SDA-RF-CH-1B.12.pdf
Media
- extracted text
-
' •'
• ' ■
SDA-RF-CH-1B.12
•
-------aj-e
explains how tins figure was reached. Our for- rice work in Banidhaman
district Almost all seasonal migrants travelWO*1.
observations and estimates suggest that poor, though most cultivate on their own bands or gangs. Gangs in the streams
several hundreds of thousands of people account back home and the livelihoods of originating in Dumka district, Puruliya/
also migrate in the three other main sea many involve multiple activities, includ Bankura districts and Purbi Singhbhum
sons for employment in rice work.
ing, in different combinations, petty trade, district8 were usually mixed sex and often
Our ethnographic research, including life processing, gathering and study. Migrants included children and infants. In contrast,
history interviews, in the study localities7 tend to be in scheduled social groups, almost all gangs in the Murshidabadsuggests that seasonal migration has in- whether castes or tribes, or according to Barddhaman stream were entirely male
creased into the destination locality and religion,
with no accompanying children. Some
out of three of the source area localities
Table 2: Characteristics of the Main Migration Seasons
over the last 30 years. Although in the ____
fourth source area locality, in Purbi Season
Conditions
Months
Singhbhum district, seasonal migration had Aman Transplanting July-August
Heavy rain - standing in water to work; migrants avoid bringing
declined, this was not necessarily typical
children; snakes; clashes with own transplanting at source
November-January
of the Purbi Singhbhum - east Medinipur Aman Harvest
Cold weather - risk of injury through threshing. Coincides with
harvest work at source.
stream as a whole. Overall, seasonal
February
Standing in water to work. In most source areas does not
Boro Transplanting
migration has continued to grow in the
coincide with own work at source.
1980s and 1990s. However, interviews Boro Harvest
Extremely hot - hard manual work during day In open fields in
April-May
temperatures of 40c and above. Risk of injury through threshing.
with employers in five destination locali
Snakes. Work very intensive because of employers’ need to
ties suggest that the source areas and
avoid rain. Employers tense and employ more workers for
recruitment mechanisms used by employ
shorter period than in aman harvest. Often leads to fever or
ers have changed over time. This change
diarrhoea as well as exhaustion among labourers. Does not
coincide with own work at source but remittances may be useful
also reflects supply factors with, for ex
for beginning of own monsoon transplanting.
ample, the development of intensive irri
gated agriculture in west Medinipur [see
Rogaly et al 2000J.
Table 3: Characteristics of the Study Localities and Sub-Regions
■
Migrant Workers and Their
Employers?
I.
Each locality was selected to be illus
trative of a particular sub-region. Table 3
summarises broad differences between
each of the sub-regions in terms of agro
ecology, economic structure, and social
identities. The migration survey showed
patterns of seasonal migration to vary across
the study localities.
Table 4 shows the number of people
migrating out of and into the five localities
during 1999-2000. Migration from the
localities in Murshidabad and Purbi
Singhbhum districts was highly gendered.
Only men migrated out from these locali
ties. Neither men nor women were pre
dominant in the mixed sex migration from
other localities. Outmigration was also
patterned by season. For example, people
from Purbi Singhbhum migrated only in
the harvest seasons, while migration from
the other study localities was common in
all four main seasons.
We can see immediately that, contrary
to popular stereotype, men travel for work
in transplanting in large numbers. More
over, contrary to what might have been
expected, adult migrants take children with
them in all seasons despite extremes of rain
and mud at aman transplanting, cold at
aman harvest, and heat at boro harvest.
The men, women and children migrating
Study Localities/
Sub-Region
Source or
Dstnation
Agro-Ecology
Agrarian Structure
Group Self-Identifications
Bagri
(Murshidabad)
Source
(West Bengal)
Men
Irrigated/
multiple
cropped/
Flood prone/
Flat
Very very high pop
density; exorbitant
land prices;
Local wage labour
market; little tenancy;
very high inequality in
locality
Muslim majority vs
Bengali; bagri vs rarh
sub-regional identity
(based on difference
between east and west
sides of Bhagirathi)
t
!
■
i
Ilf
Santhal Pargana Source
Rainfed/
Almost half pop Santhal
Low pop density;
(Dumka)
(Jharkhand)
Single cropped/ No land or wage labour (adivasi/tribe) - own
migrants women, undulating
language; strong ethnic
market; much tenancy;
men and children hills
identity vs others ‘diku’.
Relatively less
Strong sub-regional
inequality in locality
identity both as part of
Jharkhand (vs Bengal)
and within it
Purulia/Bankura
Source (West
Bengal) migrants
women, men
and children
Low pop density; land
and wage labour
markets;
Less tenancy than
Jharkhand localities;
High inequality in
locality
Many diverse caste and
tribal identities Including
Mahato, Baud, Santhal
(many vs ‘Bengali' caste
Hindus). Strong sub
regional identity- .
‘unwanted child’ ky
Purbi Singhbhum Source
Rainfed/single
(Jharkhand)
cropped plus/
migrants from
undulating
locality men only
but Santhal women
also migrate in
this stream
Low pop density; land
and wage labour
market; much tenancy;
high inequality in
locality
Many, diverse caste and
tribal identities Including
Mai, Mahato, Santhal,
Sabar. Lack of strong .
regional identity despite
separate dialect.
Barddhaman'rice bowl’
High pop density; major
wage labour market;
active land market; little
tenancy: high inequality
in locality
Central to Bengal culturally and
geographically
Strong religious (Hindu/
Mus) and caste identities
Rainfed/singlecropped plus/
drought prone/
undulating
Destination (West Irrigated/
Bengal) receives double
migrant meh,
cropped/
women and
Rat
children from all
above streams
. andothers . ~
:r,U
I
i
I
I
' I
I
■1
Economic and Political Weekly
December 8, 2001
4549
1
• •*-
J1 :
1 11 .
■
AV,
•-
?
I
It'
‘•■1.
gangs leaving the source area study locali
ties were made up of people of different
‘jati’ (as seen in relation to others in the
source locality) but to differing degrees
depending on the source locality. All
Muslim and all-Santhal groups were com
monly formed in the Murshidabad and
Dumka study localities respectively, though
there were exceptions.
Employers may be tenant farmers or
owner-occupiers. Most of the sampled
employers in the studied destination area
locality were own6r-occupiers. Although
all were Hindu, they were far from homog
enous in terms of wealth, caste, education
and occupational mix. The study locality
included two scheduled castes, Dorns and
Bagdis, high caste Mahanto Bamun, and
the peasant caste Aguris.9 Employers of
migrants included a local Bagdi labourer
who cultivated a small area of land (1
bigha, or 0.3 acres, HH 55) and could
borrow migrant workers recruited by his
own regular employer. At the other end
were employers, all Aguris, whose sons
had studied to degree level, who received
regular salaried income, and who con
trolled the trade in rice and fertiliser as well
as relatively large areas of land (36 bighas,
or 12 acres, HH 5).
In Barddhaman District as a whole, there
is also a sizeable group of Muslim culti
vators who employ migrant labourers.10
Table 7 gives the breakdown by religion
of employers.hiring gangs of workers from
the study localities in 1999 2000. There
was no one-to-one correspondence between
employer’s religion and the social compo
sition of the migrant gang they hired.
However, a larger proportion of gangs
from the predominantly Muslim
Murshidabad (Bagri) locality found work
with Muslim employers. Our findings
suggest this pattern also holds for the
Murshidabad-Barddhaman stream.11
Gangs of Migrants and Their
Recruitment
In studies of contemporary seasonal
migration in other regions of India, middle
men or brokers have been found to be
important in the fixing of wages, working
conditions and living accommodation.12
In the study region there is no cadre of
brokers involved in the recruitment of
migrant rice workers. In most of the stud
ied source localities, individual men and
women act as gang leaders or ‘sardars’.
The rote of a ‘sardar’ varies, however,
from being a person with initiative and
4550
Map of the Study Region
A
ti
l^Deihi
'
r
■-'i
f
la
BAY OP
BENGAL
BANGLADESH
□UMKA
• (\
4-'
Rampurhal •
BIRBHUM
r
yl ^jiyaqHni
Zmurshidabad
?
|
/
'l
JHARKHAND
NADIA
PUHULIYA
•Ranclu
BANKUHA
(
c
\
( 7
(
)
BANKURA
HUGLI
r>
I Jainshenpur A
\
[/
I
WEST BENGAL
\
i
PARGANA -
r—•
Z
PURBI
^INGHBHUM
A. iK0LKA,A
J f
haor
\
fl <
24 parganas
MEDINIPIJR
MAVURBHAN.I
I
ORISSA
/
I
BALESHWAR
(S)
a
BAY OF BENGAL
studied migration streams
0
km
mo I
Barddhaman district destination areas
confidence who builds up a group and regular sardars. Groups form among rela
negotiates with employers.at labour mar tives, friends and acquaintances, often
ket places, to an established
c«-------- contact and involving men from other nearby vil- >
13 Ganes
lages.13
Gangs are more like bands with
semi-permanent recruit of" a destination inopc
a
median
of five or six members (depend
who
travels
employer
regularly
area ( t ,
betwee their own village and the* em ing on the season), much smaller than the
ployer, bringing migrants according to gangs which travel out from the other
prior specifications. This variation in roles study , localities (Table 8). The larger .
exists within as well as between locali*-.
locali- , number of gangs from the Murshidabad
locality is
is due
to the
the common
ties, for example in the Puruliya study study
study locality
due to
common pracprac- ‘*
locality
tice of migrating to work for two or more
W
Exceptionally,
Exceptionally, in Jalpara, the employers in the same season [Rogaly and H
Murshidabad study locality, thete are no Rafique 2001].
»Economic and Political Weekly
December 8, 2001
1
¥
!
* •
ii .v s . ...
redistribution of land held over the ceil
From our observation’s at Katwa railway 24 Parganas and Murshidabad, and paid ing, and panchayati raj.1-8 .
.
station, an important labour market place, at a piece rate.
This was a period of consolidation for
In sum, seasonal migration is of critical
a smaller gang size appears to be typical
the
smallholdercultivators, many of whom
of the Bagri (Murshidabad)-Barddhaman importance to the material livelihoods of were actively involved in the political
stream as a whole. Table 8 also indicates many hundreds of thousands of people in parties in the Left Front coalition, particu
a relatively large median gang size (rang the study region. It involves men, women larly in the Communist Party of India
ing from 14 to 50 depending on the season) and children, mainly scheduled castes, (Marxist). The CPI(M)’s strategy became
from the Puruliya locality, where two scheduled tribes and Muslims. Employers, more focused on electoral success, through
regular institutionalised sardars worked Hindu and Muslim, are smallholders with seeking to maintain support across classes
for the same particular large-scale employ different degrees of wealth and varied of smallholders as well as agricultural wage
labour requirements. While there is some
ers each season.14
workers, and keeping the peace in the
Employers in the destination study lo cooperation in recruitment, most of this is countryside [Webster 1990,1999; Rogaly.
cality hired gangs of a wide range of sizes part of patron-client relationships between 1998b]. Smallholders with profits from
(four to sixty) and for between one and relatively large and relatively small-scale agriculture began to invest in groundwater
thirty days (Table 9). Gangs of labourers employers. Most recruitment is organised irrigation, which was more attractive in the
were recruited directly by employers or by employers acting as individuals rather less conflictive rural environment, and
through employers’ agents ( gomostha ). than in collusive groups. In Section II, we seemed likely to produce further profits,
Recruitment could be from a labour mar move on to consider the causes and con given the efforts made by the regime to
ketplace or by travelling in person to known sequences of seasonal migration for rice make access to credit and other agricul
villages in a source area. Sometimes work.
tural inputs much easier and their supply
employment arrangements were initiated
more timely. Rapid expansion of the new
III
by labour gangs or leaders of gangs (sardars) ,
boro crop and adoption of high yielding
Causes and Consequences
coming to seek work unsolicited. Evidence
varieties and associated cultivation tech
of Seasonal Migration
gathered through travelling with migrant
nologies began in the early 1980s and has
workers, observations of labour market
continued throughout the last two de
places and interviews in source areas Growing Demand for Migrant
cades.19
showed that in Barddhaman district as a Workers
One outcome of these changes was that
whole, it was common practice for em
more
manual workers were required in
The history of settlement and the begin
ployers to seek one season’s gang to commit
Barddhaman district. Although irrigation,
nings
rice
cultivation
in
the
area
now
of
to work in the following season.
Barddhaman district go back ploughing and threshing were increasingly
In the destination study locality at least known as 1----------mechanised, the transplanting and harvest
one-third of employers were supplied with over millenia [Eaton 1994]. Populations ing of rice continued to be carried out by
in
the
study
region
as
a
whole,
which
migrants by other employers. For three
hand- Ideologically, manual work was seen
seasons of the year, cultivators of smaller includes parts of present-day Jharkhand by many owners of land as. something to
amounts of land were able to borrow and West Bengal states (see Map), have be avoided by both men and women."®
migrant workers, who had been recruited never been static, and seasonal migration ; ' ‘ ■' j employers’ aspirations for
locality
and were being accommodated in the cow for work in the rice fields of south-central Study
Bengal has a history of at least 100 years.16 thei/ children and even for themselves
sheds and other outbuildings of largerTaken together, our evidence17 suggests involved spending increasing amounts of
scale cultivators. Because of a correspon
a
rapid
increase in the number of manual time out of agriculture, in urban white
dence between landholding size and jati,
collar occupations. This ‘babu’ approach
there was a jati-pattem to cross-recruit workers migrating seasonally for rice work to agriculture came to be more affordable
ment. In general. Mohanto Bamuns and in the last two decades to have taken place as profits were made from rice production
Bagdi employers required less migrants in the context of a-series of political, and wealth was accumulated. It was scoffed
than Aguris and cross-recruitment was by economic and cultural changes in the at by those other landowners who contin
Mohanto Bamuns or Bagdis from Aguri destination area.
Prior to the . coming on stream of the ued to see themselves as ‘chasis’, peasant
employers. Money did not change hands
Damodar
Valley Corporation canal system cultivators, for whom doing their own
between employers for this, though it
cultivation work alongside other family
formed a part of ongoing patron-client in the 1960s, there was little assured members and hired labourers was part of
relations. The smaller-scale employer paid irrigation in Barddhaman district. The their self-identity. However, even this group
the migrants’ wages for their period of coming of canal irrigation and the slow required manual workers in the peak sea
expansion of the entirely irrigated boro
hiring.
sons. At the same time some former local
In the season for transplanting aman paddy crop from the early 1970s took labourers aspired to withdrawal from this
rice, however, there was cooperation in place at a time of violent conflict in the kind of employment.21 Another reason
recruitment between relatively large-scale countryside, including a period of land why demand for migrant workers increased
employers, who would club together to seizures during the 1967-69 period and was the reduction in hours of work per day
cover the costs.15 Whereas for the other heavy repression in the eight years fol for local labourers. The Krishak Sabha, a
three seasons most gangs hired were from lowing. The Lert Front government, CPI(M) mass organisation made up of
the Puruliya/Bankura stream to the west which came to power in 1977, imple smallholder cultivators and wage workers,
and were paid at a time rate, for the aman mented agrarian reforms energetically in supported this direction of change in
transplanting, most of the gangs were hired its early years, in particular the registra Barddhaman.
• <
from the east, from the districts of north tion of share-croppers, the continuing
".s’
I
5
9
-
•
j
- - -J
/'••• - •
Economic and Political Weekly
>-
December 8, 2001
4551
i-l
f
i1
4
jg
I
•'ll
■7-
i L
Role of the Krishak Sabha
Other changes that the Krishak Sabha
has presided over during this period (and
which have been associated with increased
demand for labour in agriculture) have
included the closing of the gap between
male and female wages and ensuring steady
increases in the cash portion of the pre
scribed wage rate, so that real wages did
not decline. Not surprisingly, given its
cross-class composition, the Krishak Sabha
did not represent agricultural workers as
a class-for-itself.22 Indeed agricultural
labourers were not allowed to form a
separate union of their own. The changes
in wages and working conditions reflected
a carefully managed balance of power, in
which smallholder-farmerswho employed
labour were dominant. The positive
developments which were felt by wage
workers were necessary to the overall
strategy of enabling smallholders to pur
sue profit in a non-conflictive countryside
and of keeping cross-class support for the
CPI(M) in elections.
The Krishak Sabha agreed to continue
allowing migrants from outside
Barddhaman district to be employed there.
This reflected employers’ interests in two
ways. It greatly relieved their seasonal
labour shortages, and it also exerted indi
rect pressure on local labourers to comply
with working conditions or be replaced.23
The Krishak- Sabha decided in the early
1980s to compensate local labourers in
two ways: migrants would be allowed to
work only in peak seasons, and, most
importantly, they could not be employed
at lower levels of earnings per day than
local labourers.24
Accumulation of Wealth and
Growing Inequality
i.!,:
Enabled by the Krishak Sabha to bring
in migrant workers to cover the periods of
intense labour shortages, many larger-scale
smallholders accumulated further wealth.
Those in the strongest economic positions
had diversified beyond agriculture into
trade (in fertilisers and rice), processing
(of rice), salaried employment and party
politics. In Moraipur, the destination area
study locality, during the 1990s, better-off
agricultural employers were able to
build large ‘dalans’ (brick built houses)
and to purchase consumer durables,,
including dining tables and chairs,
televisions, satellite dishes and cars.25
The gini coefficient of inequality in the
distribution of assets in the study year side the vertical unequal mutuality be
was 0.58. In the district as a whole tween employers manifest in cross-recruit
inequality among smallholders and ment in the study locality, as well as the
between smallholders and labourers has cooperation over recruitment of piece rate
visibly increased26 during the 1980s and workers at aman transplanting. As men
tioned, employers here did their own re
1990s.
’ As we have seen, the Krishak Sabha cruitment (or used their own permanent
oversaw this process of accumulation in labourers) in migrants’ villages of origin
Barddhaman district, through making or at labour marketplaces. There was no
deals between classes with potentially cadre vof middlemen in the market for
conflicting interests. Accumulation has migrant rice workers.
Recruitment of labour-was not the only
increased the demand for migrant workers
(a cause of migration), while also being factor which made for the insecurity of
built on the back of migrant workers destination area employers. Profits de(a consequence of migration). The success pended on the costs of other inputs, includof the CPI(M)’s strategy can be seen in
Table 6: Social Composition of Migrant
the continued cross-class support it won
Gangs from Source Area Study
in the 2001 West Bengal state assembly
Localities
elections.
Single Jati Mixed Jati
Source Locality
Competition and Insecurity among
Employers
Although employers argued common
class positions within the Krishak Sabha,
we have described class differences be
tween employers [Rogaly et al 2000] and
found further evidence of intense rivalry
and competition among agricultural employers of the same class. This ran along-
Gang
Gang
61
39
6
1
8
7,
6
6
Jalpara, Murshidabad
Alopahari, Dumka
Bajnagarh, Puruliya
Upartola, Purbi Singhbhum
Notes: 1 Jati is used here to refer to the social group
(based on caste, religion, ethnicity), which
an individual belongs to in their home
locality as seen by others there.
2 In Jalpara, the social composition of two
gangs was not recorded.
Source: Migration survey.
Table 4: Number of Seasonal Migrants Hiring Out and Number Hired in for
Rice Work by Season (1999-2000)
Study Locality
Aman Transplanting
Aman Harvest
Boro Transplanting
Boro Harvest
Jalpara, Murshidabad
Alopahari, Dumka
54 men
130 mixed sex of
which 18 children
119 men
67 men
61 men, 55 women, 71 men, 78 women.
20 children
29 children
Bajnagarh, Puruliya
127 mixed sex of
which 26 children
35 men, 28 women, 16 men, 25 women,
and 7 children
and 11 children
0
Upartola, Purbi
Singhbhum
Moraipur, Barddhaman ?
18 men
0
143 men
49 men. 72
women and
27 children
21 men, 22
women, and
5 children
53 men
535 inmigrants +
children
500 inmigrants +
children
639 in-migrants +
childrr
I
Notes: 1 The number of women and men in-the mixed groups from Dumka district were not counted
separately in the aman transplanting season, nor in the destination area study locality.
2 The numbers of migrants for aman transplanting work from Bajnagarh, Puruliya, is inflated
£
because it is based on total gang size, rather than gang members from the study locality. .
Source: Migration survey.
1
Table 5: Sex Composition of Migrant Gangs and Number Travelling with Children or
Infants by Source Locality
Source Locality
Jalpara, Murshidabad
Alopahari, Dumka
Bajnagarh, Puruliya
Upartola, Purbi Singhbhum
Men-only
Gangs
Women-Only
Gangs
Mixed Sex
Gangs
61
2
1
7
0
0
0
0
0
28
,14
0
Gangs with
Gangs without
Children/lnfants Children/lnfants
~061
'
23 •
9
0
7
6
7
Notes: 1 Data on the sex and age composition of gangs was collected in three seasons only. Hence thejE
total number of gangs is less than might have been expected from previous tables. ®
2 Gangs travelling with children all included both men and women.
Source: Migration survey.
'
December 8, 2001 j
Economic and Political Weekly
4552
■...............................................................................................................................................................................................................................................
-
............................................................................................................
I
-/.Xv'Vv i.
< 1I 1I
■' ft
■-I
W
■
■
ing fertiliser as well'1 as diesel and electricelectric they were likely to get consecutive days
ityforgroundwaterimgation.andtheprice
for groundwater irrigation, and the price of work at the agreed rate. This was not
ity
theircropcouldfetchinthemarket.During possible at home.
■ t
Migrant workers also knew that they
the fieldwork year, it became clear that
___________
having would almost always be paid for the work
central government policies were
effect on the .prosperity of agricultural they did because non-payment would be
ant---smallholder producers. Not only were against the interests of the Krishak Sabha
subsidies on inputs being reduced, the as it was likely to lower the reputation of
prices which mills and their agents were the Krishak Sabha itself in terms of its
prepared to pay for their crops had dras- capacity/ to manage the countryside. It
tically declined. The classic method of would
v : ,J also work against the reputation of
making money on rice for relatively well- the locality. Several incidents in which the
to-do cultivators was to store it and sell ‘party’ (whether in the form of elected
it pre-harvest. Mills were now increasingly representatives in the panchayats, party
cadres or
or Krishak
Krishak Sabha
Sabha members)
members) was
was
supplied with unhusked rice by producers cadres
in Bihar and Uttar Pradesh. The price of called on to enforce proper payment were
rice was further reduced by purchases from recalled by migrants interviewed in the
Bangladesh, Thailand and elsewhere, source localities.31 These incidents were
enabled by the liberalisation of imports.27 especially striking for migrants from
It was admitted by Madan Ghosh, the Jharkhand state,32 to whom organised
secretary of the CPRM) in Barddhaman institutional wage protection, even at the
district,28 that it was conceivable that wages local level was unknown.
might have to fall and the rise eventually
agreed in November 2000 was less than Labourers’ Room for Manoeuvre in
had been anticipated, also due to flood.29 Negotiations
Earning Potential as a Cause of
Increasing Seasonal Migration
For people in the source areas surrounding Barddhaman district, the increases in
wage rates and i n the number of day s ’ work
available at the destination each season
motives for
have been very important
i ,
migration. The length of the migration
seasons was approximately 20 days (see
Appendix). During the study year the wage
the Krishak
--------- Sabha
------- in
prescribed by t.._
Barddhaman district for all agricultural
labourers was T1 rupees per day and 2 kg
was equivalent to 4-4.5 kg of
rice.30 This
'__ ____
hulled rice, which in real terms showed no
changefrom 1991-92 [Rogaly 1996:147].
Wages varied across the source area study
localities and seasonal fluctuations were
also more pronounced there. A peak sea
son wage with a total value of Rs 20 per
day (equivalent to approximately 2 kg of
hulled rice) was not untypical.
There were upward variations from the
prescribed wage in what was actually
received by migrant workers; and most
migrants in the Murshidabad-Barddhaman
stream were paid their kind portion in
cooked food rather than rice. As we have
seen in.the previous section some migrants
were paid at a piece rate, which usually
meant higher daily earnings than time rate
arrangements. Most importantly, in the
four main seasons, workers knew from
previous experience or from others that
I
■ s
•'wI
I
!'lI
JlH
"’ . W
I■
*
ft
ft
I
-s
1
Economic and Political Weekly
-----
The main proximate cause of seasonal
migration from the source areas for those
who migrated was therefore that they
expected to gain a lump sum of cash to
take home which would nor be available
to them in the source area and to eat for
the period of work. However, the specific
dynamics of seasonal migration in the study
region meant that labourers regularly found
room for manoeuvre in negotiations.
Here lies another part of the explanation
for the growth in the numbers of seasonal
migrants. Not all migrants in the study
regioni were compelled by economic circumstances to migrate each time they went.
Lack of collusion by employers contributed to migration even by people who
could get by without it. Migrants knew that
in the season they were very likely to find
work and to get paid in Barddhaman.
Negotiation was possible over wage rates,
number of days work, distance of fields
from —
accommodation, meal quality and
f.—timing.,33 Of course, once a gang of
migrants arrived at the work and living
place for the season, the employer might
renege on parts of this agreement. Employ
ers’ relatively higher material security
meant their capacity to withhold work from
a particular group gave them power in the
market unmatched by migrants’ capacity
to exit. However, this did not leave mi
grants without power altogether. Some
tested out whether the employer seemed
to be true to his word and a few gangs
December 8, 2001
••
• • •
returned to the labour marketplace surrep
titiously (if they found the reality did not
match the rhetoric and they could afford
to), leaving an employer who had invested
time and money in recruitment and more
importantly in their rice crop, with major
problems.
Divergent Trajectories
The underlying causes and the conse
quences of seasonal migration varied across
the four migration streams, as illustrated
through analysis of data from the source
area study localities. These underlying
causes reflected specific agro-ecological
conditions, economic structures and wider
social relations, as well as associated
contrasts in the potential for change in
each of these.
Murshidabad-Bagri Locality: Itinerant
petty trade has long been an important
occupation for poor residents of the
Murshidabad locality, Jalpara, but agricul
ture remains central. The sub-region has
seen a dramatic agricultural revolution in
the 1990s on the scale of that which took
place in Barddhaman district. A high water
table meant that irrigation water was readily
available beneath the surface. Those fami
lies who were able to invest early in groundTable 7: Religion of Employers of Gangs
of Migrant Workers from the Source
Area Study Localities
(Number of Gangs Each Season)
Source Locality and Season
Alopahari, Dumka
Aman harvest 1999
Boro transplanting 2000
Boro harvest 2000
Jalpara, Murshidabad
Aman harvest 1999
Boro transplanting 2000
Boro harvest 2000
Bajnagarh, Puruliya
Aman harvest 1999
Boro transplanting 2000
Boro harvest 2000
Upartola, Purbi Singhbhum
Aman harvest 1999
Boro transplanting 2000
Boro harvest 2000
Hindu
Muslim
8
7
6
2
2
3
4
4
17
18
6
11
5
4
3
o
2
o
4
0
1
1
Notes: 1 This information was not recorded in the
aman transplanting season. Employer’s
religion was also not recorded in the case
of a few gangs in the three seasons
reported here.
2 In the boro harvest there is a shift for the
Murshidabad gangs from predominantly
hiring out to Muslims to predominantly
hiring but to Hindus. This is associated
with migration to Nadia district in the boro
harvest. Six gangs reported migrating to
Nadia at the boro harvest and all but one
gang worked for Hindu employers.
Source: Migration survey.
4553
1
'ifit
livelihoods. Rather than inequality (gmi
was 0.54 in the study year) accelerating
to exclude poorer people from being able
able to dictate terms to crop cultivators to investThiand (as in the Murshidabad
within the command area. Land prices locality), migrants have invested their
have rocketed
with the
new productivity
productivity
iCKeieu wiu.
M.w ..w..
emittances into renting and cultivating
r available
land
per
has
landastenanls. Migration for some has led C°Seasona:i migration in this locality was
and as
*.
r-- cap.ta
creased further, caused by the very rapid the way out of being compelled to migrate.
clearly associated with a process of pa ia
' 1 for rice work is on the decline, delinking from dependence on the patron
Mi,..although there has been a small increase age of rajas; Wages earned through migra
Land prices in the study locality were in migration to work in brick kilns.
tion enabled conspicuous consumption
Rs 1 50,000 per acre in the study year,
In the Purbi Singhbhum study locali y, which in contrast to the employers ot
000 in the migration has also played a part tn changm .uu
to ks 1i .75
compared m
Barddhaman district meant a more secure
Barddhaman district study locality, and
social relations of jati. Different Re roof new cooking utensils or clothing.
from Rs 5,000 to Rs 60,000 (depending mg
manual work have meant that Migrants were more assertive jn their
on quality) in the other source
local.- ologm
‘3 Benias have not had access
ties Inequality has increased sharply and
g
wl— —
■smuchhigherthanintheothersourcearea
o the
the migration
m.gra
iS —ble M.l
I
.1
i
i
w
■
i
1997 and 1999 because of three consecu
tive years of drought. For those who had
previously migrated to invest in their own
land, these years meant continuing o
migrate but for different reasons - for
survival. Agriculture had failed almost
t
those dispensing development grams^rrom
the Block Development Office. Migra
has been important in enabling people to
make more choices about their economic
1 gg9
Boro transpianting 2000
0
0
13
2-14
2-30
2-30
ijn
1-30
16
15
Source: Migration survey.
Economic and Political Weekly
4554
fl
“tX" S de
study localities. Using data on. the rupees migration have ^..abled
enabled some Mai
*
a women
ed b Rogaly and Coppard (2001),
equivalent of a basket of assets including t0 become increasingly bold in taunting
taunUng
J
of ch
land, we estimated a Gmi coefficient
_ - been reduced too
the Benias who have
in individual cases. Many migrants
reliance on symbolic untouchability
untouchabi y to
migrants Continue to depend
0.69 in the study year.
Seasonal outmigration from the locality
.
“"as forEmployment and fori
maintain their sense of
of supenonty.
superiority.
offers little opportunity for syuctural
Puruliya-Bankura Locality: No such
Dumka (Santal Parganas) Locality ■
change. Men are compelled to find work irrigation development was evident in
There were no dramatic agro-ecologca •
outside the district because there is insuf Bajnagarh (Puruliya), also on the
changes in Alopahari (Dumka) either. Here
ficient employment available locally, even Chhotanagpur plateau but further no .
there was less evidence of change in
m
though, for almost all crop tasks women West Bengal. Here, boro production has economic structures, than in the Puruliya
donothireoutlabour.Thisgenderdivision
locality.noevidenceofawtopre—
of paid work is part of the cause of the
high propensity to migrate among poor
Si^hbhum^ahtyTbut at the same time
men - men in the household are. respon w pump water from the tanks they con no major regressive change in terms of
sible for oringing in income to cover the trolled. Though the Rajas were major local — major regressive change
inequality.
•• , r
Economic
' * differentiation
J: e
costs «f.all family members. In the study employers in the aman transplanting and
among tneounuia.a
----- ,
Jocality, men tended to give their rem.f harvesting seasons, there was too htt e In the study year, the gmi coefficient ot
tances to their wives or mothers and these work to go around. Moreover, people were
wealth inequality was 0.33.
were spent mainly on debt repayment and reported to have migrated in greater num
Relations between Santhals in Alopahari.
bers (both for rice work and to work in on the one hand, and their landlord and
on food.
. .
In Rogaly and Rafique (op cit), we brick kilns) during the period between
explore how dependent women in nuclear
Gang Size by Study Locality and Season
families, left behind for a season by ma e
Table 8: Median
Boro Harvest
migrants, can be reliant on their own kin
Aman Harvest
Boro
Aman
Variable
Transplanting
in Jalpara. This may involve compromises Locality
Transplanting
in behaviour, including decisions about
9
11
16
Gangs
19
- - keep close kin available A|Opahari Dumka
reproduction, to
12
4
4
Median size
4
7
for physical security, loans, child care and
7
24
Gangs
50
. • __
Bajnagarh Puruliya
14
24
29
Median size
10
her support.
suppvit.
.
other
22
10
6
6
Purbi Singhbhum
:
. JalparaMurshidabad Gangs
5
5
5
Median size
0
to the Murshidabad-Bagn locality,
■2
0
9
Gangs
9
Upartola, the Purbi Singhbhum locality, is Upartola P Sing
Median size
in an area of low population density.
However, it too has seen change
<
w in agri- Source: Migration Survey.
and Days Worked Range and Median by Season
culture, with action seen over the last decade
Table 9: Gang Size
(Destination Area)
to dam the nearby river and lift w^eT^or
---------------------------Si^e
No of Days
No of Days No DaVs
irrigation of boro rice productton. This has-----Gang Size
Gang Size
Ranae
Median
Range
come out of individual initiative and Season
Range
Med'an__________
2^25
----------------- --------------- id
' 2-25
15
contacts between the family of a long-----------2’60
0
2-14
7
4- 15
5- 35
4-27
a
December 8, 2001
I
<
>
|
I
.1
I
' i'
;5tm
■'W
I
' 1
I
■'i
■•I
■ftl
I
*B
.;s|
-
■
•->
I
through direct involvement in recruitment individually. This was part of a general
other creditors outside the locality, on the
from source areas. Employers from the picture of competition among employers,
other, remained highly unequal, however.
study locality in Barddhaman district whose religion and caste backgrounds
Migration for manual work, not only to
varied. Partly as a result of these differen-.
Barddhaman but also to the north-eastern expressed their fears about having to stay ces, labourers migrating into Barddhaman
states and further afield on ministry of in source area villages. Their sense of district had more room for manoeuver to
defence contracts, has continued to be an coming from the central place, the place make ‘small choices’ in the precise details
important part of livelihoods. The local of modem agriculture and their urban white agreed with employers at home or in the
labour market is very small and most of collar aspirations, were, strengthened by labour market places. However, employ
the Santhals and others who migrated the experience. Santhals and Muslims who ers’ anxieties stemmed not only from labour
seasonally cultivated land as tenants for a had migrated from the study localities in shortages but also from the declining
large-scale ‘dikU’ (non-Santhal, in this case Bagri/ Murshidabad and Santhal Parganas profitability of rice production.
brahmin) cultivator who lived outside the (Dumka) saw themselves more than pre
The sub-regions feeding the West Benviously as members of larger groups, an
locality.
gal
‘rice bowl’ have varied agrarian struc
Outmigration was not a major cause of expected outcome when moving from the tures, as reflected in the wide range of
known
local
area
to
a
broader
playing
structural change in the Dumka study
inequality coefficients found in the study
locality, in contrast to the effect of the field. However, although Santhals met localities. There were.clear divergences
Jharkhand political movement in the 1970s, Santhals from other sub-regions, and at between study localities in patterns of
which saw interest rates on diku loans least three marriages were reported to have structural change resulting from migration
considerably reduced. However, a few happened this way, and Bagri Muslims ot in the study localities. Migration in the
individual migrants were able to save different jati came away with a heightened Puruliya locality enabled wage workers to
enough to move out of migration here. sense of their religious and territorial bring pressure to bear on the sensibilities
Land was commonly mortgaged for cash identities, these two groups were not found of local employers, in the Purbi Singhbhum
in Alopahari. Because of special legisla to spend time together at the destination. locality there was general movement out
tion designated to protect Santhals’ and Such non-interactive encounters helped to of migration, whereas economic compul-,
harden self-identification as Santhals and
other ‘adivasis’ land from passing into the
as Muslims. However, self-identification sion continued for different reasons for
hands of dikus, it could not be sold.
migrants from the Dumka (Santhal
Remittances would be used to unmortgage did not just vary across individuals and Parganas) and Murshidabad Bagri
land, and if production resumed, it was over the life-course, different aspects of localities.
possible for some people to sustain them a persons’ group identity became promi
The stories of individual migrants and
selves without migration. However, live nent at particular points in space.
their households varied too - some used
lihoods were highly precarious and when
remittances to invest and eventually thereby
Summing Up
crisis hit, often through expenses associ
to exit migration, others continued to have
ated with ill-health, it was likely that
no choice in the decision. Decisions were
This
section
has
reported
our
analysis
of
accumulated assets would have to be liq
embedded in particular structures of intra
uidated and migration again become nec why seasonal migration for rice work has household and broader kinship relations,
increased over the last 25 years in the Jhar
essary.
khand-West Bengal region. Characterising which migration could in turn challenge.
the region as a whole, we have also high Even in the localities from which family
Changing Social Identities
lighted important contrasts between sea migration was common (Puruliya/Bankura
and Dumka), kin were important in arrang
Table 3 summarises some of the main sonal migration for rice work there and ing for care of the house, livestock and
group self-identifications in the study sub seasonal migration elsewhere in India. dependents. For younger people earlier
regions. Group self-identifications are Further, looking within the region, we have decisions to migrate might be taken with
based on combinations of ethnicity, caste, illustrated the divergent trajectories of four parents’ permission, urged by parents (rare)
religion, nation and territorial space as major migration streams.
The methodological approach of this or against their wishes. In some cases
well as class. In the process of travelling
migration against elder’s wishes had the
across the countryside, spending time at two-layer comparative study has enabled effect of changing the balance of power
us
to
consequences
of
examine
causes
and
bus or rail stations, living and working in
in the household [Rogaly et al 2000].
the destination area, migrant workers’ seasonal migration as interacting- and
The findings suggest an overall pattern
social identities changed. These changes mutually determining. Social and political in which the element of choice for mi
were complicated because they included conditions as well as technological ones grants about who they work for and where
both instrumental use of identities in labour were necessary to the take off in West they go to work has increased. Many
market place negotiations, and changing Bengal agriculture. Migrants found em migrants are able to use the lump sum of
self-identifications through interactions ployment in larger numbers. They were in remittances they return with for purposes
with other migrant workers, with destina turn essential to the continued agricultural beyond loan repayment and food consump
tion area employers, and with others, such growth at the heart of south-central West tion. However, the process is still driven
as other bus passengers, drivers and con Bengal’s prosperity, which fed back into by an economic compulsion and the choices
ductors and stallholders selling snacks, tea increasing demand for workers.'4
Instead of using middlemen for recruit remain small. This is why changing social
and trinkets.
relations, including stronger group iden
Moreover, it was not only migrant ment as is common practice elsewhere in tities and opportunities to taunt and un
workers whose social identities changed, India, employers of rice workers in West settle employers, to get the upper hand
but also employers of migrants, in part Bengal’s Barddhaman district recruited
Economic and Political Weekly
December 8, 2001
4555
)
-4
I I
and pregnant and post-natal migrant women
I I
I
more demeaning experience. At the same Child Development Scheme-funded
time, such divisions among migrant work ‘anganwadis’. In reality, anganwadis across
ers and between migrants and local
the study region (source and destination
workers, play to employers’ interests in areas) were not functioning properly for
preventing a unified rural working class local women and children. The woman in
movement.
charge in the destination study locality
Migrant rice workers in West Bengal do
expressed great surprise that we even
IV
have many common interests: forexample, asked whether migrant women could at
A Political Approach to
a place to gather peacefully to negotiate tend. The idea seemed absurd to her.
contracts, rather than being lathi-charged
Migrants’ Rights
Because of this, such measures success
f
□ as Muslim men migrants gathered on the fully won, would be of benefit to both
___________
•
’ we ,found railway lines at Katwa regularly were m
In
one source area (Dumka),
migrants and non-migrants and therefore
migrants
to have been excluded from houseA successfui campaign for
Illiyil CU1UO UV7 Iiv* » w -----------------------.
------------fnr
nnor
investment in a labour shelter more widely acceptable than measures
grants
building
earmarked
forfamipoor fami
lies the Indira Awas Yojana, because away from the railway station could lead dedicated to migrant workers alone. A bus
stand health camp at Bankura is a rare
they were deemed likely to be absent dur
to a wider claim for rights to information. example of how health provision for re
ing the stipulated period for building. In
The state could be the target and success turning migrant workers can be organised
questions"^10!" why neonatal seasonally mightmakedaytodayandseasonalchanges
by a coalition of unions, NGOs, local
migrant women did not gain access to in labour market places (number
Government bodies, shopkeepers and oth
Integrated Child Development Scheme employers
emp|oyers and workers, going wage, likely ers. Those using it included both migrants*
potential
facilities were greeted with derision.
length of season) known to potentia
and non-migrants.
misrants; it would also be of use to
From such incidents and from the high migrants;
Changing mindsets on seasonal migra- ■
...
1 lw“*
------- 3 about and/or disin- employers.
tion and persuading influential people of*
the view that migrants have been central
tcrcst in seasonal migrants in conversa
to recent growth in prosperity in destina
tions with officials across 15 districts, we
among other things, an extension of the
tion areas will not be easy. However, we
can see that migrant rice workers are not
number of state-owned buses on West
hope this research makes a contribution to
willingly accommodated by the bureauBengal’s roads and improvements to those
that agenda. One result might be for areacratic machinery. This approach to mibased projects which have a reflex reaction
in making the experience less undignified
for migrants. Such group-identification,
combined with small amounts of power in
negotiation and at the workplace can be
important to migrants’ well-being (ibid).
I
•
t-
a
1 £
••
....... ...........
I'*!'
■----- to de-
velop the opposite tendencies. It might
health when
not just beca
to rLE«j - -------treatment, but also (and especia y)
cause of loss of wages. At the very eas ,
er term_ howeVer, those havfree health care might be campaigned for,
B
more control over
have broader
broader choices.
choices,
to south-central West “ckpayisalongwayoffforcasualworkers
sickpayisalongwayoffforcasualwor er
,f (hey have
What is specific
anywhere.
.
Bengal is that migrants have been pre- anywhere
children at This requires public investment in source
children at areas to counter the unnevenness of
.... and directly under
Migrants struggle to keep cn
vented from explicitly
school, including many children
■' the wages, of
chil ren them deveiOpment and economic opportunity.
cutting
of local
workers by
local workers
by aa
selves. Social debts are made with km;
,ng“political
investment would be most effective
strong
political party and one of its affiliaffili selves. Social e ts; are: m
exams; at.
at least one
8
I- ----------? — children return to take exams
lf.
ated organisations, in order to keep cr^‘
sub.regions It might inciude tm.pport, not least at elections. This child migrated’ toj earn the
the cost of clothes
1 h
class
supr—school. However,
has been
of practical benefit to migrant to wear to school.
However the
!th schooling
h
g
.nvestment in a localily
the one
has been of practical benefit to migrant
workers, and acceptable to local workers system has not respo de to
studied in Puruliya and non-agncultural
^m^^Newcompromisesmay
rceordestination
areas.
and to employers. New compromises may requirements dictatea
py g
enterprises in Murshidabad where land is .
have to be reached, if, as seems likely, ^idte^^-—
Migrants themselves could make
Migrants’ children’s.absences
fromst‘dy
pri moreofthe remittances they returned with
there is sustained pressure on the
the profitprofit- Migrants’
chddren s
in thee Puruliya/Bankura
study
ability of rice production with lower pro- mary
sh t0 jf health and education were actually free
mary school
school in
y d;
locality could be highlighted in a push to
draw attention to primary schools’ need to
only Jo.these
for migrant workers, but
However.migranlshavenoorgamsauon change,
no"only
new inveslmenis would nol fee limchange, not
for the children of manual workers more
broadly. In the destination area study
cent drop m
"Mlarednwurromdwcnsabnietand .uvaiuj,
locality, fcne was a 40 perX'-Sy
Appendix
attendanceby
bychildren
childrenof
ofBagdis;
Bagdis - mainly
in this case caste and religious) groups, attendance
y
•
3 in
m the
th cultivation
twQ main nce crops in
Indeedthestrongersub-nationalconscious- agricultural■ •labourers
seasons.
Many
were
covering
for parents
ness developed through the migration
P
Barddhaman district as in the rest of West
earning aaareuedthatinfants
wage.
process has enabled some migrant busy
busy_earnmg
Bengal; aman, transplanted in July and
Similarly, it could be argued that infants
workers, particularly Muslims and
policies and solidarity with migrants and
their inclusion in welfare schemes when
nr? awav from home cannot be a
ttpr for bureaucratic edict alone. Such
Ii il'
E
>=
-r
“
SXET”—” i. -
gsssssx
X X.%C“Xt“-on „ -dy »ed . nug-.
Economic and Political Weekly
4556
December 8, 2001
■
I
►
,
i
I
•it-J
3
1
i
I
-
-■
August, is harvested in November and
December, and boro, transplanted in Feb
ruary, is harvested in April and May.
The net sown area of Barddhaman dis
trict is 4,73,900 ha (Barddhaman District
Statistical Handbook, 1995). In 1993-94,
4,19,000 ha were planted with rice in the
aman season and 1,58,000 ha in the boro
season (Barddhaman District Holistic
Development Plan, Society for Holistic
Approach to Planning and Evaluation,
Kolkata, no date).
Barddhaman had a population of
6,919,698 according to the 2001 census.
Although at the time of writing the 2001
census breakdown of occupations are not
available, we can use the proportion re
porting themselves in the 1991 census as
primarily agricultural labourers (6.48 per
cent) and cultivators (9.12 per cent) to
estimate the number of people in each
occupational category in 2001. We thus
estimate there to be 4,48,000 and 6,31,100
people who would have reported them
selves as primarily agricultural labourers
and cultivators respectively in the 2001
census.
Four cultivators were asked in detail
about their labour requirements in the
previous aman and boro transplanting and
harvesting seasons. Their estimates for
transplanting were 8-10 person days per
bigha (uprooting, carrying seedlings and
transplanting) and 10-19 person days per
bigha at harvest (cutting, binding, stacking
straw, threshing and storing). Boro rice
usually yields significantly more per bigha
than aman. However, the cultivators’ es
timates did not provide evidence of the
difference between them.35
The higher end of the range for harvest
were estimates by people who would not
work in these tasks themselves nor expect
their family members to do so. The lower
end was estimated by two employers who
would expect themselves and other mem
bers of their families to do harvest work.
As they were asked to detail the direct
financial cost of cultivation they did not
include the opportunity cost of their own
labour.3*5
The range of total labour required in
each season for Barddhaman district is
thus:
Aman transplanting: 25,140,000 - 31,425,000
31,425.000 - 59.707,500
Aman harvest:
Boro transplanting: 9,480,000 - 15,800,000
15,800,000 - 21,014,000
Boro harvest:
The mean number of days spent away
across the 141 groups recorded in the
Economic and Political Weekly
migration survey for the year 1999-2000
was 16.80. However, this figure was
weighted downwards by the counting of
groups in the Murshidabad locality which
travelled out more than once per season.
The mean number of days for that locality
was 12.4, whereas the means for the other
three localities were 21.65, 19.53, and
23.00. We therefore assume for the pur
poses of this calculation that the length of
a season is 20 days. This is a crude assump
tion because season lengths vary. Boro
harvesting is often done at a faster pace
over less days, especially when people are
employed on piece rates.37
Therefore, we calculate the total number
of people required for each season of rice
cultivation in Barddhaman district to fall
within the following ranges:
Aman transplanting 1.257,000 - 1,571.250
1,571,250 - 2,985,375
Aman harvest
Boro transplanting 4,74.000 - 7,90,000
7,90,000 - 1,050,700
Boro harvest
If every one of the 631,100 cultivators
and 448,000 agricultural labourers esti
mated to be resident in Barddhaman dis
trict in 2001 (a total of almost 1,080,000
people) were to work throughout each
season in these tasks, there would still be
a major shortfall of at least half a million
workers in the aman harvest. However,
many of those classifying themselves as
cultivators prefer not to do transplanting
or harvesting work and thus the shortfall
is likely to be much greater.38
Importantly, these calculations were
consistent with our own observations in
five destination localities in Galsi, Memari,
Barddhaman Sadar and Katwa blocks of
Birbhum district and in Nanoor Block of
Birbhum district. In all these villages the
number of migrant labourers was higher
than the number of local wage labourers
in the peak seasons.
Our journeys around the countryside
during the seasons of outmigration and
return have included periods of detailed
observation trying to count migrants at
labour market places. Estimates have been
made by railway staff, bus drivers and
owners and shopkeepers. These numbers
are less reliable as estimates of the number
of migrants, as our own migration survey
suggests that in three of the four source
area localities labour market places were
used relative little for recruiting labour or
finding work.39 Moreover, we only stud
ied two labour market places in detail:
Bankura bus stand and Katwa railway
station and we could only briefly visit
some of the rest, for example, Tinkonia bus
December 8, 2001
stand and the railway stations at
Barddhaman and Guskara. Others, includ
ing, importantly Nadanghat in Nadia dis
trict, and Kusumgram in Barddhaman were
not visited.
At Bankura bus stand alone, stallholders
estimated that at the peak of aman harvest
recruitment as many as 10,000 to 15,000
migrants passed through the bus stand in
a day. The station manager at Katwa sta
tion, narrow gauge line estimated that at
the end of the season 15,000 to 20,000
migrants returned each day. Given that the
peak part of each main rice work season
can last for as much as 15 days, this could
mean 1,50,000 - 2,25,000 passing through
Bankura bus stand and 2,25,000 to 3,00,000
passing through Katwa railway station.
One of us estimated that 6,000 labourers
left for Barddhaman for boro transplanting
work from Dumka bus station on one day
in late January 1999. Bus operators inter
viewed reported that this outmigrating
season lasted for 15 days. This suggests as
many as 90,000 labourers left from
Dumka bus station alone. Such estimates,
each of which refer only to one part ot one
stream, do not contradict the scale of
the estimate above based on demand
calculations. 033
Notes
1 Our thanks especially to the migrant workers,
their employers, officials and others, who spent
lime and effort engaging with us during the
study. This research would not have been
possible without the direct and indirect support
of many colleagues, including coordination
and- research assistance by Somnath
Chattopaddhyay,research assistance by Sujata
Das-Chowdhury andMalini Munshi; sustained
insightful guidance from beginning to end
from the late Sunil Sengupta: valuable advice
from Nirmala Banerjee, Sonia Bhalotra,
Debabrata Bhattacharya, Partha Chatterjee,
Samantak Das, Jean Dreze, Arjan de Haan,
David Mosse, Nitya Rao and Samita Sen; and
the reflective feedback of others including
Mukulika Banerjee, Tony Barnett, Kumkum
and Ranjit Bhattacharya, Piers Blaikie, Rahul
Bose, Srabani Chakrabarty, Surendranath
Chatterjee. Meena Dhanda, Madan Gopal
Ghosh, John Harriss, Barbara Harriss-White,
Mark Holmstrdm, Vegard Iversen, Cecile
Jackson, Praveen Jha. Catherine Locke,
Richard Palmer-Jones, Kirat Randhawa. Kunal
Sen, Dikshit Sinha, Rene Veron and Glyn
Williams. The map was drawn by Philip Judge.
Rogaly is grateful to the Centre for Studies
in Social Sciences, Kolkataand he and Coppard
to Palli Charcha Kendra, Visva-Bharati, where
they were affiliated during the research, and
to those institutions’ respective directors Partha
Chatterjee and Onkar Prasad. We thank the
Department for International Development
• (UK) for funding. Views expressed and any
4557
___
Position: 4513 (1 views)