BOMBAY'S WEAKER SECTIONS-A SURVEY OF THEIR LEVELS OF LIVING SARTHI...
BOMBAY'S WEAKER SECTIONS-A SURVEY OF THEIR
LEVELS OF LIVING
This paper presents the findings of a large sample survey conducted by the Tata Institute of Social Sciences in
1978 for understanding the nature of poverty in greater Bombay. Aspects covered in the enquiry include
migration, employment, housing and access to amenities, in addition to the enquiries of socio-economic
parameters. The results show that, contrary to the popular belief, majority of the migrants in Bombay have
come from other districts of Maharashtra rather than from other states. Other revelations relate to income
levels from a caste-based distribution of the sample; to factors responsible for entitlement to housing; to intra-
poor differences in access to amenities and finally to the prevalence among the poor of a 'resource-sharing'
phenomenon, whereby the hardships arising out of economic pressures are sought to be countered partially
Dr. Sarthi Acharya is Professor and Head of the Department of Rural Studies, Tata Institute of Social Sciences,
Deonar, Bombay-400 088.
The post World War growth of big cities in the third world has been accompanied by a
far more rapid growth of the population of the poor in these cities. The principal
reason for the population concentration is migration from the villages and towns. The
natural growth of these cities has been in the range of 1 to 2 per cent per annum
while the total growth has been 3 to 5 per cent.
Amongst the identified causes of their migration are, (a) the continued adverse
terms of trade between rural and urban areas in favour of the latter (Upton 1977), (b) a
relative neglect of small and medium towns, which has forced the people to flock
towards larger cities since the latter possess more potential for sustenance; and (c) a
(deliberate) policy of preservation of the outmoded agrarian structures in the rural
areas, which not only thwart (an egalitarian) growth process but also use the various
complex configurations of caste, class, land and political power for exploitation of the
underprivileged (Byres 1974). During the sixties and seventies, the rate of population
influx has been so rapid in absolute terms that the rate of urbanization has been
faster in the third world countries, compared to anywhere else and anytime before in
The people who migrate, are composed of the rural poor in general, the service and
tiny business classes from the small and medium towns, the young unemployed and
those who are evicted from their land. The bulk of these people are poor. They come
to the city more due to 'push' factors and less due to 'pull' factors. Keeping in view
their limited (zero) purchasing power, they come and settle down in slums and open
spaces. Much of their economic activity is a combination of miscellaneous jobs of low
skill and little work rationality, in undefined workplaces—a combination often
referred to as the urban informal sector. The informal sector today comprises a very
I. This paper forms a part of a larger report prepared by the Tata Institute of Social Sciences and financed by
the ICSSR, New Delhi and the BMC, Bombay.
100 Sarthi Acharya
cheap source of labour for all economic activities, and in many senses it has made
itself indispensable to the economies of these countries through a complicated weaving
of the formal and informal sector labour markets (Acharya, 1983, Santos, 1975).
Historically, some of the old time migrants and the fresher ones with more skills have
been able to enter the so-called formal sector. Their earnings have, however, not
changed much vis-a-vis
the cost of living since the labour productivity and wages are
generally low in all sectors. Their places of dwelling are also heterogenous. They
range from concrete structures to mud thatched huts. Some are legal settlements and
the land/tenement transaction/ownership is as per the law of the land, while others
are illegal and forced occupations which are perpetuated through bribes,
compromises and lack of alternatives.
The pavement dwellers are often the poorest though not necessarily so. Some people
may possess a living space due to their longer stay in the city but their earnings may
not be high while others may be earning more but living in more uncertain spaces.
There is, therefore, an existence entitlement' which has its origins in the duration of stay.
The city of Bombay presents a typical example of a third world megapolis, which has
achieved a high industrial growth of over 6 per cent over the last thirty years
(1950-80). This is an island city which is (inadequately) linked with the mainland.
Many industries are located on the island itself though the latest figures show that the
industrial growth has been more prominent in the mainland, while the island city is
becoming denser with commercial establishments and residential complexes over
the last two decades. The southern tip of the city presents a general high density and
a still higher spot density during working hours. The total population of Bombay was
8.24 millions as per the 1981 census, a rise of 38 per cent over the year 1971. With an
area of 603 square kilometres, the density of population is 13,665 persons per square
kilometre. This density is unevenly distributed, with certain areas like Bhuleshwar
touching 0.4 million persons per square kilometre while in some areas it is less than
one thousand. It is interesting to note, that many of the heavily congested city areas
have decongested over the period of 1971-81, while the suburbs have witnessed a
rise in the population density. The total population has, however, steadily increased
over the last four decades. The literacy rate is 68 per cent, with 74 per cent literacy
among the males and 6 1 % among the females. The average size of a household is
5.07 as per the 1981 census. About 39% of the population lives in slums, 21% of whom
are in the main city and 79% in the main and extended suburbs.
Who are the poor? Where do they come from and what do they do? Where exactly do
they live? What is their life style, their quality of life, consumption pattern and access
to basic amenities? The purpose of this paper is to answer some of these questions
through generating statistical profiles of poverty: social, economic, ethnic,
educational and attitudinal.
There are many studies which have been conducted on Bombay for a diagnostic
understanding of the labour market (Papola and Subramanyan, 1974; Deshpande,
1980), on the informal sector (Joshi and Joshi, 1975; Prakash, 1984), and on the
pavement dwellers (Ramachandran, 1976). It is not our intent here to review these
studies. This study was conceived in the late seventies when a permanent settlement'
of the slum problem was being debated along with the problem of re-building
Bombay's Weaker Sections—A Survey of their Levels of Living 101
The data for this study were collected in 1978 from 2,000 households, of which,
1,000 households were drawn from hut/slum/dilapidated settlements on a multiple
stratified sample basis, and 1,000 households were drawn from pavement dwellers,
on an arbitrary basis. Arbitrariness in the latter sample was necessitated since there
was no known scientific method of collecting data systematically on this population.
The sample represented 0.5 per cent of the identified weaker section population of
Bombay at that time.
In the next section, the demographic and socio-economic details of the sample are
presented; in section 3, the income distribution and levels of living are discussed; and
in section 4, the types of housing and other amenities are looked into. The report
ends with a short note on identification of the priority groups from the point of view of
remedial action for the removal of poverty.
II. Demographic, Economic and Social Characteristics
In Table 1, the frequency distribution of the households per ward is presented. This
table shows that the sample households are highly concentrated in wards E, F, G and
H, a characteristic seen in the census data too. These wards are in the business
district. The poor prefer to stay near their workplaces, to save travel expenses. It may
be stated that in Bombay transport is more expensive compared to most other cities
in the world in relation to the earnings. Also, the population to transport facility ratio is
heavily adverse. Between the housed and the houseless households, it is seen that
the housed households may enjoy a higher income status and thereby are able to
travel, and secondly, the housed households are already entrenched in their abodes
which they would not easily give up, in spite of transport costs. Hence, they are more
PERCENTAGE DISTRIBUTION OF THE HOUSEHOLDS BY GEOGRAPHICAL LOCATION
WITH RESPECT TO HOUSED/HOUSELESS
The average size of the households is 4.81, out of which 1.55 are earning members;
and that of the houseless is 3.70, out of which 1.60 are earning members. Both these
figures indicate the larger security and stability of the former. This aspect is discussed
PERCENTAGE DISTRIBUTION OF THE HOUSEHOLDS BY CASTE AND HOUSED/HOUSELESS
102 Sarthi Acharya
Table 2 shows the distribution of the poor by caste. It is seen that the population of
the scheduled caste is concentrated much more on the pavements compared to the
non-scheduled ones. It is further noticed that among the housed, the SCs and STs
constitute a similar proportion compared to the all India average of about 221/2 per
cent. These figures indicate that if houselessness indicates the extent of poverty, the
scheduled (castes and tribes) are poorer than others.
Illiteracy in Bombay is high but not as high as in the rest of India. Among the weaker
sections, however, this percentage is particularly high compared to the general in
Bombay standards. Table 3 shows that among the housed population it exceeds a
little over 20 per cent while among the houseless it is as high as 78 per cent. The
frequency distribution in Table 3 shows that while the modal frequency of the housed
persons is in the class of matriculates, among the pavement dwellers the modal
frequency is in the class of illiterates. It is presumed that this should be reflected in
the occupational status of the workers also.
PERCENTAGE DISTRIBUTION BY LITERACY STATUS AND HOUSED/HOUSELESS
From table 4, it is evident that there are few professionals and technically trained
persons in the housed households and negligible among the houseless. On the other
hand, the unskilled labourers are more than 23 per cent among the housed
households and over 42 per cent among the houseless. Further, it is indicated that
occupations in columns (3) and (4) in Table 4 may engage a large number of unskilled
PERCENTAGE DISTRIBUTION OF WORKING PERSONS ACCORDING TO
PROFESSION AND HOUSED/HOUSELESS
illiterate persons, which may explain the occupations of the illiterates in jobs other
than unskilled labour. A relatively large frequency concentration of the houseless in
column (4) also reflects the rural origins of people. It is of interest to note that
comparatively few persons are employed in actual production jobs. This is a clear
indication of the lack of skill which inhibits them from getting employment on the
production line. An industrial classification of the workers (not shown here)
corroborates the same result. Also implicit is the fact that industries do not attract
labour. The push factor brings people to the city.
Bombay's Weaker Sections—A Survey of their Levels of Living 103
All the above statistics show that the housed are, by and large, better off, they are
better educated, they work in supposedly better payed jobs, compared to the
houseless. Though accurate data on the length of stay of the residents in Bombay
could not be obtained, the sketchy figures show that the housed have resided in
Bombay for over 10 years while the houseless had been in the city for 5 years or less.2
Origins and Subsistence
Most of the persons in today's Bombay are migrants. The migration process had
begun early in this century when the cotton textile mills were established and labour
was brought for working in the mills from Ratnagiri and other places. Since then, both
push and pull factors have contributed to the migration in Bombay, though in the last
few decades, the push factor has been more prominent due to the persistent rural-
urban gap and the increasing population pressure on agricultural land. It is now
accepted that the urban poverty is an extension of rural poverty (see Dandekar and
In this sample, the place of origin of person or a household was identified on the basis
of whether the household has tangible roots at another place, like possession of land
or other property, has a family link, sends money out, or is a first generation migrant.
Limitations of recall have proved a hindrance to going beyond these details. The bulk
of these people are from Maharashtra, followed by Gujarat and Uttar Pradesh, among
the housed households (Table 5). In the case of the houseless households the order
is Maharashtra, Uttar Pradesh and then Gujarat.
PRECENTAGE DISTRIBUTION OF THE HOUSEHOLDS BY NATIVE ORIGIN,
AND BY HOUSED/HOUSELESS
Table 5 also shows that Bombay city is still an ethnic Marathi majority city, at least as
witnessed from the data on the weaker sections, with no immediate possibility of a
major ethnic composition change.3 The table further indicates that, proportionately,
more Maharashtrians are housed and more outsiders are houseless, which shows that
the people of the State are more firmly rooted in the city, possibly through relatives,
friends, linguistic fraternity and initial migration induced by the growth of cotton mills
in the earlier part of the century, which largely attracted people from the Konkan
districts. The findings of this table are further corroborated by the data on linguistic
distribution (not presented here).
In Table 6, the data on the distribution of the households are presented according to
the place of origin. The definitions of villages, towns and cities are the same as those
followed by the census authorities. The table shows that the majority of the city is
composed of one time villagers, which explains the reason for the unskilled
population. There are more villagers who are houseless which indicates that the poor,
2. Migration and Employment are discussed in another sub-report based on these data. It is so far unpublished.
3. This is contrary to the claims of several political parties who feel that the ethnicity is changing fast.
104 Sarthi Acharya
the landless and the illiterates tend to migrate from villages in search of livelihood.
Additionally, about 32 per cent of the houseless come from other towns and cities. It
is known for some time that the small and medium towns in the country are stagnating
and the metropolitan cities are growing at a very rapid rate. It is interesting to note
that less than 20 per cent of the housed, and only about 5 per cent of the houseless,
belong to the category of those who originally belong to Bombay The latter figure
could possibly include the physically or the mentally retarded persons while the
former category perhaps contains 'old timers' who may have voluntarily chosen to live
in slums in view of their occupation in the informal sector, their socio-economic
background, lack of opportunity, family compulsions or some combination of these.
This is why, in spite of belonging to Bombay, they are where they are.
PERCENTAGE DISTRIBUTION OF HOUSEHOLDS BY PLACE OF ORIGIN AND
It is evident from the above that the bulk of the settlers in this sample are migrants. A
logical inquiry would demand background information for the reasons and the style of
their migration and what links they maintain with the place from which they moved. In
Table 7, a frequency distribution of the households, according to the reasons for
moving is presented.
PERCENTAGE DISTRIBUTION OF THE HOUSEHOLDS BY DIFFERENT
REASONS FOR MIGRATING TO BOMBAY AND BY HOUSED/HOUSELESS
This table shows the prominence of the employment motive for migration, both
among the housed and the houseless households, though the houseless moved for
employment in larger numbers compared to the housed. If one terms the
employment motive as the 'push factor' in migration, it is noticed that this, the
compulsion to move to the city for survival, accounts for over 45 per cent of the
migration among the housed and about 60 per cent among the houseless. A change
of location during the period when the respondent was dependent on parents, is a
small number and comparable in both the samples. A migration on marriage is a form
of cultural migration. It is of interest to note that these two non-economic, and in a
way unintended motives of movement, are strikingly similar for both, the housed and
houseless. Against this, the economic compulsions are contrasting. Few move for
educational purposes, a characteristic expected in the weaker sections.
Bombay's Weaker Sections—A Survey of their Levels of Living 105
It has been stated earlier that data on the comparatively long history of the
respondents are not accurately available. A priori,
however, one can hypothesize that
the poorest sections migrate with their families and have little to look back to at home,
i.e. the place from which they came, while the not so poor would not sever their links
with the origins so easily. Some may have families back in the villages/towns and they
commute periodically while others may maintain a two way cash/kind flow. Table 8
shows a definite link the housed households maintained with their origin since the
time they migrated. Most of them left their families and came. The houseless, on the
other hand have, by and large, migrated with their families. These trends indicate two
possibilities which could co-exist. In the past, the relatively better-off people had
come to Bombay seeking work in the organized (cotton textile) sector. Due to
shortage of space and their requirement of a shelter they came alone . They lived in
chawls, houses or huts. Their off-spring possibly dwell in the same places today.
They have links with their places of origin. The data on the number of visits they
make, dependents outside Bombay (not presented here), vindicate this position for
the housed population. As for the houseless, the people who came could have largely
been pushed out from the rural areas. They could have been destitute migrants who
had little choice but to come after severing all links. Further, they may have come at a
time when Bombay was already densely populated.
PERCENTAGE DISTRIBUTION OF HOUSEHOLDS BY ACCOMPANYING M E M B E R S /
ALONE AT THE TIME OF MIGRATION AND HOUSED/HOUSELESS
Part of the hypotheses described above are verified by the recorded statements of
most of the heads of housed households about job appointments. They had already
fixed their jobs before they came to Bombay while most of the heads of the houseless
households sought jobs after they arrived in the city. Further, the heads of housed
households, to a large extent, were helped by friends and relatives through advances
in cash and kind during their process of settling down while this facility was less
available to the houseless (tables not presented here).
III. Income and Expenditure
Analyses of incomes and expenditures in India are widely available. These studies
largely draw upon data from the publications of the National Sample Survey. Very
often, these studies concentrate on India as a whole or on the States. City level
studies are few, and the samples sizes/entries often do not permit micro analysis.
Further, there are limitations when one operates with grouped data. In this section,
the different facets of the income and expenditure pattern are studied from the point
of view of identifying the homogeneity/heterogeneity of the income distribution and
types of commodities. The elasticities of income and expenditure are also estimated.
The mean per capita monthly income of the sample households was 139.45 rupees,
the mean per capita income of the housed households being 167.44 rupees and that
106 Sarthi Acharya
of the houseless households being 111.22 rupees4. The per capita annual income of
India in 1977-78 was Rs. 1250.00 at current prices, which on a monthly basis
amounts to 104.20 rupees. The figures are presented in Table 9. On the face of it, the
average per capita income of Bombay chawls, slums and squatter settlements is
higher than the national income per capita. Two possibilities are indicated here.
Firstly, there could be an overestimate of the income in this sample since these data
were collected on a one time recall, and it, therefore, leaves out possibilities of
unemployment of the casual workers in some periods of the year. Perhaps it also
leaves out possibilities of fluctuating incomes in the case of the self-employed. The
second possibility identified is the rural-urban gap which keeps the overall national
income low. Since, roughly, 45 per cent of the national income originated from rural
areas which constituted about 76 per cent of the population in 1977-78, a crude
estimate shows the average urban income to be 3.7 times the average rural income in
nominal terms and 2.3 times in real terms.5 Even in urban areas, there is a differen-
tiation since Bombay wages are higher than say, the wages in Sholapur. The reality is
a combination of these possibilities.
PER CAPITA INCOMES OF THE SAMPLED HOUSEHOLDS AND THAT OF
ALL INDIA, 1977-78, AT CURRENT PRICES
Per Capita Income
Mean Standard Deviation
The computed frequency distribution of the households according to the per capita
income interval classes (Table 10) shows a modal frequency in the interval class of
Rs. 201-250 for the housed and Rs. 51-100 for the houseless, respectively. It can,
unequivocally, be pronounced that the houseless are considerably poorer compared
to the housed.
PERCENTAGE DISTRIBUTION OF HOUSEHOLDS BY PER CAPITA HOUSEHOLD
INCOME AND HOUSED/HOUSELESS
0 - 50
5 1 - 1 0 0
201 - 2 5 0
4. All the figures are at the current prices of 1978.
5. Nominal and real figures for rural and urban areas are different since the prices in rural areas and urban
areas are different.
Bombay's Weaker Sections—A Survey of their Levels of Living 107
The two distributions also show that at least 25 per cent of the housed households
earn less than Rs. 100 per capita per month while, at least 15 per cent of the
houseless earn more than Rs. 200 per capita per month. Thus, while generally the
houseless are poorer compared to the housed, not all are, if income is the only
measure of poverty/affluence. In a city like Bombay where physical space is limited, a
relationship between housing and income is neither continuous in space nor in time.
Thus, comparatively higher income people, who may have come to the city much
later than some low income early migrants, may not be able to have the opportunity
of housing while the latter, due to their early migration to the city have access to
housing. This is so since, firstly the general price level, of house/land prices and
incomes has not moved in the same proportion over time; the land/house prices
have risen at a much faster rate. Secondly, in several cases, house/land may not be
for sale at all at any price, unlike other commodities. This particular phenomenon may
be referred to as the concept of non-exchanged entitlement', and is developed
further in the next section.
The disaggregated form of these data were used for measuring the inequality of the
income distribution. The values of the Gini coefficient for the housed and the
houseless are, 0.27 and 0.20 respectively. The inequality among the houseless is less
pronounced, which is expected since the floor income cannot be lower than the
survival level. The overall inequality also is less than the All India or State level figure,
since the top income class is also not high by absolute standards.
It has been argued earlier that the geographic distribution of the (poor) population
generally follows the pattern of demand for labour in each area. The data were
tabulated in a two-way classification with geographic locations and different income
classes.6 This distribution showed no specific recognizable pattern, both for the
housed and the houseless. There are all levels of incomes in the different settlement
wards. One can, therefore, conclude that in a sample where the variations are limited,
i.e., when the bulk of the people have low incomes and their occupations are
classified as ones which fetch them limited income, it would not make much sense to
look for explanations across space.
To adjudge the income distribution by caste, the income distribution of the
households is tabulated by caste in Table 11. The table shows that among the housed
population, both the scheduled castes and scheduled tribes are distinctly more poor
compared to the others. The modal frequency for the SCs is in the Rs. 51—100
income bracket while that of the STs is in the Rs. 0—50 bracket. Thus, between the
SCs and STs, the latter are poorer. These are the tribals who have migrated from the
PERCENTAGE DISTRIBUTION OF HOUSEHOLDS BY INCOME, CASTE AND
6. Data not presented here for saving space. In any case, its presentation does not serve any purpose.
108 Sarthi Acharya
adjoining areas. It may be stated here that Bombay is surrounded by a large tribal belt
on its east, north and south, from where the tribals migrate to Bombay since rapid
urbanization is fast encroaching upon their rural occupations. Among the houseless,
the income differences between the scheduled and the non-scheduled categories is
not so distinct. How does one explain the distinctly poorer position of the scheduled
categories among the housed while the houseless show no marked difference? One
explanation is that the houseless are generally poor, with modal frequency in the
class of Rs. 51—100 for all castes. Persons can be poor only to an extent because
below it, survival itself is threatened. Since recent history has not recorded any mass
starvation leading to death/major catastrophe in the city, it is evident that people are
earning at least a floor income. The survival is at a very low level where there is little
or no scope of inequality, since any further inequality will perhaps starve that
population which lags behind. Second, in the last section, it was observed that the
houseless are, by and large illiterate, unskilled labourers. Their earnings could be
similar in an urban setting since urban labour markets do not differentiate much
between castes for wage determination.
A more pertinent classification would be the one which would identify education-
income linkages. In Table 12, the educational status of the head of the household,7 is
cross tabulated with the income levels. The row and column percentages, given
separately, clearly show a strong relationship between education and earnings. If
Rs. 100 per capita is taken as any arbitrary cut off point, then about 40 per cent of the
illiterates earn less than this amount among the housed and over 50 per cent earn
less than this amount among the houseless. On the other hand, less than 10 per cent
of the housed and none of the houseless earn less than Rs. 100 among the category
of the persons holding educational qualifications of SSC plus.
PERCENTAGE DISTRIBUTION OF THE HEADS OF THE HOUSEHOLDS BY PER CAPITA
INCOME AND HOUSED/HOUSELESS (ROW AND COLUMN PERCENTAGES)
Upto Class IV
Upto Class IV
Class V to SSC
7. The educational status of the head of the household and that of all working persons were tabulated. The
results are strikingly similar. Therefore, only the former is presented.
PER CAPITA CONSUMER EXPENDITURE (Rs. 0.0) FOR A PERIOD OF ONE M O N T H BY BROAD GROUPS
OF ITEMS AND PER CAPITA INCOME CLASS
PER CAPITA CONSUMER EXPENDITURE (Rs. 0.0) FOR A PERIOD OF ONE M O N T H BY BROAD GROUPS
OF ITEMS AND PER CAPITA INCOME CLASS
Bombay's Weaker Sections—A Survey of their Levels of Living 111
An analysis of variance exercise was conducted to test the robustness of the
education-earnings nexus. The F value was significant at 99 per cent confidence
level, statistically validating the result.
Levels of living
Levels of living are often discussed from three points of view. Firstly, levels of living
are discussed by computing the poverty status and the extent of deprivation, so that
appropriate policies of employment, wages, prices and food rationing can be
evolved. Secondly, and this is more for rapidly growing economies, the levels of
living are discussed for calculating the income elasticities of demand with a view to
plan production. And lastly, these are discussed for comparative studies. In this
paper, we present all the three points of view. The data on levels of living are
tabulated according to the per capita expenditure interval classes evolved by the
National Sample Survey for the seventies. This was done since comparison with the
NSS data would be possible.
Data on consumer expenditure by 14 income classes are given in Tables 13 and 14
for the housed and houseless, respectively.8 A comparison with the 32nd round NSS
data on consumer expenditure for urban Maharashtra9, for the year 1977-78, shows a
marked similarity of the NSS figures with the figures for the houseless households for
both, the total expenditure in each income class as well as item-wise expenditure
(Sarvekshana, 1986). In the housed sample, there are a number of households in
expenditure classes above the upper end brackets of the NSS urban Maharashtra
tables. This shows that, in spite of the problems of habitat and utilities so often voiced
in the literature about Bombay, the levels of food consumption are higher in Bombay
compared to urban Maharashtra. The NSS has not provided data on the city since
1973-74, denying a direct comparison. Two further propositions are indicated here.
First, it is possible that the expenditure data are not an appropriate indicator of levels
of living and inequality. This is specifically true for the higher expenditure groups. The
aim should be to collect data by incomes in large NSS samples even though
experience shows that expenditure data is more reliable compared to income data in
sample data collection. Second, after a certain level of consumption, a further
propensity to consume daily items diminishes rapidly. A comparison of routine items
thus shows no major differences.
A poverty line has been described by the Government of India as one at which the
level 6f consumption is the critical minimum for the metabolic process to continue
without disturbing the reserves of a human body. This has been worked out at Rs. 89
per capita per month on food at 1977-78 prices (including carbohydrates for calories
and pulses and meat for proteins)10. According to this definition, about 30 per cent of
the housed households and about 55 per cent of the houseless households live below
the poverty line. This, however, is only an indicative exercise since, in view of the
present debate on poverty measurement, a multiplicity of indices have emerged.
There is no concensus on who are the poor (see Sen, 1980; Kakwani, 1981;
Dandekar, 1982). The most persuasive argument has been that of Sen who maintains
8. The difference between Tables 13 and 14 and the NSS tables is that the latter classifies groups by
9. The NSS has not published a city level disaggregation for the 32nd round so far.
10. See Sengupta and Joshi (1978).
112 Sarthi Acharya
that the poor are the ones who lack entitlement to the basic means of living,
irrespective of whether they consume their entitlement. Sen's concept is only
notional and as such cannot be operationalized unless a large element of subjectively
is introduced. Irrespective of these arguments however, two facts stay: that the poor
in Bombay are not so poor as elsewhere, and Bombay does continue to harbour a
large population of people who are poor from any objective criterion.
Table 15 shows the proportion of cereal consumption to total food consumption and
total food consumption to total consumption.
It is seen that cereals constitute a larger proportion in the total food basket of the
houseless compared to that of the housed. This is an expected result since cereals
are the first priority in the choice of foods (Radhakrishna and Murty, 1980). Food, as a
proportion of total expenditure, shows a secular falling trend after the Rupees 34-43
income class for the housed. As against this, the houseless population shows no
secularity in this trend. The proportion is high and fluctuates (randomly) with increase
in the income levels. This indicates that the houseless population has not stabilized its
expenditure pattern, a manifestation of the transitory nature of the group itself.
The expenditure-income relationships, normally referred to as the Engel elasticities,
have also been calculated. These show the proportional rise in the expenditure on
individual items as the overall income rises. The elasticity estimates 'e' show whether
the proportional expenditure on a particular item rises faster than (e > 1), equal to
(e = 1) or slower than (e<1) the proportionate rise in the total income. These
estimates can effectively be used for fixing prices, output targets and distribution of
essential/luxury consumer goods.
The first regression equation fitted is of the per capita income to total per capita
expenditure relationship. If Y is the total income and E is the total expenditure, then
the equation estimated from the pooled data for the housed and the houseless
households sample is,
E = 312.67+ 0.21 Y R2 = 0.37 N = 1955
where the figure in the bracket is the 't' value. It is evident that the coefficient of Y is
significant at 0.05 level of confidence. The value of the elasticity is calculated in
accordance with the following formula,
This value for the pooled sample of the housed and the houseless is, 0.40. In other
words, for every extra per cent of income earned, the amount spent is 0.40. It is
evident that the overall consumption propensity is not very high. The conventional
economics proposition, that the poor have a high spending propensity, is not
confirmed by these data.
The next equation fitted is the Engel curve for cereals. Let CE represent the per
capita expenditure on cereals. The equation is as follows:
CE = 89.97 +0.02 Y; R2 = 0.19 N = 1955
PERCENTAGE OF FOOD AND NON-FOOD TO TOTAL EXPENDITURE
114 Sarthi Acharya
The coefficient of Y is significant at 0.05 level of confidence. The Engel elasticity is
0.11, indicating that, of the additional per cent of income earned, about 0.11 will be
spent on cereals. It is indicated from this equation that the cereal demand is
(surprisingly) not very high; it appears that non-cereal items assume a precedence.
The Engel curve for non-cereal food (NCF) is as follows:
NCF = 206.5 +0.40 Y R2 = 0.18 N = 1955
The coefficient of Y is again significant at 0.05 level of confidence. The Engel
elasticity of non-cereal food is 0.77.
This is a high figure, indicating that as the income
rises, a significant portion is spent on non-cereal food, an observation corroborated
from the equation earlier.
Lastly, the Engel curve for non-food items (NF) is as follows:
NF = 508.33+ 0.61 Y R2 = 0.27 N = 1955
Once again, the co-efficient of Y is significant at 0.05 per cent confidence. The Engel
elasticity is 2.42, which indicates a very high propensity to consume/acquire non-
food commodities. These commodities are clothes, footwear, fuel, lighting and
These statistics show a design of consumption which points to a low additional
propensity to consume cereals, a little higher elasticity for non-cereal food, and a very
high elasticity for non-food consumption. The elasticity estimates suggest that the
most sought after commodity in the city is not food—there is no mass starvation—
instead, other commodities are increasingly sought after. The city's poor are not
among the poorest in the country. For a price stabilisation and commodity
augmentation policy, one could, perhaps, identify individual commodities which could
be controlled for supply and prices. Some empirical exercises at the macro level have
already been attempted for selected commodities under the 'Minimum Needs
Programme' by the Planning Commission, and the rationed commodities sold by the
fair prices shops are an outcome of these exercises. However, a disaggregated
exercise for a larger number of commodities at the city level may provide much more
spatially relevant figures. Unfortunately these data do not permit much dis-
The poor save little, according to common sense economics. Savings increase as the
incomes increase, since the propensity to consume falls after a certain consumption
level. The question often raised is, what is the extent of saving, once the incomes
increase? In other words, what is the marginal propensity to save? From these data,
however, the flow rate of savings cannot be estimated since the questions posed to
the respondents were related to the total stock savings only in cash, bank deposits,
jewellery and insurance. The data are presented in grouped form in Table 16. This
table shows that in over 80 per cent of the housed and over 95 per cent houseless
households there is no saving. The data are not subjected to further statistical
analysis since the distributions are highly skewed.
Bombay's Weaker Sections—A Survey of their Levels of Living 115
PERCENTAGE DISTRIBUTION OF HOUSEHOLDS BY MONTHLY SAVINGS
AND INCOME CLASSES AND HOUSED/HOUSELESS
Income Class Savings (Rs.)
0-50 50-700 100-150 150-200 200-250 250+ Total
(2) (3) (4) (5) <6) (7) (8)
7.7 14.4 15.5 10.2 16.2 16 80.1
0.2 0.1 0.0 0.1 0.7 0.2 1.3
0.6 0.5 0.6 0.8 1.3 1.0 4.8
0.0 0.1 0.0 0.5 1.0 0.6 2.
0.3 0.4 0.1 0.5 0.9 1.9 3.2
0.9 0.1 0.5 0.6 1.0 5.2 8.4
9.7 15.7 16.7 12.7 21.2 24.0 100
22.7 28.3 21.1 10.6 8 6.8 97.6
0.2 0.1 0.0 0.3 0.3 0.0 0.8
0.2 0.3 0.2 0.2 0.0 0.1 1.0
0.1 0.0 0.0 0.0 0.0 0.2 0.3
0.0 0.0 0.0 0.1 0.0 0.1 0.2
23.2 28.7 21.3 11.2 8.3 7.2 100
IV. Dwelling and Space
Type of Housing
The city of Bombay presents a unique contrast of extremely packed dwelling units
along with vast open spaces of thousands of acres of land (Gonsalves, 1982). The
geographical distribution shows a very heavy concentration in the southern peninsula
though the extreme south is not so heavily crowded. Open spaces are normally at
places which are away from the railway tracks or areas which have not been released
by the authorities for any purpose as yet.
In this section, it is intended to look into the spaces people live in, the facilities they
have access to, and the kinds of difficulties they face. After viewing these statistics, it
may be possible to adjudge the magnitude of the effort needed to meet these demands.
To begin with, we look into the kinds of dwelling people possess. In Table 17 the
distribution of the households is shown by the type of dwelling. This table shows that
the housed mostly live in chawls. While the exact conditions of the flats and chawls
could not be ascertained so as to further sub-divide this interval class, a casual look at
the sample flats and chawls revealed run down facades with crumbling walls. In fact,
PERCENTAGE DISTRIBUTION OF THE HOUSEHOLDS BY THE TYPE OF
DWELLING AND HOUSED/HOUSELESS
116 Sarthi Acharya
the flats were the better quality chawls. The huts were earthen/baked roof structures
supported by earthen/wooden walls. The houseless, by the very definition, live on
the pavements where they have constructed huts made out of rags, waste metal,
wood and synthetic plastic waste. The 4.4 per cent hut dwellers on the pavement are
those households which have built temporary huts in slums and open spaces, but
were unlikely to construct permanent structures. The category of others' includes
households which have made temporary/quasi-permanent arrangements with other
households to stay with the latter as household help or in some other informal capacity.
The ownership and tenancy status of the households vis-a-vis
their dwelling is shown
in Table 18. In this table, the tenants have been grouped into three categories,
namely, the general category tenants, who stay in other people's houses on contract
and do not have any immediate threat of losing their hold over the houses (they
PERCENTAGE DISTRIBUTION OF THE HOUSEHOLDS BY OWNERSHIP/
TENANCY STATUS AND HOUSED/HOUSELESS
General category tenants
include a very large percentage who have no formal agreement on
lease category tenants, who at the time of interview had a fixed time tenancy
agreement with the landlords; and the sublet category tenants who had entered into
an informal agreement with the general category tenants. The category of 'others'
includes those households who were unable to offer any clear answer regarding the
status of their dwelling. This table shows that the bulk of the people are general
category tenants among the housed population. About a fifth are owners, and
together with the tenants, they exhaust almost all of the sample. The tenants in the
chawls have been living since a long time, while those in the huts are of more recent
origin. This categorisation is, as expected, not applicable to the pavement dwellers.
In the last section, it was argued that all
the houseless need not be more poor than all
the housed. The length of stay in the city, to an extent, determines the entitlement to
a house. This entitlement cannot be exchanged with money by the houseless
population and this concept was described as 'non exchanged entitlement'. For
further analysing the validity of this argument, we tabulate the relationship between
the kind of dwelling of a household and the per capita income. In Table 19 these data
are tabulated in the same frequency intervals which were followed in section III.
This table shows no specific trend, indicating an indifferent correlation between the
increase in per capita income and the type of dwelling. The only discernible
observation is the distribution of households in the row on 'flats' in the data of the
housed. A rising trend is observed, indicating that more persons from higher income
groups stay in flats. In the table for the houseless, a falling trend is seen for the
pavement dwellers but the same inference cannot be drawn since this distribution is a
mere reflection of the overall income distribution of the houseless, considering the
Bombay's Weaker Sections—A Survey of their Levels of Living 117
PERCENTAGE DISTRIBUTION OF THE HOUSEHOLDS BY THE TYPE OF
DWELLING AND PER CAPITA INCOME CLASS FOR HOUSED/HOUSELESS
0-50 50-700 700-750 750-200 200-250 300+ All
(2) (3) (4) (5) (6) (7) (8)
0.40 0.20 0.10 0.10 0.60 1.00 2.4
1.60 1.20 1.50 2.22 0.02 11.22 26.8
5.91 11.12 12.12 8.62 11.22 9.42 58.8
1.80 3.01 2.91 1.80 1.20 0.60 11.3
0.00 0.10 0.00 0.30 0.00 0.30 0.5
0.00 0.10 0.00 0.00 0.10 0.00 0.20
9.7 15.7 16.7 12.7 22.8 22.2 100.00
0.00. 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.10 1.00 1.54 0.72 0.42 0.40 4.2
18.15 15.54 18.26 9.54 7.18 5.95 84.6
4.72 2.15 1.54 1.03 0.50 1.12 11.00
22.00 28.69 21.34 11.29 8.10 7.47 100.00
fact that more than 30 per cent of the observations are from this segment of the
population. An analysis of variance exercise was carried out to test the extent of
variation in the rows with variation in the columns. The results did not prove to be
significant at 5 per cent of confidence. It is, therefore, suggested that the hypothesis
of non-exchanged entitlement, referred to above, is supported by these data. The
observation that the housed population in this sample has been resident in Bombay
for a longer period compared to the houseless population, further strengthens the
hypothesis. The non-exchanged entitlement hypothesis can, however, be suggested
for a certain income range only. It would not hold for all income ranges. The hypothesis
can certainly not be generalized to include the upper-middle and higher income classes.
Size of the Dwellings
The area covered by each house is yet another indicator of the quality of dwelling and
this would provide information for studying the disparities 'within' the types of houses
rather than differentiate between the types of houses. In Table 20 a frequency
distribution of the households by size of the living area is given for the housed
sample. The last column of this table shows that more than half the households reside
in houses the area of which is less than 250 square feet. The single largest
concentration is in the bracket 100-150 feet. The municipal authorities have chosen
to distibute plots of about 270 square feet to the poorer sections under the sites and
services' schemes. More than 70 per cent of the households in this sample lie below
this limit. The (weighted) average size of the space is about 330 square feet, which is
higher than the average but the average is high because of the large size of the
interval classes towards the lower end of the table. The United Nations Habitat
authorities have prescribed a space size to be of a minimum of 100 square feet per
person. The per person availability for the households in this sample works out to be
at about 68 square feet. On a casual observation, it was noticed that the bulk of the
habitats constituted of one room and in some cases, the unit also contains a kitchen.
118 Sarthi Acharya
Water source and water closet facilities were mostly a common facility for the
To adjudge the relationship between the size of space and the per capita monthly
income, the data were grouped in the same size classes of the lived-in space and
tabulated in a two-way classification with the income classes. These data, also
presented in Table 20, were subjected to an analysis of variance exercise to adjudge
the variation in the area lived-in, with the rise in incomes. The F value is significant
only at 10 per cent confidence level, indicating a weak relationship between the
dwelling area and the incomes. In view of this result, it is once again indicated that the
'non-exchanged entitlement' proposition is a feasible one.
PERCENTAGE DISTRIBUTION OF HOUSEHOLDS BY THE SIZE OF DWELLING
(sq. ft.) AND INCOME CLASS FOR THE HOUSED HOUSEHOLDS
Income class Area of the space
0-50 50-700 100-150 150-200 200-250 250+ All classes
(2) (3) (4) (5) (6) (7) (8)
1.30 1.40 1.00 1.70 9.80 0.40 6.61
2.81 5.51 3.91 3.91 2.91 2.00 21.04
1.91 3.51 4.01 1.80 3.40 2.21 16.83
1.10 1.30 2.30 0.90 3.00 2.11 10.72
0.50 0.80 1.30 1.00 1.40 1.21 6.21
0.40 0.40 1.30 0.90 2.69 1.02 6.71
0.10 0.50 0.70 0.40 0.90 0.90 3.51
0.20 0.50 0.70 0.50 1.00 1.10 4.01
0.20 0.20 0.50 0.30 1.80 1.90 4.91
0.10 0.10 0.10 0.60 2.00 2.71 5.61
0.20 0.60 0.10 0.20 1.60 2.01 4.71
0.30 0.30 0.20 0.00 0.50 1.80 3.11
0.60 0.60 0.60 0.50 0.70 3.01 6.01
9.72 15.72 16.72 12.71 22.80 22.38 100.00
In this section, the extent of availability of amenities is tabulated for the housed and
the houseless sample, with a view to examining the extent of the overhead shortfalls
and requirements for the poor. This would (in a limited fashion), determine the
quantum and range of efforts the municipal authorities are required to put in, in order
to meet these needs. The overheads included here are electricity, water, bathroom
facilities and latrine facilities.
In Table 21, the distribution of households is shown according to availability/non-
availability of basic amenities: water, electricity, bathroom and latrine facilities in the
entitlement structure of the household. For electricity, the count is in accordance with
whether the house has the power connection while, for the others, the count is
according to easy accessibility, like a private/common tap near the house, a private/
common bath or a private/common latrine adjacent to the dwelling place.
For those who possessed electrical connections, over 95 per cent had legal
connections and they paid the bill, according to the meter, to the authorities or the
house owners (in case of those tenants who did not have independent meters). Less
than 2 per cent were illegal connections. These figures are for the housed. The
houseless do not have power connections.
Bombay's Weaker Sections—A Survey of their Levels of Living 119
Yes No Yes No
(2) (3) (4) (5)
87.1 12.9 0.1 99.9
52.4 47.6 33.5 66.5
78.9 21.1 3.0 97.0
85.9 14.1 4.8 95.2
Water is available easily only to about 52.4 per cent in the households from the
housed sample and to about 33.5 per cent in the houseless sample. Of those who
complained of the difficult availability of water, 82.6 per cent from the housed and
27.4 per cent from the houseless reported that acute shortage of water supply was
the sole reason
for their not being able to avail of water. The other complaints related
to the timings when water is available, excessive time wasted in fetching water,
exhorbitant payments to be made to private wells/taps, and a combination of these. A
small percentage also depended on water from drains, hotels and similar sources,
which were unhygienic and uncertain.
Bathrooms, in the context of the poor, refer to a closed space, with or without a water
connection, within or in the vicinity of the house, which is privately or collectively
used by the households for bathing/washing purposes. Of the housed population,
about 79 per cent expressed no complaints about the bathrooms, as such, except for
the shortage of water. The remaining 7 odd per cent complained of insufficiency of
space, too many people using the same facility, unclean conditions, and a
combination of these. Some 97 per cent of the houseless had no bathrooom access,
while the rest complained of having to pay excessively for availing of these services.
Again, like in the case of the bathrooms, the question posed by the respondents was
on accessibility, sufficiency and cleanliness of the latrines. These latrines are both
private as well as common facilities. Of the 86 odd per cent households among the
housed, who reported accessibility to them, 54 per cent reported no difficulty, 15 per
cent found too many people using the same facility, 9 per cent found them to be very
dirty, and the rest, 8 per cent, complained of a combination of problems stated above.
Among the houseless, over 95 per cent of the households reported no access to
latrines. About 3 per cent complained of excessive payment for using these facilities,
and the rest gave answers which indicated multiple difficulties.
In a nutshell, the single striking fact that one can infer is that the total sampled
households complained of acute water shortage. The sample of the housed
experienced degrees of insufficiency in the basic amenities. The houseless suffered
from the non-availability of amenities.
With a view to assessing people's own initiatives in solving the problems related to
different amenities, a direct question was asked about what they did, if at all, to
enhance their reach to the amenities. In Table 22, the responses are presented.
120 Sarthi Acharya
PERCENTAGE DISTRIBUTION OF THE HOUSEHOLDS ACCORDING TO THE STEPS TAKEN
FOR FACILITATING THEIR ACCESS TO AMENITIES AND HOUSED/HOUSELESS
Approached Municipal authorities
Requested the house owner
Approached a leader
Approached Municipality as well as housed owner
it is evident that the bulk of the people did nothing. The houseless households could
not have done anything since they are "unauthorised" occupants of land; they as such
cannot claim rights over common/public utilities. But the data show the placidity of
the housed people also. A direct probing question was asked about possible
corruption and bribing involved in the municipal offices for extending the amenities.
Less than 5 per cent of the respondents reported to have given bribes or complained
of corruption. This further indicates the placidity. A possible explanation is the
informal support system which the people have evolved within the community. This is
not directly captured in this study since the questionnaire was not designed for it.
Perhaps, another study can be done to understand these informal support systems.
In addition to the above, a subjective question was asked on the immediate
requirements of amenities by the households. Many households were not very
specific on their needs and listed everything as their requirement. A distribution of
their articulated needs is given in Table 23.
PERCENTAGE DISTRIBUTION OF HOUSEHOLDS BY AMENITIES REQUIRED
Water + Bathroom
Water + Latrine
Water + Bathroom + Latrine
Bathroom + Latrine
Everything (including amenities not stated above)
The table indicates two consistent problems faced by the households, the problem of
water and the problem of general dissatisfaction with the available facilities.
Bombay's Weaker Sections—A Survey of their Levels of Living 121
This table shows no clear distinction between the housed and the houseless
households in the context of demands. A further observation in this table is the dis-
similarity within the data on actual availability of amenities. For instance, about 87 per
cent of the households in the housed sample have electricity, while only about 3.7 per
cent want electricity. This leaves about 9 plus per cent in the housed sample who do
not have electricity and do not desire it. Such examples are more striking in the
houseless sample. One can find the answer in the large residual category, i.e., those
who want everything. When the extent of deprivation is all—encompassing,
identification of a single variable like power or water, perhaps, appears meaningless
to the respondents.
The data on amenities were also disaggregated by income classes to seek a possible
correlation between access to amenities and income levels. The data show a very
similar pattern to that of housing versus income class, which was discussed earlier.
This indicates that most of the amenities are linked with the type of house and space.
For the houseless too, the pattern of accessibility to amenities is very similar to that of
the housing access to amenities except for water. Water accessibility is the same for
everyone because public taps, drains, railway stations, hotels, provide it uniformly.
In any case, no significant correlation between income and water availability has been
found. These data are, therefore, not presented.
V. An Approach to Identify Priority Groups
Poverty eradication programmes often begin with identification of the impaired
groups by the type of impairment. For instance, while planning for minimum needs',
the general approach is to identify people who have no access to such facilities as
health centres, primary schools, roads and drinking water, which may have no linkage
to per capita calorific consumption per se.
A single parameter to identify the poor as
such, may not be very helpful. In this section, the priority groups are identified after
taking into consideration a set of three factors, namely, the per capita income,
houselessness and illiteracy. These are only illustrative and not an exhaustive list of
Consider a set P to be all those households whose per capita income is below a
stipulated amount, HL to be the houseless households, and I to be the illiterate
households.11 Then P, HL and I will constitute different priority groups. These
categories are not mutually exclusive. They could be overlapping. The general
category of priority groups (PG) constitute the union of these 3 priority groups. These
could be depicted as follows:
PG = P U HL U I
In a diagrammatic representation, these are shown in illustration 1. The different
categories could be rated by priorities also. For instance, the subset of households,
which is impaired by a single handicap, would receive a lower priority compared to an
overlapping category which is impaired by multiple handicaps. A Venn representation
of the different subsets is represented in Table 24. Set U is the universal set of all
11. The illiterate households are identified in accordance with the illiteracy status of the heads of the
122 Sarthi Acharya
VENN REPRESENTATION OF DIFFERENT SUBSETS OF PRIORITY GROUPS
It is evident from this table that subset 5 is constituted of those households which
have a low per capita income, who are illiterate and are houseless. Thus, they should
receive the top priority in any effort towards poverty alleviation. The next priority
category is constituted of subsets 2, 4 and 6 since they constitute those households
which are devoid of at least two needs, namely, either income and house, or income
and education or house and education. The subsets, 1, 3 and 7 are constituted of
those households which are devoid of only one need and, therefore, rated as the third
priority. The rest of the households in the set U are non-priority households, i.e.,
subset 8. It is also clear that the remedies for each subset would be different. For
example, in subset 5, the effort would be to create facilities so as to augment income,
promote literacy programmes as well as provide shelter. For the other subsets, the
effort would be a (smaller) combination of these three efforts.
The empirical computation of the subsets is presented in the last column of Table 24,
as determined from the pooled data of the samples of the housed and houseless. If
one can project this sample to represent the population, then the number of
households in each priority subset, can be determined for the city as such.
This table shows that 16.90 per cent of the sample households are very poor (subset
5), while about 33.5 per cent households are in the non-priority category. Among the
housed households, the low income illiterates are very few. Similarly among the
housed, non-low income illiterates are few. Priority II covers a total of 26.49 per cent
of the households and priority III covers a total of 23.10 per cent households.
A more detailed exercise of this type, which would cover the type of house structure,
amenities, migration status, extent of education and such variables may exactly
identify the priority subsets by the type of the help needed. This is not being
attempted here, since an exercise of that kind for a (not-necessarily representative)
sample would not make much sense.
This paper presents the survey findings of a large sample survey conducted by the
Tata Institute of Social Sciences in 1978. The conclusions can be summed up in the
Bombay's Weaker Sections—A Survey of their Levels of Living 123
REPRESENTATION OF PRIORITY GROUPS IN A VENN DIAGRAM
124 Sarthi Acharya
1. The bulk of the sampled households are migrants who had/have come to the city in
search of a livelihood. The length of their stay determines their level of living. The
majority of these people are from the state of Maharashtra itself. Between the
housed and the houseless population, the former are better endowed.
2. The income distribution shows that the poor in Bombay are less poor compared to
the general urban levels of living in Maharashtra as well as in India. Engel
elasticities show that cereals are not the most desired item of expenditure as the
incomes rise; it is non-food items. Among the poor, the scheduled categories and
the illiterates are poorer. The inequality among the poor as such is not very high,
possibly because the floor incomes cannot be lower than subsistence earnings. The
savings are near zero. Housing and incomes are not necessarily linearly associated,
indicating that factors determining entitlement to a house are different from those
3. The sizes of the houses are smaller than the area prescribed by the municipal
authorities in their rehabilitation programmes. In most dwellings, water is in acute
shortage. Most other amenities are also in short supply. Most of the respondents
seem to exhibit a sense of placid indifference towards 'non-availabilities' for
reasons which need further exploration.
4. Poverty identification by priority shows that about 16 per cent of the households
are very poor.
"Informal Sector in Developing Countries—A Macro Viewpoint",
Journal of Contemporary Asia,
Vol. 13, No. 4.
Byres T. J.
"Land Reform, Industrialization and the Marketed Surplus in India:
An Essay on the Power of Rural Bias", in D. Lehman (ed) Agrarian
Reformism: Studies of Peru, Chile, China and India. London, Faber
Dandekar V. M.
"On measurement of Poverty", Economic and Political Weekly,
Dandekar V. M. and Rath N.
Poverty in India,
Bombay: Skylark Publications.
Deshpande L. K.
"Labour Market Study in Bombay", Mimeographed, Bombay
The City in Siege,
Joshi H. and Joshi V.
Surplus Labour and City: A Study in Bombay,
New Delhi: Oxford
Kakwani N. C.
Income Inequality and Poverty,
Washington: World Bank.
Why the Poor Stay Poor—Urban Bias in World Development,
Papola T. S. and
"A Study of Labour Market in Bombay", Mimeographed, Bombay
Subramanyan K. K.
The Urban Deadend,
Bombay: Manohar Publications.
Bombay's Weaker Sections—A Survey of their Levels of Living 125
Radhakrishna R. and
"Models of complete Expenditure System in India", Paper A-2361,
Laxenberg: 11 ASA.
A Study of Pavement Dwellers in Bombay,
Shared Space: Two Circuits of an Urban Economy and Their Spatial
Journal of the NSS,
New Delhi: Government of India.
Sen A. K.
Poverty and Famines,
New Delhi: Oxford University Press.
Sengupta S. and Joshi R.
"A Note on the Determination of Poverty Line Based on NSS 27th
Round Data", Sarvekshana,
Vol. Ill, No. 1, July.