NO.

El186197
. I

WORKING ;PAPER SERIES

RESIDENTIAL AND ECONOMIC PWCTICES

OF PAVEMEhT DWELLERS I&- OLD DELHI

V & s ~ q n eDupont
&
Dhananjay Tingal

INSTITUTE OF ECONOMIC GROWTH

DELHI UNIVERSITY. XNCLAVE:
, DELHI-110007,INDIA
WORMIbk '$APE
The IEG Working Paper Series consists of research pqpers,. Project
reports, review articles etc. in preliminary draft and are for limited ,
circulation only. Comments and suggestions q$! t%R.:pêrpp,~"ar@.: s3.j , *:[?:>t,:,:
I

welcome. The responsibility ,for the views expressed rests with the ,
authors.

. -. .- . ., . .
 1

RESIDENTIAL AND ECONOMIC PRACTICES

OF PAVEMENT DWELLERS I N OLD DELHI

Véronique DUPONT* and Dhananjay TINGAL**

i
Fonds Documentaire ORSTOM
Cote: ) O s G s Ex: d.
* ORSTOM (FrenchI Institute of Scientific Research for Development through Co-operation, Dept.
I
1 - -
Societies, Urbanisation, Development, 213 Rue La Fayette, 75480 Paris cedex, France) and visiting
scholar at the Institute of Economic Growth (UniversityEnclave, Delhi 110 007, India).
** Delhi School of Economics, Dept. of Geography ( B l h i University, Delhi 110 007, India).
 2

ACKNOWLEDGEMENTS

The research for this paper was undertaken as part of a collaborative project between the
French Institute of Scientific Research for Development through Co-operation
(ORSTOM- Paris) which financed the study, and the Institute of Economic Growth
(Delhi) and the Centre for Human Sciences (Delhi) both of which provided institutional,
logistical and intellectual support. We are most grateful to these institutions for their help
and co-operation. We would especially like to thank the Director of IEG for all the
facilities provided to the research team and thank the members of the Population Research
Centre of the IEG for their continuous assistance and stimulating discussions.

Sincere thanks are due to the team of field investigators and research assistants who
helped in collecting, coding and editing the data. Those who deserve particular mention
are: Mohammed Baber Ali, Sandeep Chauhan, Bhuwan Kumar, Jay Prakash and Ravi
Shekar. We are also thankful to MsMriga Sidhu for checking the typescript.

An earlier version of this paper was presented at the 14th European Conference of
Modern South Asian Studies in Copenhagen (21-24 August, 1996) and at the seminar of
the Institute of Ecoiiomic Growth on the 22nd November, 1996. We are most grateful to
the participants at these two presentations for their many helpful comments which we
then sought to incorporate in this revised version of the paper. We would also like to
thank Ms Usha Ramhathan, Prof. Omkar Goswami, Prof. Jean-Marc Fontaine, Dr
Christophe Z. Guilmoto and Dr Pierre Audinet for their valuable comments.
 3

CONTENTS

1-INTRODUCTION P. 4

2 - SOURCES OF DATA AND RESEARCH METHODOLOGY P. 5

Secondary data available P. 5
Specific survey of pavement dwellers conducted in Old Delhi P- 6

3 - CONTEXT OF THE STUDY:THE WALLED CITY OF OLD DELHI P. 8

4 - SOCIAL MARGINALITY VERSUS INTEGRATION IN RURAL-URBAN

FAMILIALNETWORKS p. 10
Socio-demographicprofile p. 11
Migration history p. 12
Relevance of caste and community p. 18

5 - ECONOMIC MARGINALITY VERSUS INTEGRAL PART OF THE

METROPOLITAN LABOUR FORCE p. 19
Occupational pattem p. 20
Income, debts, savings and remittances p. 23
Assets in the native place and investment projects p. 26

6 - EXCLUSION FROM ACCESS TO A DWELLING VERSUS RESIDENTIAL

STRATEGY p. 27
Residential trajectory in Delhi p. 28
Financial constraints p. 28
Proximity to the workplace p. 29
Residential mobility and duration of stay in Delhi p. 30
Willingness to move and to pay for a dwelling p. 32

7. SUMMARY OF MAIN FINDINGS p. 34

REFERENCES p. 36

APPENDIX p. 38
Presentation of the research programme: 'Patterns of Population Mobility in the Delhi
Metropolitan Area'
1. Context of the study p. 38
2. Research objectives p. 39
3. Research methodology p. 40
 4

1- INTRODUCTION

The sight of people sleeping at night on the pavements in large Indian metropolises like
Bombay, Calcutta, Delhi and Madras, can be first apprehended as a manifestation of
abject poverty, an acute form of deprivarion from shelter and basic services, and the
pavement dwellers as "the unfortunate victims of diverse kinds of physical and social
crisis among our rural and urban societies" (ARORA, CHHIBBER, 1985: 1).At the macro-
level, the increase of shelterless people in urban areas is analysed as "an inevitable
outcome of the urbanisation process" (NIUA, 1986: 90j. a consequence of
industrialisation and economic development inducing migration of rural poor to the cities
with resulting pressure on land and housing (GUPTA, KAUL, PANDEY, 1993: 29).

While it is not our purpose to deny this side of the reality, we would like to propose a
more qualified appraisal of the practice of pavement dwelling. In the same way as rural to
urban migrants should not be considered merely as pawns pushed and pulled by macro-
economic forces but also as actors in a position to shape the urbanisation process,
pavement dwellers in big cities should not be considered only as the victims of striking
I
poverty, but also as dynamic agents capable of implementing their own economic
I
strategies and of finding appropriate responses to a specific urban environment.

In this perspective, we shall investigate the residential and economic practices of the
shelterless persons found at night in the Walled City of Delhi, the historical core of the
capital. The reasons for this focus will become more evident after we have outlined the
distinctive features of this area; this will allow us to better understand the context in
which the residential and economic practices of the pavement dwellers have developed.
Three pairs of interelated questions will orient our investigation:
i) To what extent is this shelterless population socially marginalised and to what extent is
it integrated into familial networks extending to the rural places of ori,gin ?
ii) To what extent are the pavement dwellers economically marginalised and to what
. extent are they an integral part of the metropolitan labour force, and not necessarily the
worse off in t e m of income levels?
iii) To what extent is this shelterless situation a consequence of a process of exclusion
from access to the urban housing system and to what extent is it a residential strategy
serving individual or familial economic betterment?
 5

As it will be shown by this exploratory study, the heterogeneity of the houseless

population and the variety of situations encounteredmakes general and definite statements
invalid, and calls for a discerning assessment, through a micro-level and detailed
analysis. Before presenting our findings, the sources of data and the research
methodology are described below.

2 - SOURCES OF DATA AND RESEARCH METHODOLOGY

Secondary data available

Although the pavement dwellers are generally qualifiedas "the poorest of the urban poor"
(JAGANNATHAN, HALDER, 1988: 1175), barring some rare exceptions1 the issue of
homelessness is generally overlooked in the abundant literarure dealing with the urban
poor or urbanisation problems2, and primary survey data are relatively few3 as compared
to the numerous studies on slums and squatter settlements' dwellers. In particular, there
is a striking lack of specific studies on pavement dwellers in Delhi, the capital city. The
main reference dates back to 1971: this is the special study on houseless population
carried out with the 1971 census operations (GANDOTRA, 1976). The other references
correspond to unpublished studies, poorly circulated: the pioneer report of the B h m t
Sevak Samaj, a non governmental organisation dedicated to the cause of the poor (1964);
a special study conducted by the Slum Wing of the Delhi Development Authority (DDA)
on the children staying in night shelters in winter (ARORA, CHHIBBER, 1985); a sm$e
survey of about 1069 pavement dwellers in Old Delhi, conducted in 1989 by the Slum
Wing of the DDA ... and whose report was unfortunately untraceable even in the
concerned administration ;a thesis of the School of Planning and Architecture based on a
sample survey of 7 1 pavement dwellers and 30 inmates of government night shelters in
different localities of Delhi (KURUVILLA, 1990-91).

A direct consequence of the paucity of studies on the houseless people in Delhi is the lack
of accurate information on this specific segment of the urban population, and to start with
on its total number. The houseless population enumerated in Delhi urban agglomerationat
the time of the 1981 census was 22,516, and about 50,000 as per the 1991 census

See for example: ALAM & ALIKHAN (1987), SINGH & SOUZA (1980), SOUZA (1983).
.?Toquote a few among some recení studies: NAGPAUL (1996), GUPTA, KAUL & PANDEY (199311),
JAIN (1996), KUNDU (1993). NIUA (1986 & 1989), PERNIA (1994), SURI (1994).
One can however mention: DHAR (1985). JAGANNATHAN & HALDER (1988-a, 1988-b & 1989),
IS1 (19771, MUKHERTEE S. (1975) for Calcutta :RAMACHANDRAN (1972), SPARC for Bombay :
ARORA, CHHIBBER (1985) ,Bharat Sewak Samaj (1964), DDA (1989), KURUVILLA (1990-91) for
Delhi. (Thislist is not exhaustive)
 6

(according to a figure quoted in an unpublished report of the Slum and Jhuggi-Jhonzpri

Department of the DDA (1994). although the corresponding census table has not been
published yet). These figures are obviously under-estimates of the actual population
living in the capital with no form of shelter whatsoever (or the so-called pavement
dwellers) due to the conditions of identification and enumeration of this specific segment
of the population during the census operations. According to a more realistic estimate
provided by the Slum andJhuggi-Jhompri Department of the DDA in 1985, the number
of houseless people in Delhi would be 1 % of its total population, which would
correspond to about 100,000 persons in the mid-nineties. The heaviest and most
conspicuous concentrations of pavement dwellers are found in the Old city and its
extensions, where the Municipality has - logically - opened 8 of its 16 night shelters (6 in
the Walled City proper), representing three fourths of the total sleeping capacity of about
4000 provided for the shelterless people in the entire urban agglomeration .

The morphological and economic characteristics of the historical core of the capital city
(which will be described in the next section) may explain the specifc attraction exerted on
a floating population without shelter. This resulting situation further explains óur choice
to focus the investigation on the pavement dwellers in Old Delhi.

Specific survey of pavement dwellers conducted in Old Delhi

The findings of this paper are based on primary data from our own socio-economic
surveys4. The target population consisted of the persons deprived of any form of shelter
of their own and sleeping at night on the pavements, under the verandas and in other
open spaces, or in the night shelters run by the Municipality5. The studied area was
further limited to the Walled City of Delhi. Two types of observation have been
combined: a statistical survey and in-depth interviews.

In a preliminary phase, the main concentrations of pavement dwellers were identified, a

head-count carried out in each location late at night, and the corresponding information
reported on a detailed map6. The 6 night shelters located in the Walled City proper were

This investigation is part of an on-going project on the pattems of population mobility in the Delhi
Metropolitan Area (see DUPONT, 1995).
See description given in the next section.
'Main concentrations' mean also conspicuous ones, hence the pavement dwellers scattered in the
backside streets of the bazaar, or taking refuge in the underground parking lots, and more generally those
sleeping in the darkest comers of the Old City were not covered by our survey. for securityreasons, since
all the filed work had to be conducted at night. Consequently. there is a risk of bias in the sample, and the
pavement dwellers surveyed should be considered as representative of the 'main stream' houseless people
in Old Delhi, while the underground section,the most marginal among them, and those more likely to be
 7

also included in the survey: for each night shelter the average attendancewas estimated on
the basis of the entrance records for the last previous month, and a map showing the
layout of the premises prepared. This set of maps (for the pavement dweLhg areas as
well as the night shelters) was used as a sampling frame to draw an area sample of about
5 per cent individuals7 for the statistical survey. The survey was conducted at night in
January and February 1996, covering a total sample of 248 individuals: 99 sleeping in the
night shelters and 149 in various open spaces*.

The survey schedule aimed at collecting information on the demographic and socio-
economic characteristics of each individual, and on his mobility pattenn including
commuting to place of work, main migratory steps, and temporary moves in and outside
Delhi. A last section dealt with the characteristics of the family members not hiving with
the respondent.

Along with the statistical survey, in-depth interviews were conducted simultaneously
with a sub-sample of 36 individuals randomly selected: 16 among the night shelter
inmates and 20 among the pavement dwellers sleeping in various open spaces. These in-
depth interviews did not aim at providing statistically representative estimates, but rather
to supplement the statistical survey by an anthropological approach. T h ~ hobjective was
to collect detailed migation, occupational and familial histones with special emphasis on
the migration to Delhi and absorption into Delhi's labour market, as well as relations
maintained with the native place; to better understand the present shelterless situation and
living conditions; to inquire about the future plans including willinpess to move to
another dwelling.

This paper however will not contain an exhaustive presentation of the results of these
surveys: rhis would go beyond the scope of this contribution, and data processing for the
statistical survey is also not complete. Rather our objective is rather to provide some
illustrations and initial reflections to highlight the issues raised above, on the basis of an '

analysis of the in-depth interviews, and some preliminary manual tabulations of the
statistical survey.

involved in illegal or criminal activities. might be under-represented. Besides, the survey a u l d not be
conducted in two of the previously identified and enumerated pavement dwelling areas, following
interferences by the police. But it is not possible to appraise the exact nature and extent of the sample
bias.
As i t will be described below, the population of pavement dwellers in Old Delhi consists almost
exclusively of individualswithout their family.
* When referring to the findings of this survey, and unless otherwise stated, the term "pavemat dwellers"
will be used to designate this segment of the population sleeping either in various open spaces or in
night shelters.
 8

3 * CONTEXT OF THE STUDY: THE WALLED CRY OF OLD DELHI

The Walled City of Old Delhi, the historical core built by the Mughals in the 17th century,
exhibits features typical of traditional Indian cities, with a mixed land use pattern
combining a high concentration of residential units with an important aggregation of
commercial and small-scale manufacturing establishments. What is however remarkable
in the case of Old Delhi is the extremely high residential densities (616 persons per
hectare on the average in 1991, with a maximum of 1596 in one of the census divisions)
combined with an equally impressive congestion of economic activities.

This situation is in fact the result of a two-pronged process which has affected the
dpamics and urban morphology of the old city core. On the one hand, there has been a
decline of the resident population, first noticeable during t k 1961-7 1 decade in certain
areas, and which has continued and spread since then. Hence, although the present
residential densities are stilI excessively high in the Walled City, they were significantly
higher in 1961, with an average of about 740 persons per hcctare. But at the same time,
the Walled City has recorded a dramatic increase in the number of its commercial
establishments (shops, workshops, warehouses and wholesale markets) as well as
manufacturing workshops, including noxious industries and hazardous trades. For
example, the number of registered commercial establishments increased by 700 per cent
in two decades, from 22,000 units in 1961 up to 155,000 lanits in 1981 (DDA, 1993).
Moreover, as rightly underlined by Mehra (1991: 50), the official statistics underestimate
the extent of the growth of economic activities, since the latter do not include the informal
sectors of employment.

While the deconcentration of the population from the urban core is in Line with one of the
proposals of the Delhi Master Plan - although the extent of the population decrease
remains far below the objective initially stated9 -,the promeration of commercial and
industrial activities, especially noxious industries and hazardous trades, is on the other
hand in contradiction with the objectives of the Plan. This shows the limitations of the
attempts at urban planning in the Old City.

For example, at the time of the preparation of the Delhi Master Plan (1958-59). the planning division
'A'. which includes the Walled City and its extension, contained a population of 607,000, and Delhi
Master Plan projected that its population in 1981 be reduced to 322,600. T h e population of this division
according to 1991 census is 616,000 indicating that the dedensification proposals of Delhi Master Plan
could not be realised"(JAIN,1996: 85).
 9

The overuse of the physical space and building infrastructure in the Walled City - both in
terms of residential and economic use - has contributed to the degradation of its housing
stock However. the "cycle of deterioration"was at the outset a perverse effect of the rent
control policies leading to low rents and insufficient income generation, hence
discoura=$ng the owners from incurring expenses for the maintenance of their building
(H3MI. 1988: 4). As a matter of fact, practically all areas of the Walled City were
graduall_vdeclared as "slum" under the Slum Areas (Improvement and Clearance) Act of
1956.

The transformation of the urban morphology in the Walled City and its decaying
condirions of housing have been linked with a certain social recomposition of the residing
population. The better off sections of the popdation are impelled to move their residence
from the old city, in search of better housing conditions in less congested afeas, leaving
behind the low-income group people, in particular tenants who would not be able to find
altanative accommodation elsewhere in the urban agglomeration for such low rentalslo.
Besides, the proliferation of commercial and manufacturing activities, as well as related
services, which provides a large number of informal job opportunities, has attracted a
floating population of male migrant workers whose residential integration remains
extremely precariousll. Thus at night many of them are found sleeping on terrace roofs,
at their work place, in the night shelters run by the Municipality for houseless people,
u n h the verandas in the bazaar, on the pavements or other open grounds.

A m g e of services specifically oriented towards the needs of the houseless population

have also developed. The Govemment itself, taking cognisance of the plight of the
houseless, started constructing night shelters in the early sixties, 6 of them - out of 16
functioning today - being located in the Wailed City proper. In the night shelters run by
the Municipality, for a nominal rate of Rs. 3 per night, each inmate is provided with a
blanket and a ground carpet, and has free access to the toilets and bathrooms usually
available in the same building. Some small private entrepreneurs have figured the
shelterless situation of so many people as a good business opportunity: they rent out
sleeping place and bedding facilities to the pavement dwellers. Quilts on hire are available
for an average rate of Rs. 5 per night, and cots with bedding for an average rate of Rs. 15
per night. Most of the entrepreneurs involved in this business also provide sleeping place
to their customers: at night they encroach some sections of the pavements, in particular
those covered by verandas, as well as pedestrian over-bridges and precincts, or other
open grounds, on which they spread plastic ground sheets or place their cots. In the

See also T
lo O , 1975: 16.
l1 SeeaIso MEHRA,1991: 46.
 10

sleeping areas which are entirely in the open, overhead plastic sheets are also arranged in
case of rain to protect the sleepers. The bedding facilities are particularly in demand in
winter, when the temperature at night can go down to 3" C. However the facility of a
relatively protected sleeping place, and of cots for the better-off pavement dwellers, finds
also customers during other seasons. Since it involves illegal encroachment on public
land, the trade of renting quilts and cots is not authorised; such situation inevitably calls
for police interference, including possible eviction of both quilts' owners and pavement
dwellers and, in order to minimise this risk, bribes to be paid by the "illegal"
entrepreneurs to the police. Yet, a substantial proportion of the pavement dwellers
(including in the sample surveyed) do not avail bedding facilities on rent, they have their
own blankets and sleep under verandas or in open spaces whose access is free - apart
from harassment by the police. Another type of service developed for the pavement
dwellers relates to the provision of hot meals, by road-side food stalls directly installed
on the pavement, most often without any structure, just for a few hours every night at
dinner time. Road-side tea staIls are also a common sight in or around the pavement
dwelling areas, sometimes operated by the entrepreneurs renting out quilts.
i

The main features of the Old Delhi 's scene being drawn, our paper will focus now on the
residential and economic practices of its shelterless population, on the basis of our own
field investigations as described in the previous section.
I-

4 -
SOCIAL MARGINALITY VERSUS INTEGRATION IN RURAL-
URBAN FAMILIAL NETWORKS

Analysing the socio-demographic profile and the migration history of the houseless
people will help us to better portray this specific population; beyond this description we
shall attempt to appraise its degree of social integration versus marginality. One set of
interelated questions arises here. Is the migration to Delhi part of familial strategies rooted
in the native place? What is the relevance of basic social institutions like the family and
the caste affiliation to understand the trajectory and present living practices of the
pavement dwellers? Or, to what extent is the present shelterless situation the result of a
crisis, the outcome of breaking off in the traditional systems of solidarity, or/and the
consequence of a process of individualisation, all leading to anomie in the metropolitan
Set-up?
The first findings presented in this paper may not allow us to evaluate accurately these
alternative processes, nor to draw definite conclusions; nonetheless we can at least
.. 11

highlight some related indications in this respect and illusnate the most significant
situations.

Socio-demographic profile

The first salient feature about d e houseless people in Old Delhi is that they are almost
exclusively male, living as single individuals. Access to the night shelters run by the
Municipality is, by rule, restricted to males. And among the main pavement dwelling
areas identified in various open spaces of the Walled City, only a very few women and
families (less than 10) could be seen during the head-count, and their presence was
moreover confined to one or two localities. Hence the sample surveyed comprises only
male individuals. The predominance of single men (unmarried or not) among pavement
dwellers is especially pronounced in the Walled City as compared to some other parts of
Delhi where pavement &-elles are also located and where familial units are more
conspicuous - like in Nizamuddh But this is also a distinctive characteristic of Delhi as
compared to other Indian metropolitan cities. This is clearly shown by the sex ratio of the
houseless population in the first four megapolises as per the 1981 censusl’: 187 females
for 1000 males in Delhi Urban Agglomeration, as against 453:lOOO in Calcutta, 278:lOOO
in Bombay and 955:lOOO in Madras. The average size of the shelterless households
further confirms the specific position of the capital city: 1.9 in Delhi, as against 4.8 in
Calcutta, 2.0 in Bombay and 4.1 in Madras.

The quasi absence of familid units on the pavements of the Walled City is reflected in the
age composition of the shelterlesspopulation. Although the presence of street children
living on their own is one of the most disquieting aspect from a humanistic point of view,
in demographic terms they represent only a very minor group among all the pavement
dwellers (in our sample, less than 5 per cent of the respondents were below 15)13. The
majority of the pavement dwellers are young people belonging to the age group 15-29
years (54 per cent of the respondents).

More significant from the point of view of its social integration, is the marital status of
this population. Thus, 78 per cent of the respondents were never married in the age group
15-29 years, and 43 per cent in the age group 30 years and above. To better appraise the
specificity of the homeless population, these figures can be compared to the

The corresponding data for the 1991census were not published when this paper was written.
Data on socio-demographic attributes of the houseless people of Old Delhi (sex, age, marital status,
literacy, percentage of scheduled castes and tribes, place of birth) are based on the total sample of 248
individuals, sleeping either in various open spaces or in night shelters.
 12

corresponding percentages in the male population of Delhi urban agglomeration at the

1991 census, that is, respectively: 62 per cent a d 3 per cent. The remarkably high
percentage of never married persons at a relatively xïvanced age can be considered as an
indicator of a certain degree of social marginality among a section of the houseless, a
manifestation of a process of individualisation, chosen or endured. The circumstances of
their mipation to Delhi and more generally their lik story will provide some explanation
to this situation (as analysed below).

The percentage of scheduled castes and tribes' pqulation enables us to appraise the
extent to which the houseiess population in Old De& hails from the most underprivileged
sections of the society. In the sample population, 13 per cent of the respondents have
reported belonging to a scheduled caste or tribe. Gir-tn the sampling error, this represents
a proportion very close to that recorded in the total nale population of the Walled City at
the 1991 census, namely 11 per cent. Interestingly, it is also almost similar to the
proportion of scheduled castes and tribes recorced at the 1971 census among the
homeless population in the territory of Delhi Munizipal Corporation, nâmely 12 % for
males. On the other hand, the proportion of scheduled castes and tribes among the
pavement dwellers of Old Delhi is significantly iower than their share in the male
population of the entire Delhi urban agglomeration.i:at is 19 per cent at the 1991 census.
Though no comparison can be done with census dan. ir is noteworthy that xhe majority of
the houseless surveyed in Old Delhi belong in fact to upper castes or communities (56 per
cent of them), the remaining share corresponding EO Other Backward Classes (31 per
cent).

I The proportion of illiterates is another indicator of sn-io-economic backwardness. In this

I respect the houseless population in the Walled City of Old Delhi appears clearly as a
disadvantaged section, counting 38 per cent of illitemes, whereas, according to the 1991
census, the proportion of illiterates in the total male pqulation of the Walled City was 24
per cent, and in the total male population of urban De-% only 18 per cent14.

Migration history

Migration is a common experience shared by almost d the houseless people in Old Delhi
(96 % of the respondents are non-native to Delhi). Nost of the migrants hail from rural
areas (almost three fourths of the sample migrants), with the major single g o u p among
migrants coming from Uttar Pradesh (47% of the migrants), followed by migrants from

l4 At the 1991 census, the numbers of literate/illiterates (art3 hence the corresponding percentages)
pertain to the population.aged 7 and above.
 13

Bihar. The share of the giant neighbouring State of Uttar Pradesh conforms to the general
pattern of migration to the capital city; what seems more remarkable is that the catchment
area of the Old Delhi pavements extends to farther eastern and southern states like West
Bengal, Orissa, Andhra Pradesh, Tamil Nadu, ...etc. These results are consistent with
the findings of the survey of pavement dwellers conducted in 1989 by the DDA in Old
Delhi, according to which 98 % of the respondents were migrans,most of them coming
from Uttar Pradesh and Bihar. For compararive purpose, it can be recalled here that, at
the 1991 census, the migrants accounted for 40 per cent of the total male urban
population of Delhi, with 45 per cent of them coming from Uttar Pradesh.

The analysis of the reasons which motivated the departure from the native place (or
home) and the migation to Delhi (or anrival to the pavement} reveals contrasting
situations. X first smking point is the impact of familial tensions, quarrels and ill
treatment in case of children: children beaten up by a drunken father or ill treated by a step
mother, cases of disputes over the familial property after the death of the father, quarrel
with the spouse, brother or other parent or relative, ... etc., are frequently quoted in the
nGgration’shistories cf the respondentsfi. When the migration to Delhi corresponds to a
departure from home under circumstances of familial crisis, the main concern for the
migrant was to escape from an unbearable situation, and no wonder that the arrival in
Delhi was not prepared, and the choice of the destination sometimes not envisaged
previously or even left to chance while catching the first departing adin. When flight from
home is evtntually followed by the severing of all links with t h e f d y , no wonder also
that, if unmarried at that time, the young migrant will generdly remain so, since he
cannot bentñt any more from the support of his family to arrange his marriage according
to the prevailing social tradition.

In greater conformity to the general pattern of migration expected for a large metropolis,
the large mE.ority of the respondents have chosen to come to Delhi for reasons related to
better and xnpler employment opportunities and economic prospects. This includes the
many migmts from rural areas or small towns, whose incomes in the native place were
not sufficient to sustain their families, as well as young people willing to get income on
their own outside familial agriculture. The choice of Delhi is often influenced by the
presence of relatives, parents or Co-villagers already working in the capital and conveying
information o n job opportunities; but it may be also merely based on the assumption that
Delhi being a big city, “everybodycati accommodate hinzselj”,“eveqbody carifilid a job
ifhard working“. In the former category, mention should be made of migration rooted in

l5 About one ~d of the 36 respondents selected for in-depth interviews have reported such type of
reasons in their migration trajectory. The statistical analysis of the entire sample will provide a better
estimate of the share of this type of reasons in the migration to Delhi.
 14

farinilial or even village tradition: this is the case of some pavement dwellers working as
cycle rickshaw drivers, handcart pullers or construction labourers in Delhi, often on a
seasonal basis during the lean season for agriculture, and who are thus perpetuating a
practice initiated by their father, or by other villagers. They follow a migration channel
already well established, going to the same labour markets, the same rickshaw garages,
and sleeping in the same pavement dwelling areas. Such groups of villagers can be found
in the wholesale market of Khari Baoli or under the verandas of Asaf Ali road. In the
latter place, for example, a group of 25-30 persons from the same village in Uttar
Pradesh was identified, staying together though from different castes, sometimes also
cooking together, moreover they return together to their village for the main festivals, and
every month one person goes back to the village and takes the remittances from all the
villagers working in Delhi to redistribute them to their respective families.

At the opposite end of these well organised migration channels, some pavement dwellers,
especially among children and young people, have been attracted by the capital city as a
place to visit ... and eventually stayed over, because they also found better economic
scope in Delhi .

Due to the different circumstances under which the migration to Delhi takes place, the
previous information about job opportunities, possibilities of accommodation and other
living conditions in the capital varies a lot. Though the sub-sample of houseless persons
I
l selected for in-depth interviews is small (36 respondents out of which 33 are 'decision
~
making' migrantsl6), it seems significant that at least one third of the migrants reported
that they did not have any information on employment opportunities in Delhi prior to their
anival, and two thirds had no specific knowledge about possibilities of accommodation.
Furthermore, while some migrants had specific information pertaining to certain jobs,
especially those coming through familial or village-based migration channels, for others
&e type of previcus information reported may be very vague, and limited to a general
knowledge on Delhi's labour market as a potential source of numerous employment
opportunities, in particular of manual jobs.

Other indicators can help us to better appraise the extent to which the houseless were and
had prepared their settling in Delhi: One third of the migrants interviewed had previously
visited Delhi - but at that time not necessarily with any prospective concern; and less than
half knew already some person(s) settled and/or working in the capital, be they relatives,
friends or Co-villagers. But only very few (4 out of 33) had made previous mangements
to secure a job in Delhi: apart from one case of transfer on government job, one case

l6 "Decision making migrants"means that follow-up dependent migrants are excluded.

 15

through contractor, z d the other two through urban contacts (relatives, Co-villagers) who
gave them assurances of introduction and help to start a work.

A question remains ZT the conditions of migration of the houseless people in Old Delhi
different from that of6Je "usual" migrants to the capital, with respect to the prior degee
of awareness about explo.ment situation in Delhi and previous arrangements for work?
Although it is d i f f i d t IO establish any statistically significant comparison, the study
conducted by B a n e x on "Rural to urban migration arid the urban labour market" in
Delhi (1986) provi% some insightsl7. Focusing on information flow and rural-based
job search, the authar notes: "The most striking feature (...) is that 28 per cent of the
migrants stated that5ey had no information on employment opportunities in Delhi prior
to their arrival (Bmejee. 1966: 79). Yet, the majority of therural migants (53per cent)
had received prevíozs assurances of urban employment, from the employer himself, or
from urban contacs who made suggestions to move (1986 82-83). This survey "also
confirms the findin,rs of earlier studies on the importance of relatives and friends in
transmitting infom-on" (1986: 79).

Thus the 'common'' flow of migrants to Delhi seems to include a rather significant
proportion of indi- who have come to Delhi without previous information on work
opportunities, and-fi3m -&s viewpoint the houseless migrants of Old Delhi would not be
exceptional. Howzss-=r.it seems that among the latter the elements of uncertainty about
absorption into thekjour market (prior to their arrival in Delhi) are much more prevalent
than in the main smam of migrant workers. This is probably related not only to the
familial and socidantest of the migration of certain pavement dwellers (as seen above),
but also to the rang of possible jobs that this largely uneducated and unskilled population
can expect to get, k i t is mainIy work as casual labourers or as own-account workers in
the informal sector ks it wilI be deveIoped in the next section).

The question of he12 received by the migrant at the time of his first arrival in Delhi is
certainly revealing in order to evaluate the degree of integration of the homeless
population in the uljan social fabric. It is however ambiguous since the perception of
help or support is subjective and hence varies from one respondent to the other.
For example, infonarion about possibilities of sleeping in this night shelter or that
pavement dwelling ET^. or about a specific labour market and employment opportunities
in a certain branch of activity may be considered as a form of help by some, while others
in the same contar would narrate their arrival in Delhi as an ordeal where they had to

l7 This study was basedon a sample of 1,615 male and decision making migrants surveyed in 1976 and
who came to Delhi in 3955 or later, out of them 1,407 coming from rural areas.
 16

manage entirely on their own without outside support, crucial though this information
might have been. In the same way, help can be provoked by the first approach and the
contacts established by the migrant in a market area or around the railway station or main
bus terminal. Given all these reservations, about one third of the 33 decision making
migrants interviewed stated that they did not receive any type of help at the time of their
arrival in Delhi. Among those who acknowledged some type of srrpport, help related to
finding a j o b or starting work as self-employedcomes in the frst place, followed by help
related to accommodation or place to sleep. The networks of relatives and Co-villagers is
approached by the migrant in the majority of the concerned cases, showing that the
present houseless condition does not mean necessarily that the pavement dwellers operate
in a familial and social vacuum, However, the economic and housing conditions of the
relatives already settled in Delhi may put limits to the type and duration of the help
extended to the new migrant, especially as far as accommodation is concemed. In a few
examples, after an initial stay upon his arrival in a slum hut or one-room tenement with
his parents or relatives, the migrant is compelled to leave due to lack of space and to stay
in a night shelter or a pavement dwelling area. The persons working in the occupation or
sector of activity in which the migrant has been eventually absorbed also play an
appreciable role in introducing him to a new work, including sometimes training, and
showing him cheap (or free) and convenient places to sleep.

Examining the relations maintained with the native place or homel8 is a-way of
i i.

investigating the extent to which the houseless population in Old Delhi remains integrated
into familial networks most often rooted in rural areas. Most: of the respondents
interviewed still have family members staying in their native place (only 3 out of 36
respondents did not) and the majority of them visit their native place and family more or
less regularly - at least once during the last two years - or intend to do so in case of very
recent migrants (less than one year). On the other hand, visits of parents and relatives to
Delhi to meet the pavement dwellers are infrequent, which is not surprising given the lack
of accommodation of the latter. A notable proportion of the houseless migrants also
maintain contacts with their families through letters. Another revealing indication is the
practice of remittances or support in kind provided to the family: about half of the
respondents still having family members in their native place do provide financial
support, often supplemented by gifts of clothes or household items at the time of visits.
Reciprocal exchanges for the benefit of the migrant in Delhi, through gifts of food or
clothes for example or even financial help, are much more rare, which is to be expected
given the fact that it is precisely the unsatisfactory economic conditions at the native place

l8 The pavement dwellers native to Delhi are also included in the following analysis, since it is also
relevant to appraise the type of relations maintained with their previous home and family.
 17

s
n-nich pushed many prospective migrants to leave. Attachment to the family is further
=vealed by future plans of returning to the native place (in the next years or much later),
5 wish that is shared by the majority of the houseless migrants interviewed. Admittedly,
.
- i to a future return to the ‘village‘and the basic community includes also a
-.a reference

zythical dimension, that may help the pavement dwellers to accept better their present
k i n g conditions in Delhi, and to justify the hardships and degrading aspects of their
shation.

-
¡ne purpose in reporting the above findings based on a limited number of in-depth
ixïrviews is - evidently - not to provide any statistical estimate of the frequency of certain
E-xiibutes or practices among t’ne houseless, but to illustrate the various situations
ezcountered and give an idea of their significance. In particular, as regards the
sig*icance of the familial unit for the houseless individuals living alone in Old Delhi,
m o distinct - and opposite - patterns can be identified, with a whole range of
kmnediuy (and transitory ?) situations. Correspondingto the highest degree of familial
icegation, one can find the seasonal migrants coming every yeaï to work in Delhi for a
€=u.months, usually during the lean season for a&culture, and supporting directly their
fmilies in the native place. Close to this group are the manied migrants whose spouse
a d children (if any) are left with the extended family in the native place and to whom
rexhtances are periodically sent, as well as the unmarried migrants contributing to the
fmilial income, both groups visiting their families regularly. These remitter-migants
essmplify familial solidarity transcending residential unity: this is a common feature of
r2-grarion process, especially of rural-urban migration, in India as well as in other
de\-tlopingcountries. What deserves to be underlined here is that the houseless condition
OZ h e migrant in the city does not prevent him from exercising this solidarity; in fact, as it
k5-X be developed in another section, it is this very condition which allows him to support

SnanciaUy his family.

Ar the lowest rung on the scale of familial attachment, one can find the individuals who
ra= away from home following an acute familial crisis often involving violence (quarrel,
&sute, ...etc.), and who eventually severed all links with their families and native place
(if it is not Delhi). While most of the children belong to this category, as also adults who
feh compelled to take this radical step when they were young, some individuals
exptrienced such crisis at a relatively advanced age. Given the circumstances of their
departure from home, these migants or escapees cannot rely on familial networks for
their integration b a new place. In this group of houseless, a process of individualisation
and anomie, more forced than chosen, may be at work, resulting from the breaking away
ko;n the basic social institution, the family.
 18

I
Such traumatic experiences at the origin of the present shelterless situation should not be
I confused with the few examples of pavement dwellers who decided deliberately to
withdraw from family life and all material attachments to live a life of renunciation - or
close to it.

._ Relevance of caste and community

Turning to the relevance of caste, community affiliation and ascriptive social status in the
living practices of the houseless people, the accounts of some individuals provide a few
insights in this issue. There are, first, the cases of several upper caste pavement dwellers
who dislike sleeping in night shelters due to overcrowding which may compel' them'to
sleep close to people from very low castes or other religions. Hence they prefer to sleep
in open spaces, where this constraint is not felt, sometimes even if they have to pay more
than the entrance fees for the night shelter in order to hire a quilt to protect themselves
from the cold during winter nights (Rs. 5 as against Rs. 3 per night). The high original
social status of some pavement dwellers also shows through the fact that, though
shelterless, they never sleep directly on the ground and always take a cot on rent for the
night. Further, at the sleeping place some pavement dwellers and inmates in night shelters
I would socialise only with people from their religious community. The network of
socialisation at the pavement site or night shelter does not appear, however, to be caste
based. In fact, the work place and the community of workers in the same type of

1 occupation provide the main network of socialisation for the houseless people during
their stay in Delhi. In the wholesale market of Khari Baoli especially, many handcart
pullers can be seen staying in groups at night, sleeping on their cart or under the verandas
of the market, and cooking food on the pavement in small groups for their dinner. For the
youngsters as well as old-aged people, the age group is also a relevant group of reference
I
to make friends and sleep in the same place. As seen above through the example of a
group of seasonal migrants from the same village staying together on the pavement site
and who managed to reconstitute a sort of community life, belonging to the same place of
origin can create links which transcend the caste differences, at least during the temporary
stay in Delhi, led by work concern. Yet, a notable proportion of the houseless people
interviewed stated that they did not socialise with anybody at their sleeping place and the
majority of them usually have their dinner alone.

Other revealing case stones with respect to ascriptive social status concern pavement
dwellers from upper castes who had to leave their native place due to stringent economic
conditions or following a sudden financial crisis. They chose Delhi as a place to migrate
 19

not only forxhe larger range of employment opportunities provided in a big metropolis,
but beforehand for the anonymity offered by such an urban environment, hence allowing
them to take up jobs (in particular manual jobs) that their social status would not permit
them to cany out in their native place. Far from the native village or town and from the
primordial community, in short far from the original social context, the range of
acceptable xcupations becomes wider for the migrant, However, the psychological cost
of such survival strategies may be high. Since nobody in their native place should know
about their morking and living conditions in Delhi, they do not give any contact address,
do not exchange any news through letters or messengers, avoid meeting people from
their village or town of origin and, above all, seem to be overwhelmed with the shame of
their "non respectable" work, compounded by their shelterless situation. This feeling of
shame proqted some of them to severe all contacts with &eir families and native place,
and, in Dellhi, to minimise their relations with other pavement dwellers, hence
aggravatingtheir isolation.
x

Living alone and without shelter does not necessarily imply familial vacuum and social
marginality. For the majority of the houseless srnveyed in Old Delhi, +he family and the
native placemain their context of reference. Furthermore, the shelterless situation of the
migrants in the capital is not a detenent factor of familial solidarity, on the other hand it is
often a precodtion.
There is some evidence of the persistent significance of ascriptive social status and
community feeling in the living practices of the houseless people. However, other
networks of ,wcialisation gather momentum in the urban set up, in particular the group of
co-workers and people from the same village or region of origin.
Yet, there is also a certain section of houseless who seem to undergo a process of
individualisation and anomie: those, especially the children, who fled their homes
following an acute familial crisis and eventually severed all links with their place and
community of origin.

5 - ECONOMIC MARGINALITYVERSUS INTEGRAL PART OF THE

METROPOLITAN LABOUR FORCE

Having exploEd the socio-demographic profile and migration history of the shelterless
population of Old Delhi in order to assess its degree and mode of integation in the urban
as well as m d society, the same type of investigation can be carried out in relation with
the economic profile of this population and the question of its economic marginality
 20

versus integration. One correlative question concerns the degree of poverty of the
houseless, in particular, can the pavement dwellers be considered as "the poorest of the
urban poor"?

Occupational pattern

Table 1. Occupational pattem of the houseless population of Old Delhi - 1996.

include all the occupations carried on in Delhi during the last 12

are excluded the occupations carried on during the reference period
native place or previous place of residence).

Table 2. Percentage distribution of the houseless population of Old Delhi by industrial

category (1996) and comparison with the urban population of Delhi at the 1991 census

1 I 1 I
Houseless Male Male urban
INDUSTRIAL CATEGORY population of population of population of
OldDelhi OldDelhi Delhi

Sample of 243 workers classified by their main occupation (non workers = 5)

(2) & (3):Cetuus of India 1991,Delhi, Primary Census Abstract.
 21

The occupziona! pattern of the houseless population of Old Delhi is analysed here on the
basis of the rotal sample of our statistical survey, taking the last twelve months preceding
the survey 3the period of reference to record all the occupations carried on in Delhi.
Hence are Excluded from this analysis the occupations carried on during the reference
period but outside Delhi, for example in the native place or any other place of previous
residence. This allows us to highlight one revealing feature of the houseless' economic
strategy: the combination of several occupations during their stay in Delhi. Thus, almost
one-third of the respondents have carried on more than one occupation, usually two.
These are okzn ahmated according to the changing opportunities of the labour market, in
particular a-cording to a seasonal pattern; sometimes different activities are also carried on
.
simultanemly throughoat the year. On the other hand, the proportion of houseless not
reporting my income generating activity19 in Delhi remains marginal (2 5% of the
respondentsr, these are essentially very recent in-migrants (less than one month) still
looking for 2job.

Examining now the different types of occupations performed (Table 11, the major groups
reflect directy the economic functions of Old Delhi. There is first the large lot of handcart
pullers and Yilshers transporting goods in or from the wholesale markets of the old city
(20 per cen: of the respondents are engaged in this work for their main occupation).
Loading and unloading activities in the markets, carriage of luggage in the two nearby
main railway stations of &e capitaI provide also appreciable employment opportunitiesfor
the pavemeEr dwellers. Tmsport of passengers by cycle rickshaw in a densely populated
and very buoyant market area, with two adjoining railway stations and a major inter-state
bus termin& is another activity attracting a significant number of houseless workers (20
per cent of fie respondents for their main activity). Altogether, the sector of transport
seems to absrb the majority of the houseless workers based in the Walled City (52 per
. cent of the respondents for their main activity). This appears as a major distinctive
characteristi: of the occupational structure of this population, as compared to the male
population cr'urban Delhi as a whole, but also to the male population residing in the
Walled City the s a t o r of transport and communication employs only 8 to 9 per cent of
the correspodng workers flable 2).

Another speC5city of OldDelhilies in the many labour markets, in the strict neo-classical
sense of the am,which rake place in different places of the Walled City. Some of them

l9 In this study. beg-: are considered as part of the workers, contrary to the convention applied in
official statistics &e tbose issuedfrom censuses and National Sample Surveys.
 2.2

I are specialised in recruiting various types of workers (in particular waiters, cooks and
related service workers) for marriage parties and other functions requiring catering
~

services and the set up of temporary tent structures to host large numbers of guests. The
demand for this type of workers is submitted to seasonal fluctuations, with peaks
corresponding to the most auspicious periods for marriage ceremonies. Such employment
opportunities draw many houseless workers, especially for a supplementary seasonal or
temporary activity: while the occupational category of waiters. cooks and related service
workers*O accounts for 22 per cent ofthe main occupations reported by the respondents,
it accounts for 56 per cent of the supplementary occupations reported (Table 1).

On the other hand, marginal economic activities which are often associated with the
condition of the urban poor, iike rag picking and begging, concem only very minor
groups of the houseless surveyed in Old Delhi (2 and 1 per cent respectively in the
sample). Rag picking is more specificdiy canied out by children and teenagers.

A significant feature of the occupational pattern of the houseless is the general lack of
employment security or lack of parantee in getting sufficient work. Going daily or
periodically to the labour market to get recruited by a contractor or to the wholesale
markets and bazaar areas to find assignments, is the common fate of construction
workers, service workers for caterers, loaders and unloaders, handcart pushers (helping
the main puller and recruited by himi. and other casual labourers ... Among the houseless
working as employees, only a v e q small minority have secured salaried jobs. As for
self-employed workers like handcart pullers and cycle rickshaw drivers, they have to hire
their cart or rickshaw everyday ... uithout any guarantee about the number of trips -for
transport of goods or passengers - they will be able to get. The situation of street
vendors, roadside mechanics, and other self-employed in the informal sector, is similar
from the viewpoint of precariousness and correlative irregularity in income.

In fact, the combination of several occupations and flexibility in changing work - a rather
frequent practice as seen above - is a response to the risk of unemployment and
irregularity of work. Although, eventually, most of the houseless surveyed in Old Delhi
were not affected by unemployment during last year and, among those who were, only
few reported significant unemployment periods, the insecurity of employment and hence
the lack of guarantee of regular income, is a critical fact to understand the shelterless
situation of these workers (as we shall elaborate in the next section).

2o In the context of the survey, this occupational category corresponds mostly to service workers for
caterers, but it also includes cooks, waiters andrelated workers employed in restaurants and tea stalls.
 23

Income, debts, savings and remittances

Though the data pertaiiig to income, debts, savings and remitmces are based only on
the 36 in-depthinterviews with a sub-sample of houseless, some ncaningful features can
be stressed. At the ouset, the houseless workers appear to f o x a very heterogeneous
section of &e urban labur force in terms of earned money incomt: the average monthly
earnings (fw the last welve months preceding the 1996 survey1 range from Rs. 520
(case of achild rag picker) up to Rs. 4500 (case of a pavement dw-Ellerinvolved in illegal
trade combined with other legal occupations)”, with half of tkt respondents earning
between Rs. 1000 and Rs. 2000 per month: Furthermore, in=-occupational income
differentihare very l q e , as shown by the following examples?
- from Rs. 1100 to 2250 per month for cycle rickshaw drivers;
- from Rs. 900 to 2500 for handcart pullers;
- from Rs.500 to 20N €or handcart pushers;
- from Rs. 500 to 2700 for street vendors (incidentally, these two extreme cases sell the
same goods, namely pens);
- íì-omRs. 520 to 180G for rag pickers;
- from Rs. SOO to 1701 €or waiters, according to the place of work, in restaurants or for
caterers.
If certain @ r e s for t
k monthly earnings by occupation given a t ” are lower than the
minimum average monthly earning computed for the preceding year, it is because the
concerned individuals did not stay in this occupation throughont the year: they left it
precisely ia escape ficintoo exploitative conditions and changed for another better-paid
work.

With a vie%-to the discussion on urban poverty and houseless population, it is interesting
ro compare the average monthly earnings of the individuals s w e y e d with the poverty
line. We refer here to k e usual official definition (discussed later in this section), namely
the level ofexpenditux required for a daily calorie intake of 2,100 per person in urban
areas (as per the norm determined by the Nutrition Expert Groupì. and estimated at Rs.
310.50 per capita per month at 1995-96 price$. According ïo this criteria, all the

In fact thehighest incomt reported corresponds to the specific case of a m s i e n t pavement dweller, a
ticket checkerhaving a pmanent governmentjob in the railways, with a monthly salary of Rs. 4820.
and having k n transferred to Delhi three weeks ago. While waiting to get a gwemment accommodation,
he sleeps in= open -ground located just opposite the Old Delhi Railway Sta&m, his work place, hiring a
cot and bedding for the night
22 The mondIy incomes $\-en here correspond to an activity performed on a fuI1 time basis, but with
possible variarions in the number of working hours in a day and the number of work-days in a month.
23 This fig= is calculated on the basis of the last published official estimate, that is Rs. 209.50 per
capita per month in urban areas at 1991-92 prices, converted to present d u e by appIying the index
numbers of m u m e r pnca for industrial workers in Delhi for the correspondingperiod.
 I 24

I
respondents of our sub-sample are above the poverty line24, in so far w e consider that

II their earnings are available for their own and single consumption and they do not have to
share them with other family members not living with them in Delhi. Thus, to have a
better appraisal of the economic condition of the houseless, some indications about their
pattern of debts, savings afid remittances are required.

Incumng debts for everyday expenses does not emerge as a widespread nor regular
practice among the houseless interviewed, moreover, the amount borrowed is usually
limited (the highest recurrent amount reported was Rs. 150 per month). However, the
effective debts incurred could underestimate the real needs, for two reasons, pointed out
by some of the respondents themselves. On the potential debtor's side, the lack of
guarantee can be an obstacle to the pant of a loan, or at least is likely to put constraints
on the amount of the loan. On the potential creditors' side, the network of acquaintances
usually approached involves also some restrictions: the Co-villagers, co-workers or other
pavement dwellers known often share similar economic conditions, hence limiting the
amount of financial help they can extend; as for the employers or contractors, a certain
degree of regularity in the work relationship is required to enter in a debtor/creditor
agreement, which is precisely lacking in the case of casual labourers.

On the other hand, a large majority of the respondents report a saving practice, with a
varied range of situations, depending not only on the income level but also on the way of
living in Delhi and on the relations maintained with their famaes. The a v e r q a m o u n t of
saving varies from zero to Rs. 1100 per monthz, and in terms of percentage of income,
from zero to 65 per cent. There is no systematic inverse relation between savings and
- debts: individuals who do not save are not necessarily those who are often indebted, and
incurring debts does not necessarily prevent from regular savings. There is no simple
linear relation either between the capacity to save (measured as a percentage of income)
and the income level. For example, among the workers saving more than 50 per cent of
_ I

their monthly income, one can mention a beggar earning Rs. 800 per month as well as a
cycle rickshaw driver earning Rs. 2000 per month.

One strong motivation to save is remittance, and the houseless migrants who contribute
financially to the expenses of their families in their native place usually send the entire
amount of their savings or most of it. Hence, the amount,of remittances as well as the
corresponding percentage of income show ranges of variation similar to those obsemed

24 The statistical analysis of the entire sample w i l l allow us to determine to what extent this finding can
be generalised
25 Barring one exceptional case of Rs. 2000 per month corresponding to the government employee
already referred to in note 21.
 25

for savings. and m o n g the top remitters (proportionally to their income), we could
mention again the w o examples already given above. However, all the persons who save
more or less regduly do not remit money to their families (in the sub-sample of 36
kpondents: 24 E against 16 respectively, although 33 still have family members in
their place of ori&). In this matter, the circumstances of the migration and the nature of
rhe relations mainrrined with the family in the place of origin also intervene, as already
discussed in the prcvious section. ,

Retuming to the question of poverty line, we can consider now the average monthly
income available Ster remittances, which makes the comparison with the level of per
capita expenditure corresponding to the poverty line more relevant in order to appraise the
economic condition of the houseless population in Old Delhi. This net income shows a
range of variation wider than the gross income, from a minimum of Rs. 350 per month
(case of a beggar mding more than 50 per cent of his income to his family), to the same
mwimum of Rs. 2500. The majority of the respondents have a net monthly income
ranging benveen Rs. 500 and Rs. 1500 (as against half between Rs. 1000 and Rs. 2000
zs per their goss income). Yet, all the respondents in our sub-sample, including those
engaged in marginzl economic activities like rag picking and begging, remain above the
poverty line of Rs. 3 10.50 per capita per month, considered as the expenditure required
for a daily calorie k - d e of 2,100 per person in urban areas. Thus, the incomes generated
by the houseless inz3iewed in Old Delhi place them in a position to meet their minimum
numtional requirements. But other basic Reeds, especially shelter and clothing, which
should be also cocsidered in a comprehensive approach of urban poverty, are not
nccessarily coverei- The deprivation of shelter is obvious in the case of the pavement
dwellers; as for clo5ng, the sartorial condition of some of them testifies by itself to the
non fulfilled needs. Thus, several respondents acknowledged that they had only one set
of clothes, to be w o 3 until the clothes are completely torn and have to be replaced by a
new set. Nonetheles, the lack of purchasing power is not the only explanatory factor; the
lack of safe place i o keep one's belongings is another major problem faced by many
houseless people26.

The official definition of the poverty line, based on a nominal level of calories, has
already attracted criticism (NIUA, 1986: 11-12). And our finding that the houseless
people surveyed in Old Delhi stay above the poverty line can be considered as another

26 Some houseless peqle manage to keep their belongings (and savings) at their work place; others
confide them to a shopkeeper or a tea-stall owner's care; or resort to the safekeeping of another
acquaintance settled in Delhi under better housing conditions; a few pavement dwellers mentioned the
washerman for storing meir spare sets of clothes; as for savings only a few respondents keep them in a
bank account; .... while other houseless people have no alternative solution but to keep the very few
things they possess with them .....or have no belonging nor saving to look after.
 26

evidence of the limitation of this official definition. A subsequent question arises here: if
the pavement dwellers in Delhi -at least a section of them- are above the poverty line, who
and where are the ”poor” of Delhi, namely the 17 per cent of the populatidn of the capital
estimated to be under the poverty line” ? Assuming that most of the urban poor are
likely to be concentrated in the city slums (h?UA, 1989), it would mean that the
houseless people living alone on the pavements in Old Delhi are in a better position to
meet their minimum numtional requirements than many slum dwellers living with their
families and having housing accommodations,precarious though they may be. This calls
for a better scrutiny of the survival strategy of the pavement dwellers (as it will be
developed in the next section).

Assets in the native place and investment projects

The economic condition of the pavement dwellers and night shelter inmates in Old Delhi
has also to be considered in a larger perspective than their present stay in the city as
homeless people, and replaced in the context of the economic situation iptheir native
place.

The first interesting point to underline is that the houseless do not hail necessarily from
the poorest rural families. Most of the respondents for in-depth interviews reported a
familial house in their native place, and the majority of them had familial agricultural land,
though generally small properties. Certain respondents even hail from well-off families,
their migration in that case is more likely to have been motivated by familial disputes than
5,
by economic factors. The familial assets in the native place bear some relevance to
appraise the current economic condition of the houseless in Delhi only for those migrants
who have maintained relationships with their families, still the majority of them (see
previous section).

Projecting now the houseless people into the future, among the migrants who wish to
return to their native place, future plans of investments there are frequent, in particular to
open a general store or another type of shop in the village, to buy more agricultural land
or more generally to invest in aZg5culture. To realise their investment project, the
individuals interviewed plan to raise funds from their own savings, supplemented if
necessary by familial contribution. Some of these investment projects may not be
realised, yet some seem viable given the saving capacity of the concerned workers. This

27 This percentage corresponds to the 1987-88 estimate given in the Reporr of the Expert Group on
Estinlafion of Proportion and number of Poor, Planning Commission, Government of India, July 1993.
 27

shows a definite degree of economic dynamism among certain houseless persons, and
conveys an image which is not that of abject poverty, but more of an economic
calculation entailing a temporary sacrifice in t e m of housing conditions in lhe city, in
o r e r to improve the economic conditions in the native place.

*
This first investigation of the economic profile of the houseless people in Old Delhi
allows us to highlight some significant features. At the outset, this shelterless population
proves to be an integral part of the normal menopolitan labour force, wZlich is in
acordance wirh previous studies on pavement dwellers and night shelter inmates in Delhi
(ser ARORA B CHHIBBER, 1985: 5). Furthermore, the incomes generated by the
hocseless workers surveyed place them in the low-income group but above thepoverty
l i n c even after taking into account the remittances sent to the family outside Delhi. In
prolortion to their income level, the saving capacizy of the majority of the houseless is
also far from being marginal. Remittances to the family and future plans of investment
a f t c returning to the native place are other encouraging indicators of the economic
potential of a good number of them. In this respect, the findings of om survey
corraborate the conclusions of other studies on the urban poor, as underlined by Kundu
(1943: 23): "the thesis regarding economic marginality of the people in urban informal
sectors, slum dwellers, pavement dwellers and other is an exaggerationq8. Yet,
insecurity of employment or uncertainty in getting sufficient work, and hence the lack of
guaranted and regular income, constitute a widespread concern. A last striking fact
rega-ding the economic condition of the houseless peaple in Old Delhi is the large variety
of sltuations encountered, including very contrasting cases. Thus, qualifying the
pzvement dwellers as the "poorest of the urban poor"' without further distinction appears
as a simplistic statement.

6 -EXCLUSION FROM ACCESS TO A DWELLING VERSUS

RESIDENTIAL STRATEGY

In this section, we shall further investigate the present shelterless situation of the
paveznent dwellers surveyed, including: circumstances of the arrival on the pavement or
in a night shelter, residential mobility in Delhi, choice of sleeping places - especially in
relation to workplaces and employment opportunities, and willingness to move and to

**Quoted from: LEE (M.), 1986. "The mobilisation of informal sector savings: the USAID experience",
paper presented at the International Workshop on MobiIisation of Informal Sector Savings, 8-12
December, Societyfor Development Studies, New Delhi.
 28

pay for a dwelling. We shall attempt to appraise the elements of constraint and the
elements of choice, with an underlying question: is the shelterless situation merely the
consequence of a process of exclusion from access to the urban housing system? or does
it correspond also to a residential strategy aimed at improving individual or familial
economic conditions? Another question pertains to the perennial versus temporar). nature
of the shelterless situation: is it a transitory stage preceding a better integration in the
urban housing system ? or does it correspond to a permanent way of life or at least a long
lasting one for the duration of the stay in Delhi?

Residential trajectory in Delhi

Several types of residential trajectory eventually leading to staying on the pavement or in

a night shelter can be identified. In the most frequent pattern, the shelterless siruation
started upon the arrival in Delhi as a migrant and continued thereafter: without knowing
any altemative place to stay, and given his financial constraints, the migrant resolved to
sleep on the pavement or any open ground, or approached a night shelter foliowing some
indications grasped from people at the railway station, bus terminal, or after a few days
of exploration in the city. Moreover, the frrst stay as shelterless in Delhi corresponds also
in most of the cases to the first mipration in the capital. Another noticeable trajectory,
though less frequent, starts in Delhi with a stay at the workplace (restaurant. shop,
factory, workshop, construction site...) and ends on the pavement or in a night shelter
'
following the termination of the work contract or in search of better employment. As
already noticed in a previous section, there are also a few cases of migrants staying
initially with their relatives, and who were compelled to leave due to lack of space or for
better proximity to their workplace. A last mention is also deserved for the cases of
natives of Delhi who fled from their home following familial problems, especially
children, and stayed on the pavement for lack of alternative accommodation

Financial constraints

, Financial constraints are put forward by the large majority of the houseless people
I
1 interviewed to explain the origin of their shelterless situation. However, the significance
l of this factor has to be appraised in relation with other contributing factors and to be
1,
I considered in a long term perspective. The financial constraints are likely to be more
stringent at the initial stage of pavement dwelling in particular when it corresponds to the
arrival in Delhi, and that the migrant has to get absorbed into the labour market. Then the
 29

financial constraints may result not as much from the average level of income than from
the lack of guarantee of regular income. This element of uncertainty restrains many casual
workers staying on the pavement or in a night shelter to envisage an accommodation on
rent even if they have the financial potential for it, because this would.entai1 regular and
fixed expenses which cannot be adjusted to the actual earnings. On the other hand
expenses for hiring a quilt or a cot outside or for entrance to a night shelter (for those
availing these facilities) are incurred on a daily basis and hence can be easily adjusted LO
the daily earnings. Even for the houseless with a sufficient saving capacity to rent a room
[alone or by sharing it with one or two other workers), priority may be given to
remittances to the family or to long term savings for future investment in the native place.
In other words, preference may be given to the family's living conditions in the native
place over the migmt's living conditions in Delhi, and to future over present. In that
case, it cannot be said that there are absolute financial constraints preventing the
houseless migrant from renting a room, but rather relative ones resulting from his ONTI
choice and priorities. The priority given by the houseless migrants to remittances over
their present housing conditions in town does not mean however that they do not fulfil
first their own basic needs in terms of food requirements, in order to insure the
reproduction of their own work force. This may explain why the houseless workers
living alone in Old Delhi are able to stay above the poverty line (as per the definition
$ven above), while it may not be the case for many slum dwellers living with their
families.

Proximity to the workplace

Another major factor to understand the shelterless situation of many workers in Old Delhi
and the choice of their sleeping place, is the location of their workplace. Better proximity
to the place of work or the source of employment opportunities is one of the reasons
reported explicitly by certain respondents for staying initially on the pavement or in a
night shelter, and more frequently to explain the choice of a specific place to sleep. The
actual 'residential' location of the houseless is more revealing than these explicit answers.
Thus, most of the respondents interviewed in Old Delhi work in the Walled City itself or
in adjoining areas, within a walking distance from their place of sleep, often within 10
minutes walk or less. The cost of transportation to commute to the workplace is
consequently reduced to nil for them. Even among the respondents who said they
exercised preference in their location's choice for the sleeping environment, especially the
presence of acquaintances and the facilities available, most of them stay also de facto in a
walking distance from their place of work or labour market.
 30

The importance of staying close to the source of employment opportunities depends on

the type of occupation. For the casual workers having to go daily to a labour market to
get recruited, like workers for catering services and construction labourers, this
proximity factor appem primordial. In order to get more job offers, and to be in a better
position to bargain with the contractors, it is necessary to reach the labour market early
morning, and hence not EO have to spend time on commuting.
For handcart pullers or pushers and loaders working in market areas, transportation
activities do not start very early in the morning (usually at 10) but they often continue till
late in the night, which makes it more convenient to sleep in the market itself. and more
profitable to get assignments. Since the nature of the work requires intense physical
strength, the transportation workers are usually exhausted after long working hours,
hence sleeping at the same place (or nearby) enables them to avoid the additional tiredness
of commuting. In addition, in the market they can sleep on their handcarts or under the
verandas of the buildings, and do not face harassment by the police since they are known
to work there.
The position of the cycle rickshaw drivers (whose nature of work is also physically
demanding) is mixed. Those keeping their rickshaw at night can sleep on it and have more
flexibility in their location’s choice within the zone where they ply, providing they can
park their rickshaw safely. But for the drivers having to take their rickshaw every morning
from the owner‘s garage, staying in proximity to the garage is also important, as explained
by one respondent: “ r f y mstay fur, you have to srurt early iii the niornittg utid do not get
pi-oyer rest. And if you start late, your work suffers und your inconle becomes less”.

Even for the houseless workers whose nature of occupation and mode of recruitment do
not require necessarily to stay near the source of employment opportunities, proximity
between sleeping place and workplace is sought-after in order to reduce - or cut entirely -
transportation expenses for commuting.

Residential mobility and duration of stay in Delhi

Examining the residential mobility in Delhi of the houseless people allows us to highlight
some revealing facets of their living conditions. Change of residential location during the
year is a frequent practice among the pavement dwellers surveyed in Old Delhi, according
to two main reasons.
 31

Seasonal pattern can be first of ah observed, with various possible combinations. For
example, in summers preference is given to sleeping on open grounds, non covered
pavements. road dividers, or in parks; during the rainy season, verandas, night shelters,
cots or mattresses on hire in sleening areas protected with ground and overhead plastic
sheets, are more in demand; and ia winter preference goes to quilts on hire, night shelters
and verandas. For the pavement dwellers who never avail bedding facilities on rent and
never go to night shelters, the change of sleeping place may be limited to crossing the
road, from a veranda in winter ari during the rains to a nearby park in summer, or even
to the divider of the same road. Some pavement dwellers have been repeating the same
seasonal pattern with the same combination of locations for years. This flexibility in
sleeping places also helps us understand how the pavement dwellers who can afford to
spend Rs.3 per night for access to a night shelter, or Rs. 5 to hire a quilt, up to Rs. 15
for cot and bedding, are not neceszuily ready to spend the equivalent monthly amount to
rent a room. Whereas taking a roc,m on rent entails regular and fixed expenses throughout
the year, resorting to a night shelter or bedding on hire involves daily expenses which can
be entirely cut during certain periods of the year.

The second main reason for residential mobility in Delhi is directly related to occupational
mobility. The houseless people a&o adjust the location of their sleeping place according
to employment opportunities, to the location of a particular labour market, or to the
possibilities of sleeping at the workplace. The rationale behind these changes of
residential location is the search for a better proximity to the place of work or the source
of potential employment.

On the other hand, there is also a category of houseless people staying in the same night
shelter or on the same section of pavement throughout the year (when in Delhi) and for
several years, with duration of stay extending in a few remarkable cases up to ten years
or more.

Considering now the entire duration of stay in Delhi as houseless, there is a notable
proportion of houseless people who have been living under this condition for ten or more
than ten years29 - which does not exclude however regular stays in the native place. For
this category of pavement dwellers, and especially for those staying in the same night
shelter or sleeping place for many years, or repeating exactly the same seasonal pattern
sometimes for the last ten or even fifteen years, their houseless condition seems to be a

2g In the sub-sample of 36 respondentsselectedfor in-depth interviews, one third have been living in
Delhi as houseless since 1985 or before. The actual proportion will have to be confirmed by the
statisticalanalysisof the entire sample.
 32

permanent way of life in Delhi, or for the duration of their working life befare the definite
return to their native place.

Willingness to move and to pay for a dwelling

Investigating the future plans of the houseless people, in particular thee willingness to
move and to pay for a dwelling, can bring further elements of answer to %e question of
the perennial versus transitory nature of the shelterless situation in Delhi.

Among the respondents of our sub-sample, the capacity and willingness to pay for a
dwelling varies from zero to Rs. 500 per month30, and in terms of percen2ge of income,
from zero to 35 per cent maximum. However, the willingness to move as expressed
through the answers of the respondents may remain in a good number ai cases abstract
wishes, subject to several conditions. The constraints are sometimes so 5mngent that
they make the possibility to move very unlikely. Here we find again the factors whose
importance has been already underlined to understand the shelterless sitetion, namely:
financial constraints, r e p l a r incomes or even getting a permanent job as E Trerequisite to
move, dwelling in proximity to the place of work. Among the house:.tss who have
already attempted to move to a proper dwelling (a minority group), <XE reasons for
eventually returning to the pavement or a night shelter are also interestir, to point out.
Some mentioned difficulties to adjust with other persons with whom r.ky shared the
same room - a common way to reduce housing expenses. Others realised iîat their work
and income were suffering from the fact of staying too far from their Flace of work.
Hence, wishes and even actud attempts to move are not sufficient indiczors to foresee
the shelterless situation of the concerned individuals as a transitory stag? preceding a
better integration in the urban housing system.

The arguments of the houseless persons who stated that they had no v.3lingness nor
intention to move to a better accommodation, and who form an appreck-ie group, are
also revealing in order to appraise the transitory versus perennial nature of -;?e shelterless
situation. Interestingly, those reporting a 'zero' willingness to pay for a dxelling are not
necessarily those who have no saving capacity and hence no financial pcrtntial. In that
case, two types of reasons are put forward to explain the lack of willingness to pay for a
dwelling. Some pavement dwellers do not intend to stay in Delhi for a lozg period, and
hence would find it irrelevant to take a room on rent. Others give a clear priority to the
maximisation of their savings, especially in order to send remittances to rheir families,

30 Barring one exceptional case of&. 1400 per month, corresponding again to the govemment employee
already mentionned, who earns a monthly salary of Rs. 4 820.
 33

and hence try to minimise expenses for housing and transportation, or even to reduce
them to zero. Being alone, without their family in Delhi. some of the houseless migrants
do not perceive a proper accommodation as a need for thcm. As a matter of fact. most of
the houseless interviewed do not plan to settle in Delhi permanentli, but to return one day
to their native place or to migate to another city. Hence u e y perceive their stay in Delhi
as limited in time, even if this transitory situation may e~sntuallylast for the duration of
their working life31, and consequently a dwelling is not perceived as a priority need.

Coming back to the questions raised at the beginning of this section concerning the
interpretation of the shelterless situation, we can sum up the main arguments which have
emerged at several points from the in-depth interviews with houseless people in Old
Delhi. Financial constraints undoubtedly prevent or limit the possibilities of access to a
dwelling. Nevertheless this factor has to be considered in combination with other
explanatory factors, forming a system in which choice is often present. The residential
practices of the majority of the houseless reveal an economic rationale that aims at
"ising savings and reminances to their families in their native place, by minimising
their housing and transportation expenses. Proximity between the sleeping place and the
place of work provides another important clue in understanding the shelterless option and
choice of location by the pavement dwellers. A location near the workplace or the labour
market enables them not only to cut their transport expenses but also to avoid a tiring
commute. This ensures a proper rest, which is vital €or manual workers involved in
occupations demanding intense physical strength. In addition, for casual labourers,
proximity to the source of employment opportunities also increases the probability of
ge&g daily work. When the logic of staying shelterless in Delhi is an integral part of
familial strategies rooted in the native place, priority being given to the economic
condition of the family in the village at the expense of the migrants' living conditions in
Delhi, then the shelterless situation is likely to last for the duration of their stay in the
capital.

The part of rational choice in the residential practices of the houseless people has been
highlighted in other studies - although the exercise of choict for this segment of the urban
population is restrictea to a choice under strong economic and social constraints. For
example, in the conclusion of his primary survey of pavement dwellers and night shelter
inmates in Delhi, Kuruvilla states: "The choice of the pavement is mainly for reduction of

31 ....O
r may be even their entire remaining life, in so far the future plan of returning to the native place
mightbe in some cases more mvthical than realistic (as suggested above).
 34

expenses on housing, proximity to employment opportunities, (....), availability of

facilities, services, food and water and maximise savings to send back home. Thus it
becomes a deliberate rational decision to live on the pavement" (1991-92: 85-86).
Jagannathan and Halder, in their study of the pavement dwellers in Calcutta, also infer:
"Pavement dwellers of the main stream vocations have chosen this life style to protect
their access to earning opportunities. In addition (...) a substantial proportion are
temporary migrants, who remit savings home to the village". Further: "The majority of
pavement dwellers live without shelter as a deliberate rational decision, by which the
expenditure on housing is reduced to zero" (1988-a: 1177).

The importance of proximity to the source of livelihood is also emphasized in the two
studies quoted above. This factor is crucial not only to understand the residential pracdces
and location choices of the houseless, but more generally of the urban poor3'. The failure
of many attempts to relocate slum and squatter settlement dwellers far from their initial
residence is thus due to inadequate consideration for easy physical access to earning
opportunities.

7. SUMMARY OF MAIN FINDINGS

Old Delhi, the historical core of the capital city, is characterised by extremely high
population densities combined with a remarkable concentration of commercial and
manufacturing establishments. While a process of population deconcentration from the
old and deteriorating housing stock is at work, economic activities have proliferated. This
has attracted a floating population of male migrant workers, most of them unskilled,
whose residential integation is extremely precarious. Surveys conducted on a sample of
shelterless persons in the Old City allowed us to examine their residential and related
economic practices and contribute to the discussion on the social and economic
marginality of the houseless population.

Only a few have broken away from the basic social institution, the family. These are
those who have left their homes foIlowing acute familial tensions, especially with
parents. Although they live alone in Delhi, the majority of the houseless migrants
maintain a regular link with their families in their native place, which remains their pIace
of reference.

32 See for example: GUPTA, KAUL, PANDEY (1993: 86),SURI (1994: 273). KUNDU (1993: 65).
 35

The houseless population of Old Delhi also forms an integral part of the metropolitan
labour force, which, in terms of income, seems to be able to stay above the poverty line.
Yet, the lack of guaranted and regular income constitutes a general concern. At the same
time their saving capacity, remittances and plans of future investment, all represent
encouraging indicators of the economic potential of a notable share of the houseless.
Finally, thelarge variety of individual situations encountered indicates that the houseless
are not a simgle category of 'urban poor', nor are they necessarily 'the poorest of the
urban poor'.

Although financial constraints form the background of the shelterless situation, the
residential practices of pavement dwellers and night shelter inmates should not be seen
only as the consequence of a process of exclusion from access to a dwelling. One should
also appreciate the economic rationales of individual migrants who try to maximise
remittances to their families in the village by cutting their housing and transport expenses.
Priority is therefore given by them to a location near the workplace or near the labour
market. For casual labourers, in fact, proximity between the sleeping place and the source
of employment opportunities often increases their probability of getting daily work. Thus
the condition of the houseless has to be seen in relation to their needs and priorities. This
is a prerequisite for formulating appropriate urban housing policies.
 36

REFERENCES
ALAM (S. Manzoor), ALIKHAN (F.) (eds.), 1987. Poverty in Metropolitan Cities,
Concept Publishing Company, New Delhi.
ARORA (R.K.), CE-IHIBBER (A.K), 1985. "Chilled nights on pavements in Delhi. A
socio-economic study of the children staying in ran baseras in winter". Socio-economic
division, Slum and Jhuggi-Jhompri Department, Delhi Development Authoriry, New
Delhi, mimeo.
BANERJEE (El.), 1986. Rural to Urban Migration arid the Urban Labour Market (A case
study ofDelhi). Himalaya Publishing House, Delhi.

Bharat Sew& Samaj, 1964. "A roof over the head", Delhi School of Social Work, Delhi.
DDA, 1989. "Survey of pavement dwellers in Old Delhi", unpublished study, Slum and
Jhuggi-Jhompri Department, Delhi Development Authority, New Delhi.
DDA, 1993. Zonal Plun Walled City. Draft.Delhi Development Authority, New Delhi.
DDA, 1994. "Prog"e of night shelters for the homeless in Delhi.", Slum and Jhuggi-
Jhompri Department. Delhi Development Authority, New Delhi, mimeo.
DHAR (P.), 1985. "Pavement dwellers in Calcutta. Strategies for rehabilitation and
urban development". Unpublished thesis, School of Planning and Architecture, Delhi.
DUPONT (V.), 1995. "Patterns of population mobility in the Delhi Metropolitan Area
and their impact on the urban dynamics. Presentation of the system of investigation",
Paper presented to the seminar on 'Migration and urbanisation in Vietnam', Ho Chi Minh
City, 5-3 December.
GANDOTRA (S.R.), 1976. Houseless in Delhi, Census of India 1971, Series 27,
Delhi, Part X (a), Special study, Government of India, Delhi.
GUPTA (D.B.), KAUL (S.), PANDEY (R.), 1993. Housing and India's Urban Poor.
Har-Anand Publications, New Delhi.
HSMI, 1988. Renewal of historical housing stock in Old Delhi. Indian Human
Settlement Programme, HSMI Studies 1, Human Settlement Management Institute, New
Delhi.
ISI, 1977. "Calcutta 1976: A socio-economic survey of pavement dwellers", Indian
Staitstical Institute, Calcutta.
JAGANNATHAN (N.V.), HALDER (A-), 1988-a - "Income-housing linkages. A case
study of pavement dwellers in Calcutta". Economic aiid Political Weekly, vol XXIII, No
23 (June 4 ), pp. 1175-1178.
JAGANNATHAN (N.V.), HALDER (A.), 1988-b. "A case study of pavement dwellers
in Calcutta. Occupation, mobility and rural-urban linkages". Economic and Political
Weekly, vol XXIII, No49 (3 December), pp. 2602- 2605.
JAGANNATAHN (N.V.) HALDER (A.), 1989. "A case study of pavement dwellers in
Calcutta. Family characteristicsof the urban poor". Ecoizoniic and Political Weekly, vol.
=IV, No 6 (February Il), pp. 315-318.
 37

JAIN (A.K.), 1996. The Indian Megacity and Economic Reform. Management
Publishing Co,New Delhi.
KUNDU (A.)7 1993. In the Name of the Urban Poor. Access to Basic Amenities. Sage,
New Delhi.
KURUVILLA (J.), 1990-91. "Pavement dwelling in metropolitan cities. Case study
Delhi". Thesis, School of Planning and Architecture, Dept. of Housing, New Delhi,
Mimeo.
MEHRA (A.K.), 1991. The Poliiics of Urban Redevelopment. A Study of Old Delhi.
Sage Publications, New Delhi.
MUKHERJEE (S.), 1975. "Under the shadow of the metropolis - They are citizens too".
Calcutta Metropolitan Development Authority & Calcutta Metropolitan Planning
Organisation, Cacutta, mimeo.
NAGPAUL (H.), 1996.Modernisation and Urbanisation in India: Problems and Issues.
Rawat Publications, Jaipur and New Delhi.
NIUA, 1986. Dimension of Urbanpoverty. A Situational Analysis. National Institute of
Urban Affairs, Research Study Series No 25, New Delhi.
NIUA, 1989. Profile of the Urban Poor: an Investigation into their Demographic,
Economic and Shelter Characteristics. National Institute of Urban Affairs, Research
Study Series No40, New Delhi.
PERNIA (E.M.) (ed.),1994. Urban Poverty in Asia. A Survey of Critical Issues. Oxford
University Press, Hongkong.
RAMACHANDRAN (P.), 1972. "Pavement dwellers in Bombay City", Tata Institute of
Social Sciences, Series N026, Bombay.
SINGH (A. M.), SOUZA (A. de), 1980. The Urban Poor, Slum and Pavement Dwellers
in Major Cities of India, Manohar, Delhi.
SPARC, 1985. "We the invisible", Report on pavement dwellers in Bombay, SPARC,
Bombay.
SOUZA (A. de) (ed.), 1983. The Indian City. Poverty, Ecology and Urban
Development. Manohar, Delhi.
'
SURI (P.), 1994. Urban Poor, Their Housing Needs and Governement Response. Har-
Anand Publications, Delhi.
TCPO, 1975. Seminar on Redevelopment of Shahjahanabad: the Walled City of Delhi
(New Delhi, 3Ist Jan.- Ist Feb. 1995), Report and Selected Paper. Resume of the
Seminar. Town and Country Planning Organisation, Government of India, Ministry of
Works and Housing, New Delhi.
 38

APPENDIX

Presentation of the research programme:

PATTERNS OF POPULATION MOBILITY IN THE DELHI
METROPOLITAN AREA

1 - Context of the study

Promoted the capital of the British Indian Empire in 1911, then capital of independent
India in 1947, Delhi is the Indian metropolis (over a million inhabitants in 1981) which
has experienced the highest demographic growth over the last decades. Its population has
expanded from 1.4 million in 1951 to 8.4 million in 1991. The rate of growth, however,
slows down over that same period: 5.1% per year from 1951 to 1961, 4.5% to 4.6%
from 1961 to 1981, and 3.9% per year between 1981 and 1991.
Today, Delhi is a city i n full expansion: although traditionally, it has always had a
political and administrative role to play, it is now becoming one of the country's major
centres of economic growth. In addition, the capital has pioneered original experiences in
town and country planning, with a rigorous planning of the city, policy of urban public
transport, the creation of satellite towns, measures of land control, relocation of slum
dwellers. However, despite these policy measures, speculation on land and formation of
slums could not be avoided. Delhi thus provides an example of how a capital with several
millions of inhabitants has developed, with a sustained growth in the context of wilful
urban policies.
Moreover, this capital's dynamics is part of an urbanisation process that has two
characteristics at a national level. Firstly, with a level of urbanisation that is relatively
moderate (26% in 1991), India remains a predominantly rural country - and will remain
so in the medium term. Secondly, the urban scene is dominated by the big metropolises,
and the urbanisation process has gone hand in hand with a growing concentration of the
urban population in the metropolises of a million- and multi-million inhabitants.
In Delhi, like in other large metropolises in the world, the process of urbanisation seems
to be linked to other specific phenomena:
- the deconcentration of the urban core;
- a process of suburbanisation, with the rapid development of suburbs and the formation
of satellite towns;
- an increased differentiation ahd segmentation of the metropolitan area;
- the development of commuting between place of residence and place of work as well as
the development of other forms of circular mobility between different places of residence
inside and outside the town, resulting in complex and multi-polar spatial residential
patterns which go beyond the urban/rural dichotomy.
The development of a metropolis like Delhi, with several millions of inhabitants, raises
crucial and specific problems in terms of urban administration. Furthermore, the rate of
demographic growth, in spite of its slow down, remains quite significant, which
compounds the task of town planners.
Understanding how cities function and, in the f i s t place, how their population dynamics
function, is an indispensable prerequisite for any attempt at town planning. In this
research project on Delhi, the study of population movements is used as a preferential
tool to analyse and understand better the dynamics of the metropolis: its development as
well as its intemal structuring and transformations.
 39

2. Research objectives
The main objective of this research project is to improve our knowledge of the different
forms of spatial mobility and of the complex spatial residential patterns in a large
metropolis, and to analyse the different types of mobility practised by the population in
relation to its absorption into the urban labour market and conditions of access to a
dwelling.
Then, it aims to analyse the impact of such residential and occupational strategies on
urban dynamics, and this at three levels:
- Impact on the global dynamics of the town, in terms of geographical spread of
the urban agglomeration and development of suburbs and satellite towns, that is, more
generally, impact on the process of metropolisation and suburbanistion;
- Impact on the internal structure of the metropolitan area, in terms of population
redistribution within the urban space and in terms of differential dynamics of certain
neighbourhoods, and, more generally, role of residential strategies in the process of
spatial segmentation ;
- By putting Delhi into its regional context, the different forms of population
mobility will be also examined in order to better undexstand the structure of the exchanges
that take place between the metropolis and the neighbouring states of North India, or
other places of the national territory, or even international space.
In this perspective, we shall also examine the links between the urban policies at the
macro-level and the individual's residential practices at the micro-level.
The following two questions guide our investigation:
- What are the residential practices developed by the population, its strategies as regards
the occupation of the geographical and economic space of Delhi metropolitan area? What
are the occupational, familial, etc. .. determinants of such practices?
- What impact do these residential practices have on the global and intra-urban dynamics
of Delhi?
Particular attention is given to the forms of temporary and circular mobility and their
evolution. The permanent versus transitory character of circular mobility will be
questioned: do these complex residential pattems represent a step towards a settling down
into the city or do they tend to become perennial? What lessons can be drawn as to the
impact of such pattems on urban dynamics?
The objective of such research in the Indian national context is also to understand to what
extent the development of the forms of circular mobility can help to maintain a relatively
moderate rate of urbanisation.
The research project on Delhi is also conducted in the perspective of an international
comparison with another metropolis of the developing world, Bogota ( C ~ l o m b i a ) This
~~.

33 The research programme on Bogota is conducted by Françoise DUREAU (ORSTOM, France) in

collaboration with the CEDE (University of Andes, Colombia). The comparative programme on the
residential practices of the popdations and their impact on the dynamics and segmentation of large
metropoliseshas received a financial support from CNRS - PIR-Villes (Paris).
 40

will also allow us to compare the different patterns of spatial mobility in two large
metropolises which face some similar problems although they belong to contrasting
geographical and socio-political contexts.

3. Research methodology
The official systems of demographic data production in India (the censuses and the
National Sample Survey) do not provide adequate information on the different forms of
spatial mobility in relation to urban dynamics: first they focus mainly of migration seen
as a relatively permanent transfer of residence, and secondly the level of aggregation of
the data on migration does not allow a differential analysis of the neighbourhoods inside a
given urban agglomeration. Hence, in addition to making use of the bibliographical and
statistical data available on Delhi, this research programme relies essentially on specific
surveys carried out on samples of population in seven selected neighbourhoods, in Delhi
urban agglomeration and in two satellite towns of the metropolitan area, and covering
different types of settlement including a sample of houseless people.
In order to get a better understanding of the urban transformations linked to the residential
behaviour and to the spatial mobility practised by individuals and their families, the
system of observation follows three basic principles:
- To take into consideration the set of all types of mobility, regardless of the distance
(intra-urban movements in the metropolitan area and migrations towards and from Delhi)
or the duration of the movement (permanent or temporary migrations and daily
commuting).
- To introduce a longitudinal approach, which helps us to understand the way in which
people combine different residential practices throughout the different stages of their life
cycle.
- To take into consideration the family units in the observation and-analysis of the
migratory practices, in order to relocate individuals’ mobility behaviours into their
familial context, and thereby recover the collective dimension of the mobility logic,
evidenced in numerous socio-anthropological works.

Only a combination of the quantitative and qualitative approaches can adequately meet the
three principles stated above, and help us to determine the direction and socio-
demographic compositions of the principal population flows towards, from and inside the
Delhi metropolitan area, as well as the individual and family strategies under which these
population movements take place. Thus, the system of observation adopted combines a
quantitative and a qualitative approach, and includes two main phases:
A statistical survey was carried out on a sample of approximately 1,700 households
in seven selected zones of the metropolitan area. The data collection was based on a
structured questionnaire, and information was collected regarding:
- housing conditions,
- demographic and socio-economic characteristics of the members of the household,
- commuting between the place of residence and the place of work or study,
- main migratory steps,
-temporary moves in and outside Delhi,
- characteristics of family members not living in the household surveyed.
The general survey was conducted from mid-February to the end of April 1995 in six
zones, and covered a total sample of 1413 households. The specific survey of pavement
dwellers in Old Delhi was conducted in January-February 1996, and covered a sample of
248 houseless people.

...
 41

An anthropological observation supplements the quantitative observation, and

consists of semi-directed interviews carried out on a sub-sample of people selected from
the file of the statistical survey. The purpose of these in-depth interviews is to collect
detailed migration, occupational and familial histones, as well as information on the
circumstances and reasons of the rni-mtion to Delhi. the conditions of absorption into the
C i e and the urban labour market, and the relations maintained with the native place. This
will allow us to deepen in the understanding of the population's mobility behaviours.
Till today, about 50 in-depth interviews were conducted: with houseless people in Old
Delhi, and with inhabitans (natives as well as migrants) of an urban village of NOIDA.
This will be completed by 2 other series: one with slum dwellers in a centrally located
area, and the other with occupants (owners as well as tenants) of flats built by the Delhi
Development Authority.
This two-fold observation of spatialmobility and residential practices was supplemented
by the collection of basic information on the contextual. background of each
neighbourhood selected for the survey: civic amenities available, land use pattern, history
of the neighbourhood, how the area has developed over the last decades in terms of
housing, civic amenities and economic activities, whether it has benefited from specific
urban policies or whether it results mainly from non govemmental initiatives (including
unauthorised settlementsj. This wilt allow us to analyse the extent to which the main
characteristics of the neighbourhood (in terms of housing and economic activities in
particular) may influence certain residential strategies at the individual and household
levels, and, reciprocally, to which estent the residential strategies have an impact on the
dynamics of the neighbourhood and the structuring of the city.
I .' .

Recent Working Papers

Title Author (s) Name Paper No.

Water Pollution Abatement: A.J. James E/177/96

A Taxes-and-Standards M.N. Murty
Approach for Indian
Industry.
Natural Resource Accounting: M.N. Murty E/178/96
Measuring Value Added by
Environmental Resources with
Illustrations from Indian
Industry.
Accounting for Costs of Smita Misra E/ 179/96
Water Pollution Abatement:
A Case Study of Nandesari
Industrial Area.
Environmental Impact: A A.J. James E/180/96
Basis for Increasing Costs
in Wheat Cultivation in
the Punjab?
Water.Pollution Abatement R.N. Agarwal E/181/96
and Industrial Growth: A B.N. Goldar
Study of Distilleries in D.B. Gupta
India.
Environmental Degradation, Kanchan Chopra E/ 182/96 4

Property Rights f Population S.C. Gulati

Movements: Hypotheses and
Evidence from Rajasthan.
Structural Reforms and Moneer Alam E/183/96
Employment Issues in S.N. Mishra
India: A Case of Industrial
Labour.
Linking Indicators of Wietze Lise E/184/97
Participation in Forest
Management and Additional
Incomes From Participating: -
An Estimation Procedure for
Non-Cooperative Games.
Measuring Benefits from Smita Misra E/185/97
Industrial Water Pollution
Abatement: Use of Contingent
Valuation Method in Nandesari
Industrial Area of Gujarat in
India.