per cent. The other
expenditure items showed relatively small variations as between family
income groups.
B.

INCOME AND EXPENDITURE PER CONSUMPTION U N I T

When account was taken of the number and age of persons dependent
on family income, it was possible to compute for the year 1931-1932
the following distribution according to income per consumption unit
for 992 wage earners' families in 10 cities in Japan:
TABLE X 1 . —

DISTRIBUTION OF FAMILIES, CONSUMPTION UNITS AND
PERSONS ACCORDING TO INCOME PER UNIT .

Total family income in
yen per year per unit

U p to
„
,
,
,
,
,
,
,

100
150
200
250
300
350
400
450
500

.
.
.
.
.
.
.
.
.

Total . . .

Cumulative percentage of
Families

Consumption
units

Children
under 14

0
0.8
9.2
19.3
48.7
63.8
78.4
88.2
91.4
100.0

0
1.2
12.2
24.5
56.6
71.0
83.6
91.5
94.0
100.0

0
1.6
14.2
28.0
64.6
79.5
90.1
95.9
97.8
100.0

i SOURCE: Computed from data presented in the family budget study for 1931-1932.
The published table gave a distribution of families according to 7 monthly family income
groups, (current income, i.e. not including changes in assets and liabilities) in intervals of
10 yen per month, namely under 50, 50-59, 60-69, etc., to 100 and over. Within each income
group there was a further distribution of families according to number of members in
household (ranging from 2 to 7 inclusive) and indication as to how many of these were
children below 14 years of age. In making the computation each family within a given
income interval was treated as though it had the average income of that interval, and each
child under 14 was counted as one half-unit and each person of 14 or over as one unit.
Annual income was obtained by multiplying monthly income by 12.
1

1 yen equals about 1.25 Swiss francs at present exchange rates.

— 91 —
The median family income per unit shown in the preceding table,
slightly over 300 yen, is higher t h a n the corresponding figure for all
urban Japanese families, the exact size of which is not known 1.
The recent Japanese budget studies do not present d a t a on expenditure
per consumption unit. In 1926-1927, however, the basic data on expenditures were presented according to family income further subdivided
according t o family size. This classification made possible a calculation
of food expenditure per consumption unit, following methods similar to
those used in t h e computation of d a t a shown in table X. For later years
published d a t a were limited to average food expenditure per family
in each family income class, and the number of families of each size
in each income class was given but not their corresponding food expenditures. A means of approximating these food expenditures was made
possible by a characteristic of the analysis for 1926-1927: this was, t h a t
the relative deviations in 1926-1927 from the average food expenditure
in each family income class as calculated separately for families with
1, 2, 3, 4 and 5 children were found to be remarkably stable for each
family size, irrespective of income level. These relative deviations (as
1
The families selected for inclusion in the budget studies by the nature of
the requirements of the investigations were somewhat above the average in
income: they were families of employed workers with incomes of at least
50 yen per month per family, with at least 2 and not more than 7 members of
the household and with no boarders or lodgers, who were willing to keep
accounts for a twelve-month period. Comparison of the frequency of family
sizes included in the budget study of 1931-1932, with those shown in the census
for 1930 (excluding those of less than 2 or more than 7) showed an underrepresentation in the budget sample of family sizes associated with the lower
incomes per unit.
For Kumamoto, a city of 172,000 inhabitants (closely similar in size to the
seventh, eighth and ninth largest of the 10 cities included in the budget
study), data are available regarding a distribution in the fiscal year 1931-1932
of all households in the city according to their payment of the household rate
tax, a municipal tax which is in effect a family income tax. The same deductions
were made from incomes of families included in the budget study for 1931-1932
for dependants and for size of income as allowed in assessments of the tax in
Kumamoto. When the two family income distributions were thus put on a
comparable basis, it was found that the median family income in Kumamoto
was slightly under 300 yen per family per year, while in the family budget
study for the same year it was slightly under 600 after deductions in accordance with tax regulations (Professor S. SHIOMI: " Survey of the Distribution
of the People's Income in the Light of the Household Rate ", Kyoto University
Economic Review, July 1933, p. 36).
The average annual income per family for the total Japanese population
in 1930 was estimated by Professor Kawada to be 809 yen, as compared with a
figure of approximately 1,000 yen for the families in the budget study in
1931-1932 (Shiro KAWADA: " T h e Income and Living Conditions of the
Agrarian Population in Japan ", Journal of the Osaka University of Commerce,
No. IV, Dec. 1936, p. 3).
Finally, the degree of continuity and stability of the successive budget
samples from 1931-1932 to 1935-1936 warrants the conclusion that the samples
for later years as well as that for 1931-1932 over-represent the higher income
per unit families and under-represent the low ones. Any statement, therefore,
of the percentage of families in the budget samples having incomes inadequate
to meet a certain norm is undoubtedly an under-statement rather than an
over-statement for the urban population as a whole. Regarding the total
population, including agricultural families, the situation cannot be judged
accurately, but it may be said that Professor Kawada (op. cit.) estimated the
total family income of the rural population in 1930 to be 336 yen on the
average, as compared with about 1,000 for the budget sample in 1931-1932.

— 92 —
found in 1926-1927) were then applied to the average food expenditure
in each family income class for 1935-1936, to obtain the estimated food
expenditure for families of each size in each family income class. The
results so obtained were compared with the estimated cost of a specimen
diet as detailed in the ensuing section.
C.

FOOD CONSUMPTION

As in other countries in the East, rice constitutes the most important
item in the Japanese diet. Other cereals included in the diet are barley,
wheat, millet, " kibi " and " hie ", which are sometimes supplemented
by potatoes and corn. Likewise, vegetables and fruits, fish and meat,
besides various sea products, are served with rice but, in general, the
average diet is deficient in vegetables, fruits and animal products—
especially dairy products.
The most obvious method of testing the adequacy of food consumption
consists in evaluating the actual diet in terms of the various nutritional
constituents (e.g. proteins, carbohydrates, vitamins, etc.) of the quantities consumed per " unit " and comparing the result with some agreed
norm. Unfortunately, however, the more recent Japanese family budget
enquiries do not give figures for the quantities of the various foodstuffs
consumed, and this method of physiological evaluation had thus to be
discarded.
Another less satisfactory method consists in evaluating at current
retail prices the money cost of a set of food quantities that (1) contain all
the nutritive elements required in a physiologically well-balanced diet,
and (2) is adjusted to the consumption'habits of the people in question.
By comparison of the actual food expenditures per consumption unit
incurred by the families studied with the money amount thus computed,
an approximate measurement of the adequacy of their food consumption
may be obtained. In spite of the many recognised shortcomings of this
method, this is the only method whose application is possible in the case
of Japan.
As a basis of computations, the dietary standards recommended by
the Institute of Nutrition of Japan have been used. These standards 1
are expressed in calories and proteins by age levels. Very briefly put
they allow for 80 grammes of protein and 2,400 calories per day per
consumption unit. By an analysis of a diet recommended by the same
Institute 2 for nutrition work among the peasant population, it is possible
to determine the respective quantities of fats and carbohydrates (namely,
29 grammes of fat and 447 grammes of carbohydrates per consumption
unit) considered to be sufficient to meet the energy needs of an adult
Japanese engaged in moderate work. The average cost in 1936 in the
cities included in the family budget enquiries of a diet fulfilling these
requirements was computed to be about 21 sen per consumption unit
per day 3.
1
LEAGUE OF NATIONS (HEALTH ORG.): Report of the Intergovernmental
Conference of Far-Eastern Countries on Rural Hygiene. No. A.19.1937.Ill,
Geneva,
p. 85.
2
Sonraku eiyo kaizen jisshi hokoku. (Report of experiences in nutrition
work among the rural population.) Published by the Dept. of Gumma, 1933,
p. 8151.
In order to determine the approximate cost of such a diet, a list of actual
foods yielding the proper proportions of these dietary requirements were

— 93 —
Comparing the estimated food expenditures per unit, explained in
the previous section, with the cost of such a d i e t l , it is possible to estimate
the proportion of families with sufficient and insufficient food expenditure
or to construct an index of the adequacy of food expenditure. Table X I
gives the results of these calculations 2 , for 1,107 wage earners' families
in 10 cities in Japan in 1935-1936.
TABLE XI.

DISTRIBUTION OF FAMILIES, PERSONS AND
ACCORDING TO I N D E X OF FOOD ADEQUACY

Index of food adequacy
(per cent, of estimated actual food expenditure per
unit to weighted average
cost of specimen diet)

80 to
„
„
„
„
„
„
Total

90
100
110
120
130
140
150

.
.
.
.
.
.

. .
.
.
.
.
.
.

.
.
.
.
.
.
.

CHILDREN

Cumulative percentage of

Families

Consumption
units

Children

0.7
9.6
25.0
45.6
65.0
78.2
83.9
100.0

0.9
12.3
30.3
52.3
73.3
84.6
88.8
100.0

1.4
16.2
37.3
61.3
84.1
94.1
96.3
100.0

The results of these various computations show t h a t the average
number of children, as well as of persons, per family diminishes with
increasing income per unit and t h a t the diet improves steadily with
estimated on the basis of specimen diets of the Institute of Nutrition, taking
reasonable account of local tastes and food habits. This list of foods was
priced for each of the cities covered by the budget enquiries according to
prices published for 1936 by the Ministry of Commerce and Industry. (For
three cities for which retail price quotations were not available, prices were
used for three substitute cities which were nearest geographically and closest
in size.)
1
The amounts spent for food in the budget sample were not analysed
separately by cities, hence the question arises as to the best weighted average
minimum cost of the diet to be used for comparison with these figures. The
most reasonable procedure, since the data for the budget sample are an
average for the 10 cities, is to use a weighted average (the total cost of the
diet in each city weighted by the number of families in the budget sample
for that city). It is, of course, possible that the sums spent on food by a
household in cities with lower than average costs may in fact have exceeded
the cost in those cities, although it was below the average cost for the 10 cities ;
and in this event the family would be classified as' not having sufficient food
expenditure; but it is assumed that such cases are balanced by those of
households in the cities with higher than average costs which actually spent
less than the minimum required in those cities but higher than the average
for the 10 cities.
2
The analysis is, of course, subject to the reservations discussed in the
preceding section, that the individual families of each family income class
are assigned to their respective percentage of food adequacy not according
to the actual facts for each family, which are not available, but according
to an estimate (based upon an application of results from a similar budget
study for an earlier period) of the average result for families of the same size
and income class.

— 94 —
increasing income per unit. In the group spending for food less than
the estimated cost of the adequate diet and, therefore, probably unable
to maintain health and efficiency in the long run were 10 per cent, of
the families, 13 per cent, of the persons and, more seriously, 16 per
cent, of the children below 14 years of age. It is not possible to say
how much margin should be allowed for irrationality in spending or
inefficient marketing, to say nothing of cooking efficiency. If, however,
a small margin—say 10 per cent.—be added to the cost of the adequate diet to allow for these factors, the proportions spending less
than this amount come to 25 per cent, of the families, 32 per cent, of
the persons and 37 per cent, of the children.
It must be emphasised that the foregoing calculations are very rough
and are subject to important limitations. Nevertheless, it seems quite
likely that the percentages of families and of children spending for food
less than 10 per cent, above the cost of this diet had a food intake
which, in the long run, would not guarantee the maintenance of either
normal working efficiency or of health.
It must further be pointed out that the budget sample is made up
of families which are relatively better off than the urban population as
a whole. It is particularly the frequencies in the lowest consumption
unit income groups that are under-represented in the budget sample,
whereas the frequencies in the higher income groups are over-represented.
Therefore, if the figures in table XI were based on total urban families,
the figures would be higher in the lower adequacy groups.
D.

HOUSING

In summarising housing conditions in Japan, a housing standard
which allows a minimum of comfort by Japanese standards has been
selected. This standard is based on the plan followed by the Dojunkai
Association in slum-clearance programmes. In 1927, the Dojunkai
Association, founded in Tokyo in 1923 and subsidised by the State
for the reconstruction of slum areas destroyed in the great earthquake of 1923, carried out a slum-clearance programme in Yokohama
and the houses were reconstructed according to the following plan 1 :
Number of persons
per household

Number of
rooms

Number of
mats i

Floor space in
square feet

1
2-3
4-5
6-7
8 and over

1 or 2
2
2
2
3

4 % or 71/2

79.3 or 133.2
133.2
159.8
177.6
239.8

9

ioy 2
13

1 A mat is used in Japan as the unit of measurement of floor space in houses and also
as the basis for determining rents. It is a rectangular shaped rug about 2.98 feet in width
and about 5.97 feet in length covering thus about 17.79 square feet. The floors of Japanese
homes are usually completely covered by these mats.

This housing plan was set up primarily to meet the minimum requirements of the poorest classes of the urban population as the new structures
1
D O J U N K A I : Report on the Clearance
Yokohama, 1934, p. 14.

of Slums

in Minami-Otamachi

in

— 95 —
were reoccupied by the former tenants. Although these plans were
designed to meet the minimum requirements of an urban dweller, it
must be pointed out that all the buildings included, apart from the rooms',
an equipped kitchen, a store-room, a drying-room, running water and
gas. On the other hand, as is the case in most Japanese houses, no
special means of heating were provided.
In converting this standard into terms of the rent required to meet it,
it has been necessary to use the family budget enquiry of 1926-1927, as
it is the only enquiry which includes data relative to rents per mat for
dwellings by income groups. Thus, an average has been taken of the
rents per mat for all the dwellings covered by this enquiry, including
those of salaried employee households. Then by multiplying this average by the number of mats indicated in the plan of Dojunkai, the
following rent standards have been determined:
Yen
Average rent per mat
1.12
Rent standard for a household of:
1 person
5.04 or 8.40
2-3 persons
8.40
4-5 persons
. .
10.08
6-7 persons
11.76
8 persons and over
14.56
In analysing the data relative to rent expenditure of the families by
income groups and by the size of family, it has been possible to determine
the frequency of households according to the housing conditions for four
classifications of households which vary according to the number of
persons per family as shown in the following table :
TABLE XII. — PERCENTAGE DISTRIBUTION OF HOUSEHOLDS (1926-1927)
ACCORDING TO THE DEGREE OF ADEQUACY
Indices of housing
adequacy

Households having the following number
of persons:
2-3

80-90
90-100
100-110
110-120
120-130
130-140
140-150
150-160
160-170
170 and over . . .
Total. . . .

7.5

—

23.8

—

32.4

—
—

4-5
24.4
13.2
16.7
8.4
12.5

—

20.0
7.4
8.9

16.5
3.2
1.6
3.5

100.0

100.0

|

6-7

8 and over

44.9
6.6
19.2
3.1
15.9
1.2
2.0
5.1

45.3
16.6

Total

2.0

14.3

22.4
7.3
19.5
4.3
20.1
0.5
7.8
9.5
3.3
5.3

100.0

100.0

100.0

—

—
—
—

23.8

—
—
—

Thus from the-table it can be seen that for households of 2 to 3 persons,
7.5 per cent, of the households belonging to this group are below the
norm and for households of 4 to 5 persons, 37.6 per cent, are below the
norm; for households of 6 to 7 persons, 51.5 per cent. ; and for, households
having 8 or more persons, 61.9 per cent, are living in conditions below
minimum requirements. Thus, the percentage shows a steady increase
as the number of persons per household increases. Of the total of all

— 96 —
workers' households included in the enquiry, 29.7 per cent, are classed
below the standard of minimum requirements.
However, in studying the table it is necessary to have the following
facts in mind: (1) The percentages relate only to wage earners' households
(i.e., textile, mine and transport workers and day labourers). (2) The
figures do not include households composed of single persons, for reasons
relative to the method of enquiry. (3) As was shown in the discussion of income, there is a tendency towards an over-statement of frequencies of households in the higher income groups and, therefore, a
relative under-statement of the frequencies of those in the lower income
groups. (4) In calculating the rent standard from the rents per mat for
all households included in the enquiry, those of salaried employees have
been included as well. Inasmuch as the rents paid by the latter tend
to be higher than those paid by the wage earners' households, the rent
standard is somewhat higher than it would be if based on the rent paid
by the wage earners' households only.
In view of these facts, caution is necessary in using the table and the
figures are offered only in so far as they do give, in an approximate
manner, some indication of housing conditions in the urban areas of
Japan.
E.

HEALTH

Birth and death rates in Japan, although showing a tendency to
decrease, continue at relatively high levels. The birth rate went down
after the peak year of 1920 until 1935 when it rose to 31.6 per 1,000,
keeping above that of most other countries and contributing considerably
to the problems resulting from a rapidly increasing population. However, the death rate shows a gradual, but steady, decline, and in 1936
was 17.51 per 1,000 compared with 18.98 in 1931 and 19.2 in 1926.
It is interesting to note that the rural death rate is higher than the urban
rate in Japan. In 1934, the urban death rate was 16.00 per 1,000 while
that for the rural districts was 19.07. It is also noted that, although
both rates are declining, the urban rates show a more rapid decline than
the rural. The infant mortality rate in 1936 was 117 per 1,000 as
compared with 131 in 1931 and 137 in 1926 \
One-tenth of the total number of deaths in Japan is due to tuberculosis and it is estimated that over 1,000,000 persons are suffering from
the disease. Its incidence is especially high among the young people
of the country and in 1935 the number of deaths from this cause was
132,151 or about 2.0 per 1,000, showing an increase over 1934 of almost
1,000 deaths. Besides tuberculosis, the other most frequent causes of
death in Japan are cerebral haemorrhage and embolism, diarrhoea and
enteritis, and pneumonia. Each of these three causes accounted for 9
to 10 per cent, of the total deaths in Japan in 1935.
In an attempt to determine the extent of sickness in the country,
the Nippon Ishikai (the Japan Physicians' Society) carried out a survey
of medical conditions in 1936 by collecting reports from its members.
On the basis of the survey it was estimated that over 2,044.000 persons
visited doctors on 10 October 1936, the day of the survey. This
number forms 2.9 per cent, of the total population in Japan proper.
1
BUREAU DE STATISTIQUE DU CABINET: Dai Nippon Teikoku Tokei nenkwan
(Resume statistique de l'Empire du Japon), Tokyo, 1937. See also: The Japan
Year-Book 1937. Published by the Foreign Affairs Association of Japan, 1937,
pp. 829-831.

— 97 —
It is important to note that the percentage of female patients to the total
female population was 2.5 as compared with 3.3 of male patients.
F.

LITERACY

As a result of the enforcement of compulsory education and also of
the eagerness of the people for education, school attendance is almost
100 per cent, in Japanese primary schools. In 1935, there were
11,150,824 children of school age (6 to 14 years) and of these 11,103,920
or 99.58 per cent, were reported to be in attendance.
The problem of illiteracy is practically non-existent in Japan. Complete data on literacy are not given, but figures showing the educational
standing of military conscripts give some indication of the extent of
illiteracy in Japan. According to the figures for 1936, among 630,882
conscripts only 2,044 or 0.32 per cent, did not know how to read and
write 1 . These figures relate only to males, but as a result of the compulsory system it is expected that the same rate would apply in a general
way to females as well.

1

S. L.

Résumé statistique de l'Empire du Japon, op. cit.
7

CHAPTER IV
SUMMARY

It may be useful to review briefly the discussion of the preceding
chapters and to summarise the evidence presented of low levels
of living.
It was pointed out, to begin with, that if standards of living
are to be the subject of economic and social policy, it is necessary
to define them in objective terms and to give them, as much as
possible, a quantitative content. The definition of the standard
of living adopted in this Report, without minimising the importance
of subjective well-being, stresses such objective elements of welfare
as income, consumption, social services and working conditions.
These elements of the standard of living are measurable to a
considerable extent. Thus, the money incomes of persons, families
and groups can be appraised, and the quantity, weight, chemical composition and calorific and nutritive values of foods can be assessed.
It is less easy to establish objective norms for such items as
housing, clothing, health services, education, but in practice
minimum standards with regard to these items are formulated in
different countries, on the basis of national and local conditions,
which acquire validity as they are gradually accepted by public
opinion and are incorporated in social legislation.
It was found, further, from a survey of existing material that
the data needed for a description and evaluation of workers'
living standards are incomplete and that in many countries even
the simplest facts are not available, or are presented in a form
difficult to use. Nevertheless, it was indicated how family budget
enquiries and other statistical data may be and have been used, at
least in the most important industrial countries, for approximate
descriptions of standards of living among the working population of
these countries.
Finally, in Chapter III, a partial survey of actual levels of family
living was presented.for four countries—the United States, Poland,

— 99 —
India and Japan. As already indicated, this survey was not
intended either as a comprehensive analysis of living conditions of
the workers in the countries considered or for purposes of comparison. I t was primarily an illustration of method in using available
materials for the study of the problem. In presenting the data, only
four elements in the workers' standard of living were taken into
account—food, housing, health and literacy. These elements
serve to indicate in an objective manner the content of living,
especially at the lower income levels. Other elements such as
clothing, working conditions, leisure and recreational facilities, and
social services were not considered, despite their basic importance.
Furthermore, in setting up norms even for nutrition and housing
in the selected countries, certain qualifying assumptions and
reservations were necessary owing to the limited character of
available data. Nevertheless, despite the limitations of the data
and the complexities of the problem, certain broad generalisations
emerge from the survey which it may be helpful to restate briefly.
With regard to nutrition, it is seen that varying proportions of
the population of the world have an insufficient diet when measured
not only by the London standards but even by the tentative norms
adopted by the various countries. As income per consumption unit
in the family rises, especially at the lower income levels, food consumption increases and the diet shows a tendency to become more varied
and better balanced. However, diet adequacy is affected also,
and to an increasing extent as family income increases, by food
habits, tastes, and knowledge in the purchase and utilisation of
foods.
The analysis of housing conditions likewise shows that varying
portions of the population are inadequately housed when judged
even by the minimum standards set up in the different countries.
Inadequate housing is in large measure also related to low
earnings and to insufficient family income. But the evidence
presented shows that poor housing conditions and overcrowding
are related also to the rate of increase of the population and its
distribution, and to general habits of social life which are influenced
in large measure by education.
Public health and literacy statistics show clearly the wide variations in birth and death rates, disease incidence, and illiteracy.
The evidence points to the close interrelation between low nutritional and housing conditions and variations in conditions of
health.
In general, the survey made in Chapter III suggests that the

— 100 —
level of family income and the size of the family are the chief
determinants of the workers' standard of living. It is among the
groups with the lowest family incomes that the lowest levels of
living are found. The intensity of poverty at the lower income levels
is greatest since the largest families are found consistently in the lower
per unit income groups. In other words, there is a greater concentration of consumption units than of families at the lower per unit income
levels—the largest number of children being in those groups where
food consumption and living conditions are the most unsatisfactory.
Malnutrition, overcrowded dwellings, disease incidence and
illiteracy are all more prevalent at the lower income levels.
The facts of this Report tend to reinforce and to sharpen convictions already widely held concerning the prevalence of poverty and
low standards of living and the need for considering economic and social policies to improve existing conditions. Obviously, plans of action
can be laid more effectively only on the basis of more complete
knowledge than we now possess of the facts concerning the character and causes of low living standards in each particular country.
This opens up a wide field for enquiries to fill the gaps in our information pointed out in Chapter II and to ascertain the nature of low
standards along the lines indicated in Chapter III. The incontestable facts concerning the inadequacy of nutrition, housing, etc.,
clearly point to the need for studying the possibilities of increasing
output and the productiveness of human labour, especially in
the poorer countries, as well as of investigating the methods of
ensuring a better distribution of the product. However, while
the main reason for low living standards in most parts of the
world is poverty, an improvement in conditions of living
depends also upon a clearer undertanding of the part played
by education and by social customs in determining the
worker's standard of living. Furthermore, as the social services
are in most countries an important factor in determining health,
earning capacity and security, the possibility of extending these
services is a basic question for further study. Also in view of the
fact that in countries with a rapid rate of population growth,
measures for the improvement of health and social conditions
may result in an even more rapid rate of population increase and
therefore in increased population pressure, it is of importance to
pursue the study of problems of the " optimum " population for
each country and of questions connected with the most effective
distribution of the world's population. Last, but not least, there
are the problems involved in protecting existing standards against

— 101 —
the threat of severe economic depressions and against the dislocations due to rapid technical and structural changes in industry
and trade.
In general, the preliminary study made here suggests that the
relations between the various determinants of the workers' standard of living are complex and that the pattern differs considerably
between countries. It is clear that this pattern depends not only on
the stage of development of the national economy, but also on
national habits and traditions. There is thus a large field here for
social-economic research which would throw light on the nature
of the determinants of the worker's standard of living in particular
countries, and on the relation of these determinants to problems of
international co-operation.