The Commercialization of Agriculture in Colonial India: Production, Subsistence and

Reproduction in the 'Dry South', c. 1870-1930

Author(s): David Washbrook
Source: Modern Asian Studies, Vol. 28, No. 1 (Feb., 1994), pp. 129-164
Published by: Cambridge University Press
Stable URL: http://www.jstor.org/stable/312924
Accessed: 01-04-2015 08:52 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact [email protected].

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to Modern Asian Studies.

http://www.jstor.org

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
Modern Asian Studies 28, i (i994), pp. 129-164. Printed in Great Britain.

of Agriculturein
The Commercialization
Subsistence
ColonialIndia: Production, and
in the 'Dry South' c. I87o0-930
Reproduction
DAVID WASHBROOK

Universityof Oxford

Although it would now seem established beyond question that agri-

culture in most parts of India had been exposed to commercial
influences from medieval times, there can be little doubt that a variety
of developments from the second half of the nineteenth century greatly
strengthened those influences.1 Railways and road transport made
possible a huge expansion in cash cropping, for national and interna-
tional markets, and production regimes across the subcontinent were
placed in a new context of opportunity-and of pressure.2 While so
much would scarcely be disputed among historians, what has
become-and remained-more controversial, however, is an under-
standing of the implications of this extended commercial logic for
agrarian economy and society. Since colonial times, opinions would
seem to have been divided between 'optimists', for whom com-
mercialization marked progress and a growing prosperity for all;
'pessimists', for whom it marked regress into deepening class stratifi-
cation and mass pauperization; and 'sceptics' who held that it made
very little difference and that its impact was largely absorbed by pre-
existing structures of wealth accumulation and power on the land.3
1
For the growth of commerce in late medieval South India, see S. Subrahmanyam,
ThePoliticalEconomyof Commerce: SouthernIndia I500o-650 (Cambridge, 1990); also S.
Subrahmanyam (ed.), Merchants,Marketsand the State in Early ModernIndia (Delhi,
1990).
2 See J. Hurd, 'Railways and the Expansion of Markets in India, I861-1921',
Explorationsin EconomicHistory12, 1975; M. McAlpin, 'Railroads, Prices and Peasant
Rationality: India 86o- goo', Journalof Economic Histo?y34, 1974.
3 These three
positions can be traced back to debates in the later nineteenth
century. For a recent optimistic or 'meliorist' account of, particularly, Western India,
see M. McAlpin, Subjectto Famine:Food Crisis and EconomicChangein WesternIndia,
1860-1920 (Princeton, 1983); for a more pessimistic view, S. Guha, The Agrarian

oo26-749X/94/$5.oo + .00oo 994 Cambridge University Press

I29

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
I30 DAVID WASHBROOK

One historical context in which a debate along these lines has been
conducted, in the pages of this journal and elsewhere, is that of 'the
dry' interior of South India. My own 'pessimistic-to-sceptical'
account of its colonial development saw most of the benefits of
expanded cash cropping in cotton and groundnuts passing to a small
group of larger-farmer 'magnates', whom I took to dominate the
markets in credit, commodities and employment. These magnates
were established prior to and, in many ways, independently of the
expansion of commerce, through positions in the revenue and kinship
systems, which enabled them to take up a highly advantageous posi-
tion in the new market context as well. From these positions, they
denied access to market benefits to their smaller clients and used the
forces of commercialization to strengthen their own social and politi-
cal grip on the countryside.4
My account of the 'dry' South was directly challenged by Bruce
Robert who, from a detailed study of one of the 'driest' districts of all
(Bellary district), came to an opposite and very 'optimistic' conclu-
sion. For him, the deepening commercialization of the period pro-
vided the context for a liberation of the small peasantry and for an
increase in their prosperity. He drew attention to a significant
increase in the incidence of landholding, particularly the holding of
small plots of land which were intensively cultivated with high value
cash crops and yielded better returns per acre than the lands of larger
farms. He further argued, from the detailed studies of the Cotton
Committee (1927), that there was no evidence of a monopsonization
of the cotton market, such as would give small producers lower
returns than larger ones. Nor did he see significant restrictions in the
credit market, such as would make small producers the consumption
debtors of large ones. Finally, he took the growth of agricultural
production in the region to be clearly shown by evidence of increasing
cultivated acreage and of capital investment, particularly in carts.5
Needless to say, my own initial reaction to Robert's case, which was
largely based on the same sources that I thought I had consulted, was

Economyof the BombayDeccan(Delhi, 1985); for a sceptical view, N. Charlesworth,

PeasantsandImperialRule:Agriculture
andAgrarianSocietyin theBombayPresidency,i850-
I935 (Cambridge, 1985).
4 See
my, TheEmergence of ProvincialPolitics: TheMadrasPresidency
1870-9g2o (Cam-
bridge, 1976); also, 'Country Politics: Madras I870-1930', ModernAsian Studies7,
I973; also 'Economic Stratification in Rural Madras' in A. Hopkins and C. Dewey
(eds), TheImperialImpact(London, 1978).
5 B. Robert, 'Economic Change and
Agrarian Organization in "Dry" South India
1890-I940: A Re-interpretation', Modem Asian Studies 17, 1983.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA 1870-1930 I31
one of sheer puzzlement. On certain points, I had no choice but to
concede best to him: I had seriously underestimated the evidence of
deepening market penetration in the region and the way that it had
come to be reflected in the growth of a small-holding agriculture
devoted to cash crop production. At the same time, however, Robert's
translation of this evidence into a story of growing prosperity and
small peasant liberation did not ring true in the light of other
evidence.
Bellary district was distinguished in the Madras Presidency as the
only district with a negative rate of population growth between 1891
and I93I.6 It was at the epicentre of the Southern 'famine risk' zone
and experienced repeated crop failures and dearths-in 1876-78, 1892,
I896-97, I900, 1917, I922-24. Admittedly, after the coming of the
new famine code in I88os, these dearths did not produce deaths on the
same scale as before (between 1876 and 1878 one-quarter of the
population had died).7 Nonetheless, the havoc that they wrought with
crop production was clearly evidenced in the records.8Further, there
was little evidence of increasing productivity to off-set these regular
losses. Bellary (and, in fact, the whole of the Madras Deccan) were
again statistically distinguished in the Presidency for showing no
increases in average crop yields across this period.9 Further, while the
statistics on commercial capital certainly showed 'progress', all the
indices of 'productive' capital-from wells to cattle to ploughs-showed
a common decline.'1 This pointed to a deepening paradox: that while
there had undoubtedly been an expansion of commercialization, it
was to be associated not with a broadening prosperity, but with a
progressive crisis in agricultural production and social reproduction.
The possibility of this crisis was something which, in our different

6
Robert calculated this decline at a rate of o.o6 per cent per annum between I891
and I931. Robert, 'Economic Change', p. 63.
7
MadrasDistrictGazetteer.W. Francis, BellaryDistrict(Madras, 1904), p. 135.
8 For example,the droughtsof I891-92 and 1896-97 reducedcroppedarea by 25
and I8 per cent, respectively, from that of their previous seasons. Bad seasons
between 1917-18 and 1923-24 kept average cropped area 6 per cent lower than in
1916-I7. See Government of India, Agricultural Statisticsof BritishIndia,Quinquennial
Series, 'Madras: Bellary District' (Calcutta and Delhi, various).
9 See Government of Madras, Seasonand CropReports(Madras, annual). In fact,
Bellary's crop yield 'norm' was briefly reduced for a few years in the I9ios, before
being raised back to pre-9goo levels, which the 'seasonal adjustment' factor indicates
it never reached. Bellary's stagnation is marked against the apparent dynamism of
the Southern cotton belt in Coimbatore-Tinnevelly, where, particularly, cotton yields
rose noticeably over the period.
'0 See below, Section V.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
I32 DAVID WASHBROOK

ways, both Robert and I had missed: he from the supposition that
commercial growth must have produced economic growth; I from the
supposition that the absence of economic growth must have been the
result of an incomplete penetration of market forces. Neither of us
considered that the expansion of the market economy, itself, might
have had negative and deleterious effects on the bases of production
and social reproduction. In retrospect, our mutual mistake was to put
too much faith in conventional 'economic' theory and in models of
'the market' as an obvious and automatic source of growth.
But if, in Bellary, the expansion of the market economy was
attended, at least in time, by a crisis of production and reproduction,
why should this have been so; and what might an answer tell us for
the study of other Indian regions, even those where economic growth
did more assuredly take place?

II

Bellary district, of some 5900 square miles,11 was one of the four
'Ceded' districts in the Madras Deccan (and is now in Karnataka
state). In 187I, it had a population density of about 172 to the square
mile, which, following various vicissitudes and recoveries from famine
and disease, stood at 170 in i93 .12 It was one of the 'driest' districts
in South India with only about 2.5 per cent of its cultivation irrigated
at any time before I930.13 Its agriculture was almost entirely
dependent on the rains and was based, first and foremost, on the
production of millet dry grains and pulses, which covered over 70 per
cent of acreage in the I88os and over 65 per cent even in the late
I920S.'4 The district was divided between two contrasting soil types:
heavy black soil, particularly good for cotton production but difficult
to work, predominated in Bellary, Adoni and Alur taluks; poorer
quality red soil, which was much less productive until the coming of
11The district
originally included a number of taluks which, after the Great
Famine, were separated off into the separate Anantapur district. Subsequent to that,
too, taluk boundaries underwent several changes. At various times between 1890 and
1930, the district varied between 5730 and 5975 square miles.
12 Because of the
changes in boundaries, it is easiest to express population changes
in these terms. See Censusof India, i87i, Reporton theCensusof theMadrasPresidency, vol.
i (Madras, 1874), p. 68; 193I, vol. 14, pt 2, p. 2.
13
See Appendix to Bellary District Gazetteer, 1930, p. i6.
14 Food crops occupied between I.4 and 1.6 million acres of Bellary's cultivation,
outside drought years, fairly continuously between 1890 and I930. Seasonand Crop
Reports.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA 1870-1930 I33

groundnut cropping in the early twentieth century, covered most of

Rayadrug taluk. By the late I920S, groundnut had generally joined
cotton even in black-soil areas as the second principal commercial
crop.15 Between 1890 and I929, the total acreage under cultivation
increased by about 2o per cent. Cotton acreage over the same period
rose by about 60 per cent; and, between I9IO and I928, groundnut
acreage increased from almost nothing to about 13 per cent of total
cultivation. 16
An examination of the available data on landholding suggests that,
as Robert argued, the principal force behind the expansion of cropped
acreage was the proliferation of small farms. Between 1890 and 1930,
the number of pattas, issued by the government for the payment of
less than Rs IO in land revenue increased from 63,000 to I13,000 and
the amount of land covered by these pattas rose from about 580,000 to
800,000 acres.17 Of course, not all small farms were new: some com-
prised the division of older large farms. However, in Bellary this
would account for only a small part of the increase. Both 'medium'
pattas of Rs 10-50 and 'large' pattas of Rs 50+ increased their num-
bers and acreages over the period although not by anything so con-
siderable an extent. Medium pattas rose from 32,000 to 43,000 and
their acreage from about 800oo000to 9oo,ooo; and large pattas from
4000 to 4800 and their acreage from 300,000 to 340,000.18
What lay behind this remarkable expansion and proliferation of,
especially, small farming? As indicated previously, given Bellary's
history of negative demographic growth, the pressure of population
can be discounted. Robert supposed that would-be small farmerswere
'induced' into cultivation by the prospects of high cash crop returns.19
And there seems no doubt that small farmers did concentrate on the
principal cotton cash crop: the Cotton Commission (I927) reported
15On the ontheMarketing
groundnut'revolution',see Governmentof India,Report of
Groundnuts in IndiaandBurma(Delhi, I941).
16 In 1890-9 , total cropped area was 2. Io million acres, of which 346,000 were
under cotton. In I928-29, 2.46 million acres were cropped, of which 593,000 were
Statisticsof BritishIndia,Madras 1890/91-95/96, p. 98; Season
under cotton. Agricultural
and CropReport1928/29. For purposes of comparison, there should be a downward
adjustment of 08,000 acres in the first estimate of cropped area to allow for boundary
changes. But the cotton figures require little alteration as the changes did not affect
the principal cotton-producing taluks.
17 Calculated from 'Statement of the Rent-roll' in Government of Madras, Reporton
theSettlementof theLandRevenue in theMadrasPresidency (Madras, annual series).
18LandRevenueReports.
19'The
higher prices for farm commodities induced cultivation of new lands'.
Robert, 'Economic Change', p. 63.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
I34 DAVID WASHBROOK

that 87 per cent of the crop, in the Bellary villages which it surveyed,
was grown on plots of less than 25 acres; and 21 per cent of it on plots
of less than 5 acres.20A very large part of the 200,000+ acre increase
in cotton cultivation between the I89os and late I920S would seem to
have resulted from the activities of small farmers.
However, whether this activity was meaningfully 'induced' by the
buoyancy of cash crop markets may be another matter. The economic
history of the district was extremely bumpy and might best be divided
into four distinctive 'mini-epochs'. The years I890 to about 1904 were
difficult for agriculture. There were droughts in I892, I896-97 and
1900 and cotton prices were little more than steady.21 From about
I904 to the onset of the First World War, Bellary enjoyed something
of a 'golden age' with no serious droughts, cotton prices rising by
about 40 per cent and the new cash crop of groundnut starting to
make its entry.22The Wartime years produced something close to
economic mayhem with rocketing cotton prices, the collapse of the
groundnut market and, from I917-19, crop failure and famine.23 The
decade of the I920s was disappointing: opening in famine, drought
seasons recurred in 1922 and 1924 and, although the weather then
became more stable, cotton prices went into slow decline for several
years before the Great Depression of I930 put them into complete
collapse.24
A model of small farm proliferation 'induced' by rising market

20
Government of India, Indian CentralCottonCommittee.GeneralReporton Eight
Investigationsinto the Finance and Marketingof Cultivators'Cotton (Bombay, I929).
'Madras', p. 51. The Madras investigation, which centred on Bellary district, was
conducted in 1926-27.
21 Cultivated
acreage both opened and closed the decade of the I89os at about 2. I
million acres but plummeted twice between in response to severe droughts. Bellary
cotton prices rose only from Rs 15.5 to Rs 17 per imperial maund. See Agricultural
Statistics,'Madras: Bellary district'.
22
Cropped area rose from 2.2 million acres in 1901-02 to 2.4 million in I914-15.
Cotton acreage rose over the same period from 287,000 to 411,000 acres. Cotton
prices rose from Rs 17.5 to Rs 24. Seasonand CropReports,and Government of India,
Pricesand Wagesin Indiai86i to I9g2 (Calcutta, 1923).
23
Cropped area fell from 2.4 million acres in 1914-15 to 2.2 million in I918-I9.
Cotton prices peaked in I 918 at Rs 76 per maund and then halved again over the next
two years. Cholum/jowar prices rose from Rs 2.25 per maund in 1915-16 to Rs 7.25 in
1918-19 and Rs 6.5 in I919-20. Season and Crop Reportsand Prices and Wages.
24 Cropped area remained static at around 2.2 million acres from
1920-24 and then
rose to 2.45 million acres by the end of the decade. Groundnut acreage increased from
20,000 to 315,000 acres and cotton from 446,000 to 593,000 acres. Cotton prices were
extremely unstable (as was acreage): dropping sharply in the immediate aftermath of
the War; recovering between 1923 and 1925; and then declining slowly until 1929
when they halved, from Rs 24 to Rs 12 per maund, at the onset of the Great
Depression. Seasonand CropReportsand Pricesand Wages.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA 1870-1930 I35
returns must anticipate that the growth of small holdings would have
been fastest during Bellary's 'golden age' between I905 and I9I5, and
slowest during the hard times of the I89os and the problematic 1920S.
In fact, however, the patta data indicate exactly the reverse. Between
I890 and 1901, small pattas proliferated at the amazing rate of about
I640 per annum. Between 1921 and 1929, they increased at the rate of
1345 per annum. But between 1905 and 1915, when market condi-
tions were at their best, they grew at the rate of only 540 a year. In
effect, the proliferation of smallholdings appears to have been more a
response to economic adversity than to market opportunity.25
But why should the Bellary 'peasantry' have responded to adversity
in this way? To answer this question it is necessary to look at the
context from which they came to small farming. The evidence from
the middle of the nineteenth century would seem to suggest that they
came, predominantly, from a background of landless farm labour.
Although Bellary experienced a slight de-urbanization between I89I
and 193I (the result, mainly, of the removal of an army barracks from
Bellary town), the consequent increase in rural population is too small
to account for many of the new farmers.26In the early I87os, before
the Great Famine, the ratio of pattas to population in the district had
stood at about I:12, indicating (if we take a family to consist of five
people) a large rural population without land. After the Great
Famine, as cultivation came to recover in the i88os, the ratio fell to
I:9 and, by the end of the 1920S, it stood at 1:6, which would suggest
almost nobody without some kind of access to land.27
Evidence on land distribution and farming systems in the middle of
the nineteenth century helps further to clarify the picture. According
to contemporaries, the centre of local production regimes was pro-
vided by a small elite of 'magnate' families who commanded huge
landed resources and who worked them, predominantly, with perma-
nent farm servants and gang labour. It was these magnates who, at
this time (whence my original 'magnate' model was derived), produ-
ced most of the cotton and also controlled local 'grain heaps', whose
storage facilities were vital for fending off periodic droughts and
famines. Such magnates built their houses on top of large grain pits
from which they distributed subsistence wages to employees and grain
loans to clients.28
25 Calculated from 'Statement of the Rent-roll', in LandRevenueReports.
26
From I41,928 to 138,070. Census of India, 189i, vol. XIV, p. 6; 9g3y,Madras, vol. 2,
p. 8.
27Calculated from 'Statement of the Rent-roll', LandRevenueReports.
28 '... the
bigger ryots-those who own wide acres, employ many hands and are as

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
136 DAVID WASHBROOK

The position of significance which the magnates had achieved in

the district by the i86os and I87os may have owed much to the
operations of the revenue system in the earlier decades of the century.
Initial rates of assessment on Bellary district were penal: in money
terms they were almost as high in the I82os as in the 192os. This put
great power in the hands of village officer families who were able to
manipulate them on the ground. The power was expressed most
particularly in relation to the classification of inam land, which bore
privileged (and often nil) rates of assessment. Thomas Munro, in
making the first settlement of the district, had been all too conscious of
the avariciousness of his employers and, to off-set their greed, had
granted extensive rights of inam protection. Under his I804 settle-
ment, nearly half of the cultivated land (and all the best land) was
classified as inam-and nearly 30 per cent of cultivation, as late as the
I920S, remained so.29 Under the confiscatory revenue policies of the
Company state, it was only the protection offered by the inam shield
that permitted any profits raised out of agriculture to be retained in
the district. As Bellary Collectors repeatedly claimed, cultivation was
heavily concentrated on inam lands 'to the detriment of the revenues'
and policies were introduced to try to force inamdars to take up
valueless government land to at least the extent of inam production-
thereby giving large inamdars reciprocally large holdings of revenue-
paying land.30
And, as matters unforeseen by Munro were to turn out, the inam-
dars of Bellary were to be large. Munro's initial settlement had reflec-
ted a wide distribution of inam lands across broad sections of the
population. However, power over the revenue system enabled the
principal village officer families to shift the distribution very much in
their own favour. In I804, principal village officers had held only 22
per cent of inam land.3' By the time of the Inam Commission in I869,
this had changed to over 60 per cent.32 In effect, the 2000 or so
principal village officers in Bellary's Iooo revenue villages had come
to hold nearly 400,ooo acres of the best quality land in addition to
their 'ordinary' ryotwari holdings.

often as not traders in produce and moneylenders as well as landholders'. Francis,

BellaiyDistrict,p. 99; also seeJ. Kelsall, A Manualof theBellaryDistrict(Madras, 1872),
pp. 260-70; and my 'Economic Stratification'.
29B. Stein, 'Does Culture Make Practice Perfect?'in B.
Stein, All theKings'Manna
(Madras, 1984).
30 See N. Mukherjee, TheRyotwariSystemin Madras(Calcutta, 1962).
31
Stein, 'Does Culture'.
32
By the late i86os, patels and kurnams held 386,918 of Bellary's 635,25I acres of
inam land. Kelsall, Manual,p. 19 .

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA I870-I930 I37
Given, before the revised ryotwari system of the I85os (which
greatly reduced rates of assessment), the economic impossibility of
cultivating at profit without inam land, the central role played by
'magnates' in the Bellary production regime, and the large number of
'landless' labourers within it, become readily explicable. For a large
section of rural society, working as the farm servant of an inamdar
brought a 'share' of revenue-protected production;33 whereas
independent farm production brought only tax-bills. But there was a
further logic to 'large farm' production in this context as well.
Cotton was, and had long been, the principal cash crop of the
district. In the circumstances of the period, however, the best methods
of producing and marketing it favoured the large producer with
capital and land to spare. To gain the best cotton crops from black-
soil lands, for example, required regular deep ploughing with heavy
metal ploughs drawn by upwards of a dozen bullocks. Only 'magnate'
farmers had the capital for such a form of cultivation. Equally, cotton
production was extremely soil-exhausting and was best pursued
where land could be fallowed for extensive periods: only cropping
regimes where land was not a scarce factor of production could afford
lengthy fallows. Further, poor road transport conditions (before the
construction work completed during the Great Famine of I876-78,
which almost doubled Bellary's road mileage) made it hard for small
farmers to get their crops to market and gave sellers who brought
large quantities of crop to the auction block substantial advantages.34
Farm labourers who were paid in terms of a 'share' of the product
accruing to these advantages-and a sizeable share reckoned, in the
I86os, at a customary one-third of the crop-were plainly better off
than had they tried to cultivate, in a small way, on their own.35
Service on the large magnate farms also gave them entitlements to
share in magnate grain stores, the difference between life and death in
times of famine.
What had changed, by the later nineteenth century, to break up
this magnate-centred organization of production and to convert large
numbers of erstwhile farm labourers into small, independent pro-
ducers?As indicated earlier, the specific timing of the conversion must
raise suspicions about the extent to which it was the result of market
'inducement'. An examination of the precise conditions of production
and marketing for small farmers broadens these suspicions. Admit-
33 For a discussion of the 'share' economy, see B. Stein, 'Politics, Peasants and the
Deconstruction of Feudalism in Medieval India', Journalof PeasantStudies12, 985.
34 See Kelsall, Manual, pp. 262-7.
35 Ibid., p. 262.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
138 DAVID WASHBROOK

tedly, by the later nineteenth century, land revenue charges were now
no longer significant and transport improvements made it possible for
small peasants to sell their crops without great disadvantage.
However (and pace Robert), there is no evidence of any production
advantages in Bellary-type farming to small-scale production
methods and several of the old disadvantages still operated.36As in
other parts of the Deccan studied by the Farm Management Survey of
the 1950s, the celebrated 'inverse farm size:productivity ratio' did not
apply and most of the surveys of Bellary conducted in the 192os and
1930S actually found farm size:productivity ratios to be direct.37The
Cotton Committee (1927), for example, noted the highest cotton
yields to come from the largest cotton fields.38
In the context of production methods, this is not surprising. Cattle
were extremely expensive to buy (Rs 200-400 a pair in the late 1920s)
and to keep (Rs 92 p.a. in the early 1930s).39 Small farmers simply
could not afford to own them-with the result that either they had to
hire them (which raised their costs of production) or, more usually,
they had to skimp on ploughing, weeding and manuring.40Indeed,
lack of cattle also kept them out of the new market in groundnut since
its principal cost of production (nearly 55 per cent, according to the
Imperial Council of Agricultural Research's close study of Bellary
production methods in the early 1930s) came from cattle and
manure.41Commercial farming in Bellary without 'owned-cattle' was
either a very expensive or a very restricted business.
36 Robert
(p. 75) implied that Bellary farming could be drawn under a general
'inverse farm size:productivity' rubric derived from the Farm Management Surveys.
But, as Bharadwaj has shown, the rubric principally operated in conditions of
irrigated agriculture. The surveyed districts with 'dry' production conditions most
similar to Bellary's were Amraoti and Akola districts, further North across the Dec-
can. They possessed no significant inverse ratio. See K. Bharadwaj, ProductionCondi-
tionsin IndianAgriculture(Cambridge, 1972).
37 The
sample sizes in both surveys were too small, and too biased towards larger
producers, to make this evidence conclusive. But it can be said that, in both cases, the
farms with the highest per acre productivities were large. MadrasProvincialBanking
EnquiryCommittee,vol. V (Madras, 1930), pp. 272-350; Imperial Council of Agri-
cultural Research, Reporton the Costof Production of Cropsin the PrincipalSugarcaneand
CottonTractsof India,vol. IV (Delhi, i938-39), pp. I I-200.
38
The Committee noted output of io8 lbs per acre on cotton areas over 50 acres
but just go lbs on those under 5 acres and 85 on those between 5 and 25 acres. Cotton
39 MPBCE, II, p.
Committee,p. 5I. 297; V, pp. 3I0-I2; ICAR, IV, p. 14.
40 MPBCE, V, p. 272, 298; Francis, Bellary District, p. 85.
41The ICAR imputed a rental charge of c. Rs. 3.5 per acre to production costs. If
this is removed, as irrelevant to the circumstances of most groundnut farmers,
charges imputed to bullocks and fertiliser come to about 55 per cent of costs of
production. ICAR, IV, pp. 192-8. On the importance of cattle to groundnut, also see
C. Baker, An IndianRuralEconomy880o-955 (Oxford, 1984), pp. 145-53.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA I870-I930 I39
In fact, not only production but also market conditions make the
'decision' of small farmers to enter cotton production difficult to
understand. During both the i89os and 1920s, when small farming
was expanding at its fastest, cotton prices were not at their best. Their
greatest rise in this period came across the igios when, as we have
seen, small patta formation actually slowed down. Indeed, compared
to the price of grain, cotton prices underwent a relative decline across
the whole of this period. Robert seriously miscalculated the ratio of
increase in cotton and grain prices at about parity (56 per cent and 60
per cent respectively). While cotton prices would seem to have risen
by about this amount, the rise in grain prices was much nearer the
order of I6O per cent (from an average of Rs 1.45 per maund in the
last quinquennium of the I88os to Rs 3.75 per maund in the four years
preceding I930-none of these years being famine years).42In effect,
the grain: cotton price ratio almost halved between the I89os and the
late I920S, making the small peasantry's 'decision' to concentrate on
cotton farming, notionally at the expense of grain production, all the
more wonderful.
Behind this decline in relative cotton prices there lay general prob-
lems in the market for short-staple 'Northerns and Westerns', the
local varieties of Deccani cotton which, as a late I920S commentator
put it, 'have lost much of their former reputation'.43With the growth
of longer-stapled and watered varieties of cotton in other parts of
India and the South, the market for Bellary cotton went into decline
after the First World War. Paradoxically, it was the market conditions
associated with this decline, which Robert reported as particularly
favourable to small producers. He noted, from the Cotton Committee
(I927) report, that, allowing for transport costs, there was little dif-
ference between village and town prices for cotton; that the bulk of the
crop was purchased by itinerant commission agents who bought from
large and small producers alike at comparable rates; that there was no
evidence of'magnate' interventions to corner the crop and gain specu-
lative profits.44
What Robert failed to note, however, were the reasons given for this
by the Cotton Committee itself. In the depressed Bellary market,
virtually the whole of the crop was forward-contracted by the pur-
chasing houses at fixed rates and quotas. As these quotas were easily

42
Robert, 'Economic Change', p. 63; AgriculturalStatistics,'Madras: Bellary dis-
trict' and Seasonand CropReports.
BellaryDistrict(Supplement) (Madras, 1930), p. 68.
43 MadrasDistrictGazetteer,
44Robert, 'Economic Change', p. 74.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
140 DAVID WASHBROOK

TABLE I
Bellary Quinquennial Wage Census

1900* i906 I9I I916 1921* I926

I00 131 125 233 78 II0

*Years of serious dearth and high grain prices.

Source: Robert, 'Economic Change', p. 76; derived from QuinqennialWagesCensuses,I906-1936.
Tamilnadu Archives.

filled (in fact, overfilled since 12 per cent of the crop was left unsold
and unsaleable at the season's end), there was no possibility of any
speculative profit in the market. The 'commonness' of price seen
throughout the market was a commonness of bottom price, induced
by purchasing-house monopsony, not a commonness of top price,
beaten up by fierce competition, as Robert supposed.45
But why, then, should small farmers decide to engage in, especially,
cotton cultivation at just the times when the market for it had gone
flat? A more comprehensible answer comes from a view of the situa-
tion whence they came. As Table i shows, the 'golden age' of Bellary
farming, when small patta formation was minimal, coincided with a
period of significant rises in effective wage rates. By contrast, the
192os, when small patta formation expanded prodigiously, saw wages
first in crisis and then continuously depressed. Small patta formation,
here, would seem to have been a response to instabilities in the wage
sector.
But what was happening in the wage sector to precipitate this
response? There seems little doubt that the 'magnate-centred' produc-
tion regime began to change in the i86os and I87os, partly as a result
of the 'new' ryotwari revenue system but, perhaps mostly, because of
the coming of the railways. During the American Civil War, the
district had responded to boom cotton prices with a very considerable
expansion in production, which showed it firmly engaging in a much
larger scale of market activity.46 With the post-War fall in cotton
prices, cotton acreages fell back but market engagement continued in
the grain trade.
As Michelle McAlpin, among others, has argued, the coming of the
railways can be linked to the very considerable rise in grain prices,
which occurred nearly continuously from the I86os to the late 1920S.
Local spatial limitations on grain markets were broken and grain was
45 Cotton Committee,pp. 11-33.
46 Kelsall, Manual, pp.
318-19.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA 1870-1 930 14I

'freed' to find its best price over wider areas.47 In the case of the
market for millets, this 'best price' proved to be spectacularly high:
millet prices rose faster than those of any other major food-grain-for
long periods, faster than those of any commercial crop and of
manufactured goods. The reasons for this would seem to lie in the
extent to which millets were associated with 'dry' production regimes
subject to constant interruption by the weather. There was likely to
be, at least, a dearth somewhere in the millet zone almost every year
with the result that its 'dearth' prices became spread through the
market hiking up the cost of grain everywhere.
There is evidence that, in the post-American Civil War 'depression'
and the early I870s, 'magnate' farmers were starting to run down
their local grain stores and to take advantage of rising market prices.
Indeed, this was suggested as one reason why the Great Famine of
I876-78 had been so severe in terms of loss of life, particularly for
labourers and small peasants who depended on access to these stores
as their own insurance mechanisms.48 It was noticeable that the
death-toll was heaviest in the most commercially-advanced taluks of
the district (Adoni and Alur where nearly a third of the population
was lost).4' After the Great Famine, the run down of local storage
facilities would seem to have continued and was noted again as a
feature of the 1896-97 famine.50By the I920s, large-scale local storage
of grain was held to be a thing of the past and, in effect, subsistence
relations had been put on a commercial and cash basis. The credit
surveys of the 1920s particularly remarked on how, in comparison to
thirty years before, intra-rural borrowings and lendings, which had
been dominated by transactions in grain, were now run largely
through the medium of cash.5'
Needless to say, the commercialization of grain trading had far-
reaching implications for the employment of labour since it was shares
47 M. McAlpin, 'Railroads, Cultivation Patterns and Foodgrain Availability',
Indian Economicand Social History Review 12, I975.
48 See ParliamentaryPapers,I881, vol. LXXI, pt 2: Reportof theIndianFamineCommis-
sion, i88i, Appendix 3. Also my 'Economic Stratification' and 'Country Politics'; D.
Arnold, 'Famine in Peasant Consciousness and Peasant Action: Madras 1876-78' in
R. Guha (ed.), SubalternStudiesIII (Delhi, I984).
49 Francis, BellaryDistrict,p. 135.
50 See ParliamentaryPapers,1898, vol. XXXII: Reportof theIndianFamineCommission,
i898, Appendix, 'Madras'. Also my 'Economic Stratification' and 'Country Politics';
D. Arnold, 'Famine'.
51 See A. Kolliner, 'The Structure of Rural Credit in the Ceded Districts of the
Madras Presidency', paper presented at Conference of Rural Agrarian History,
University of Pennsylvania, I975, pp. 39-49.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
I42 DAVID WASHBROOK

in, and entitlements to, magnate grain stores that had articulated
magnate-centred relations of production. The most obvious conse-
quence would appear to have been a general disemployment of per-
manent and tied farm servants and a casualization of wage-labour.
Certainly, if the, admittedly impressionistic, data on the situation in
the mid-nineteenth century can be believed, Bellary large farming, by
the I920S, was distinguished by its absence of permanent farm
servants. The Banking Enquiry (1929) and the ICAR survey reported
farms of upwards of oo00acres operating with seldom more than two or
three, and sometimes none.52 And this was not because magnates
were deploying their own family labour instead: the ICAR noted
Bellary (large) farming to operate with the highest ratio of hired to
family labour of any part of British India that it surveyed.53
Further evidence of the casualization of labour comes from the
wage census where the Bellary reporter, at the turn of the twentieth
century, was among the first in South India to claim to discern a
definite shift from payments in grain and by 'custom' to payments in
cash or by price-related grain dole.54 The latter style of payment is
scarcely compatible with permanent farm labour.
Besides the rising value of grain, which magnate farmers no longer
wished to 'share' with their employees, the move towards a casualized
labour force also fitted with other developments in Bellary large farm-
ing. Except for the 'golden age' of the i9ios, the cotton market in
Bellary scarcely recovered the significance which it had had in the
American Civil War era and, in the long term, proved itself much less
profitable than the grain and, later, the groundnut markets. The
corollary to the move of cotton towards small farm production was
that it moved off large farm production. The Banking Enquiry (1929)
and the ICAR report noted that, while large farms still obviously
grew it, they could scarcely be said to specialize in it but produced it
as a limited part of mixes dominated by grain and supported by
groundnut.55 The Cotton Committee (I927) found large-scale (50

52 The
employment of permanent farm servants seemed closely related to the
number of cattle and ploughs kept and amount of wet cultivation undertaken.
MPBCE, V, pp. 301, 330, 332.
53This was rated across all Bellary farming at 5 days of family labour for every 2 I
days hired. But in purely 'dry' farming it was considerably lower-5.2 days of family
labour for 8. I hired. ICAR, IV, pp. I 7, 66-7.
54 G.O. 3628 (Revenue) dated 30 November i909, Tamilnadu Archives.
55The three villages examined by the Banking Enquiry were on excellent cotton
soil and very close to Bellary town. Nonetheless only between 20 and 30 per cent of
their acreages were under cotton. MPBCE, V, pp. 273, 296, 323.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA 1870-I-930 I43
acres +) production to be responsible for only 2 per cent of the total.56
The shift away from cotton and towards grain and groundnut
cultivation, however, had drastic implications for the employment of
labour. Cotton was an extremely labour intensive crop: the ICAR put
labour costs at close to 30 per cent of total production costs.57 Grain
and groundnut, by contrast, were labour extensive crops (the latter
being capital intensive) whose demands for labour were low. The
ICAR recorded labour inputs on Bellary large farms at little more
than 13 work-days per acre per year, the lowest in all the areas of
British India that it examined.58At this level of utilization, permanent
farm servants became a very expensive luxury.
How expensive may be seen from data collected by the Banking
Enquiry (1929), which reckoned the annual cost of a permanent farm
servant as between Rs 75 and Rs 90 a year.59Daily wage rates at the
time were thought to have been about 3 annas for an adult male,
making the actual wage-time paid to permanent farm servants the
equivalent of 399 to 480 days per year.60Moreover, the active agri-
cultural season lasted only about 8 months, making the effective work-
time paid for the equivalent of a 598 to 720 day year.61Even if the
permanent farm servant's work-time included that of his wife and
children (paid casually at half the male rate), the value looks very
questionable.
In effect, then, the case would seem much stronger that the expan-
sion of small farming in Bellary was the corollary to changes in labour
practices on the larger farms, and the break-up of the old magnate-
centred production system, than that it was 'induced' by the prospects
of entrepreneurial profit. Casualized labour had to find alternative
means of subsistence for the times when it was no longer being paid
and, in the absence of alternatives, found it in small farming. That
this was, in no sense, a preferred alternative may be judged from
behaviour during 'the golden age', when a price rise temporarily
revived the cotton market and the demand for labour and when
labourers, immediately, ceased taking up small farms.
56 CottonCommittee, p. 5.I
7 Calculated after 'disallowing' for rent. ICAR, IV, p. 21.
58
Ibid., pp. 66-7.
59MPBCE, V, pp. 324, 332-
60
Of course, local wage rates varied greatly. This figure is based on the 'com-
monest' rate found for male labour in the late i920S. QuinquennialWage Censuses
(I926).
61 The rains came in June or July and the last harvests in the black-soil areas took
place in March.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
144 DAVID WASHBROOK

It may also be judged from strong evidence of resistance to the new

economic logic, which occurred at times of famine. In both I876-78
and 1896-97, magnate farmers found themselves castigated for their
grain-selling activities and facing prospective grain riots as their
clients and employees refused to allow sales from their grain-pits.62
Indeed, it may have been in response to their vulnerability as targets
of public opprobrium that magnates decided to run down their grain
pits on a much greater scale after the I896-97 famine, so that they
would possess no objects which could become the focus of hostility.
That, in spite of protest and resistance, this run-down should,
nonetheless, have taken place, and the new economic logic been
imposed, may be seen as the result of three factors.
First, the appalling death-toll of the i876-78 Great Famine broke
the back both of the old economic system and of much potential
resistance to the new. With a third of the population dead in the most
commercialized taluks, the break with the past was almost physical
and, as recovery took place in the I88os, few of the survivors were in a
position to threaten the magnate capital which they needed to restore
their own livelihoods. Further, the 88os revealed the new system in a
particularly kindly light. A series of good seasons kept grain prices
relatively low (and thus wages high) and cotton cropping had not, as
yet, lost its profitability. By the time that bad times returned in the
I89os, it was too late to resuscitate the past.
Second, the later nineteenth century witnessed, parallel to the
deepening penetration of market forces, the deepening penetration of
the state. Magnate farmers could rely increasingly on the effective
support of the police to sustain their 'private property' rights in the
face of local resistance: and, thus, depended less upon deferring to the
demands, material and moral, of their local communities.63And third,
and relatedly, after the Great Famine, new famine codes passed the
major responsibility for the reproduction of the local labour force in
times of trouble onto the state. Famine was not again to bring the
direct loss of life, which had taken place in 1876-78, as disemployed
labourers and indigent small peasants could find some kind of succour
at new government relief camps.64Magnate farmers were then left to
enjoy the full profits which came from grain trading at famine prices.

62
Arnold, 'Famine'; also, my Emergence of ProvincialPolitics,ch. 2.
63
My Emergence of ProvincialPolitics, chs 2, 3; Baker, A Rural Economy,ch. 5; D.
Arnold, PolicePowerand ColonialRule,Madrass859-1947 (Delhi, I987).
64See McAlpin, Subjectto Famine.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA 1870-1930 I45

III

But if, as suggested above, the expansion of small farming followed

from the 'expulsion' and casualization of large-farm labour, and
represented a new search for subsistence to off-set wage losses, why
should it have concentrated on cash crop production, particularly of
cotton, rather than direct food provision? In straightforward terms,
this 'decision' would seem to make no sense of a subsistence strategy.
It meant producing a crop whose relative value against grain halved
across this period. It also involved its producers in a three-sided
structure of risk: from the climate, from the oscillation of grain prices
and from the oscillation in cotton prices which, being internationally
determined, were scarcely calculable within Bellary itself. Seeking
subsistence through the cotton market offered the prospects of ending
up with very little of an inedible crop or piles of a crop that was
unsellable or returns on a sold crop which were too low to buy much
food.
If the production and market conditions of small farm agriculture
are examined, however, it must be doubted whether producers had
much choice in the matter for, given their specific factor endowment,
cotton gave much the best returns and was, in a real sense, the only
crop that they could afford to grow. In the first place, operating on
fewer than 10 acres, they could make very little dent on their food
requirements if they concentrated on grain production.
What the precise calculation of 'minimum subsistence needs'
should be remains a very controversial question.65However, if we take
what the (notoriously ungenerous) colonial state of the period
thought, and offered for famine relief work (I lb of grain and one anna
per day for adult males and I lb + 0.5 annas for women and working
children), the result, at the prices of the later I920S, comes to the
equivalent of 3285 Ibs of grain per year for a family of five.66At what
would seem to have been the average levels of production at the time
(see below), that would represent the produce of about o1 acres. But
grain was not costless to grow. Andrew Kolliner, working on the
evidence of the Banking Enquiry for Bellary district, has estimated the
minimal cost of grain production, excluding imputed charges for
65
McAlpin, for example, estimated basic grain needs at 460 lbs per person per
year, but this does not include allowances for other 'necessities'. McAlpin,
'Railroads'.
66 Francis, Bellary District, II, p. 139.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
146 DAVID WASHBROOK

family labour, at Rs 5 per acre.67 At current market prices, Rs 5

represented the equivalent of IO6 lbs per acre, which would have
needed to be raised and sold to cover costs. That would have required
a further 3.5 acres of production which, with its own costs added,
would have required a further i acre. In effect, and as many con-
temporaries noted, it would have taken about 15 acres of average
quality land for a family of five to produce directly its own subsistence
needs annually.68 The problem was that the average size of farms
among the 70+ per cent of farmers who paid less than Rs 10 in land
revenue was about 7 acres.
As Robert correctly noted, necessary farm size could be cut very
considerably if cotton, even at the low prices of the i92os, were
produced.69Kolliner estimated the costs of cultivating cotton, under
the same conditions, at about Rs I2-15 per acre. But, for small
farmers, this would have to be raised to pay for bullock-hire, say to Rs
I4-I7.70 In the year of the Banking Enquiry (1929), gross returns to
cotton cultivation on the farms which it surveyed, where all produce
(kapas, seed, straw and interplanted korra) was sold, ranged between
Rs 28 and Rs 34.71 Small farmers, however, did not get the highest
returns and, for them, Rs 16 per acre would have been an excellent
profit. Nonetheless, on that basis, 3285 lbs of grain could have been
purchased off the 'profits' of 9.5 acres. For land-short peasants, cot-
ton's higher returns per acre provided a better chance of approaching
subsistence targets than did grain cultivation itself-even if, at 9.5
necessary acres, the majority of small farmers would still not have
been able quite to reach it. Of course, groundnut which, on Kolliner's
calculations, yielded 'profits' of Rs 25+ would have been better still
but, without cattle, this was uncultivable.72
67
Kolliner, 'Structure of Credit', p. 15.
68
See my 'Economic Stratification'; also MPBEC, V, p. 350.
69 Robert, 'Economic
Change', pp. 6o-i.
70
Kolliner, 'Structure of Credit', p. i5. The ICAR put the costs of cotton cultiva-
tion in the early I93os at Rs I2-13 per acre. ICAR, IV, p. 37.
71 Robert claimed that
Kolliner and the Banking Enquiry indicated net profits per
acre of Rs 30 for cotton, Rs I8 for groundnut and Rs 10 for cholum. But it is hard to
see how these figures are derived. The highest gross return to cotton acreage in the
Banking Enquiry's survey was just Rs 34, which, allowing even for low cultivation
costs, could not have yielded a net return of more than Rs 22 per acre. Kolliner's
actual conclusion was that net profits per farm averaged between Rs o and Rs 25 per
acre, depending on what was grown. Robert, 'Economic Change', pp. 63, 74-5;
Kolliner, 'Structure of Credit', p. 15.
72 Gross returns to
groundnut acreage varied between Rs 37 and Rs 45 against
costs of production, recalculated by Kolliner, at about Rs I5-I8. MPBCE, V, pp.
275-333; Kolliner, 'Structure of Credit', p. 15. Robert's attempt, by manipulation of

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA 1870-I930 I47
In various other ways too, cotton suited the needs of small farmers.
It was highly drought resistant and thus promised some returns even
when grain crops failed. It also, as noted before, was highly labour
intensive, which meant that it favoured the one factor of production
which small farmers possessed in greater abundance than large
farmers--unpaid family labour. Indeed, had Kolliner imputed labour
costs to cotton production (as large farm management would require),
the profitability of the crop would have become very seriously redu-
ced. At the ratio of labour costs seen by the ICAR survey, cotton
production with hired labour would then appear to have cost Rs i6-
20 per acre and 'profits' to have dropped to Rs 8-18 (say Rs 12). At a
cost: profit ratio of just about I o.8 cotton would have been the least
profitable of any of the three major crops of the region: the returns to
labour-extensive grain and groundnut production, at Kolliner's
figures, give possibilities of ratios of I: 1.4 and I: I.5, respectively.73
These ratios suggest very clearly why large farmers, for whom land
scarcity was not a significant issue and who farmed for returns on
capital, were withdrawing from cotton production. And also, they
indicate that the principal 'advantage' which small farmers possessed,
and which made cotton production still 'profitable' for them in the
depressed conditions of the I92os cotton market, was cheap family
labour.

IV

But 'advantages' and 'profits' for whom? It can hardly be held as of

much meaningful benefit to small farmers, themselves, that their
appearance of market competitiveness should depend upon their
receiving nothing for their labour. Moreover, as soon as the question
of retained 'profits' and even 'income' is raised, a large schedule of
other and hidden costs against small farm production comes to light.

price and acreage statistics, to demonstrate 'small farm' cropping choice as including
groundnut is extremely curious. Besides the problem of production conditions, there
is also one of location. Before the mid-g92os, 80 per cent of the cotton and the
groundnut crops were produced in different (black-soil and red-soil) taluks: if farmers
did make price-rational decisions about choosing between them, their farms must
have been spread over dozens of miles! Robert, 'Economic Change', pp. 74-5.
73 The ICAR found a 'business' income of Rs
7 per acre on cotton against produc-
tion costs of Rs 13. Although cotton prices were lower in the mid-I93os than in the
late I920S. ICAR, IV, pp. io6-7. Cholum returned Rs I-14 per acre against costs of
Rs 5-8 per acre. MPBCE, V, pp. 275-333; Kolliner, 'Structure of Credit', p. 15.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
I48 DAVID WASHBROOK

TABLE 2
Distributionof Debtby RevenueValueof Land

Pattas % Debt % Revenue Value

Rso-io 34 26.5
Rs o-30 34 36.5
Rs3o-50 13 15
Rs5o+ I9 23

Source:DerivedfromRobert,'EconomicChange',p.
69, and 'Statementof the Rebt-roll'in LandRevene
Report1925/26.

Some of these we have seen already-in, for example, the need to hire
cattle. Others, however, come as soon as we consider how small
peasants may have gained access to land and have financed their
cultivation of it. Although Bellary land-values were very low (in the
I920S, the equivalent of 1.5-2 years' gross yields), if land was bought
these still had to be met.74If land was rented, as some 30 per cent of
inam land (although very little ryotwari land) was, there were heavy
additional charges to meet. According to the Banking Enquiry (1929),
the standard rental rate in the villages which it examined was Rs 5 per
acre or five-times the equivalent land revenue charge.75 Even if
'virgin' land were cleared, the necessity of digging out deep-rooted
nath grass with a heavy iron plough and bullock team made its costs
of reclamation by no means negligible.
Then there were the costs of farming itself, which could only be met
off future income at the end of the harvest. Cotton farming had several
unavoidable cash expenditures-for seed and bullocks-and, in addi-
tion, there was the family's subsistence through the growing season,
which casual labour on large farms was likely to meet only in part. In
effect, small farmers needed credit in order to cultivate at all and, as
Table 2 demonstrates, their cultivation was more heavily burdened
with debt, in relation to the revenue-based asset value of their land,
than that of larger farmers.76The figures reflected in this Table come
from the Banking Enquiry (1929), which estimated the general level
of debt on Ceded Districts' farming at Rs 23 per acre.77On this basis
74 Between 1926 and
1930, black-soil land averaged about Rs 53 per acre, which
was down by about 35 per cent on values during the boom years of the First World
War. Government of Madras.StatisticalAtlas of the MadrasPresidency(Madras, 1936), p.
337- 75E.g., MPBCE, V, p. 300.
76 Due to variations in land
fertility, it seems preferable to quote revenue asset
values (which were, albeit loosely, related to fertility) than simple acreages.
77
MPBCE, I, p. 76.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA 1870-I930 I49

(which is almost certainly an underestimate since it overlooked petty

hand and grain loans), the debt per acre on small farming was over Rs
30 per acre or about the equivalent of the gross value of the cotton
crop. Such credit, of course, had to be paid for-and heavily.
Robert claimed that 'most loans were short-term production and/or
consumption credits which carried interest rates at 9 to 12 per cent.'78
This is a remarkable summary of the report which Robert claimed as
his authority but which actually found:
A good firstmortgagecan ordinarilybe got anywhereat 12 per cent but for
unsecuredloans and for large doubtfulmortgages,15 to I8 per cent are
commonrates,while 24 per cent, particularlyin the CededDistricts,where
crops are precariousand credit proportionatelylow, is not an uncommon
rate.79
Admittedly, a superficial reading of the Banking Enquiry, and of
other credit surveys of the 1920S, may give the impression that credit
was not a serious problem in the region. These surveys concentrated
in the main on the situation of the larger farmers, whom colonial
officials considered, anachronistically, to be the principal commercial
producers, and who had access to a variety of different credit sources
and at moderate rates of interest.80The rather different situation of
the mass of the poorer peasantry received much less attention. But it
is hard to disagree with Kolliner that, when teased out from the data,
what these reports reveal is that 'the wealthy ryot looking to take a
large loan and able to offer his land as security was in a far better
bargaining position than the cultivator who existed on the margin'.81
Without much security, he found it difficult to get loans; had to pay
the highest rates for them (of 25 and sometimes 50 per cent); and, not
infrequently, had to perform labour and other services as part of the
arrangement.82
Indebted up to and sometimes over, the total value of his crops, and
paying the heaviest interest charges of all, it must be seriously doub-
ted whether many small farmers made any meaningful 'profits' out of
their cultivation. Interest payments at 24 per cent on a principal of Rs
30, would have come to Rs 7.2 per acre of cotton cultivation and
halved the notional profit figure calculated by Kolliner (and hence
would have doubled the effective acreage needed to 'buy' subsistence

78Robert, 'Economic Change', p. 68.

79
MPBCE, I, p. 82.
80
Kolliner, 'Structure of Credit', pp. 53-5.
81
Ibid., p. 53.
82 Ibid., pp. 54-6; my 'Economic Stratification'.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
I50 DAVID WASHBROOK

to close to 20 acres). But perhaps making 'profit' was not the point for,
looked at another way, what the new economic system did permit
small farmers to do was to live on, and by, credit-which was itself a
valuable source of subsistence.
Credit advances represented consumption in the present against
costs in the future. Should crops fail and the small peasant debtor
prove unable to repay his loans, he had at least eaten that loan in the
first place; and his creditor was unlikely to have much recall against
him afterwards. At the land values current in Bellary, it was hardly
worth going through the procedures of repossession, if, indeed, the
land belonged to the debtor and was not rented in the first place. The
best way in which a creditor stood to get anything back from his
defaulting debtor was, in fact, to advance him more credit for the next
season in the hope that its crops might be better.
In the precarious circumstances of the Ceded Districts, being a
'consumption debtor' was by no means a bad option for labourers-
cum-small-farmers to pursue. It enabled them to re-form the connec-
tions of a shared subsistence with significant economic actors, which
the break-up of the old production regime had severed. One thing
which evidence from such people to the banking and credit surveys of
the period makes clear is that, for them, there was no problem of
'indebtedness' as such: rather the problem was expressed as one of
'credit' and of their difficulties in getting hold of it in sufficient quanti-
ties to be able to immerse themselves in the security of'debt'.83
Their need for credit may be seen as another factor pushing them
towards 'independent' farming and cotton production. Land and cash
crops represented some security for loans and the higher the per acre
value of the crops, the higher was the volume of credit likely to be
available to the producer. That the potential profit of these crops
might be absorbed in interest charges, made all the heavier by the
extra costs of production that had to be borrowed in order to produce
them, mattered little when the principal strategy was simply to get
hold of credit adequate for subsistence in the first place. Small
peasants, in effect, reproduced themselves and their cultivation, from
day-to-day and season-to-season, through the credit system.
And, at the end of all this, were they any better off than they had
been under the old production system? It would, perhaps, be easier to
see their situation as 'different' rather than 'better' in any qualitative
sense. As clients and dependants in the old magnate-centred
83
MPBCE, II, p. 297; IV, p. 74. See also, V. V. Sayana, The Agrarian Problems of
Madras Province (Madras, I949), espec. pp. 15I-7.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA 1870-1930 151

economy, they had had few opportunities of significantly 'improving

their lot' and had had to share in the deadly risks of famine borne by
all semi-isolated local production regimes in the pre-railway age.
There had been bad times, but also some noticeably good ones. The
'share' basis of customary payment schemes meant bonanza harvests
and boom profits were distributed (albeit very differentially) through
society. There were moments of plenty to help off-set those of desper-
ate scarcity. Indeed, and perhaps ironically, in periods after a famine
had been survived, conditions could temporarily become exception-
ally good. With labour scarce and valuable, employers and patrons
were unusually supportive of its welfare: in the I88os, the Bellary
population recovered nearly 60 per cent of its Great Famine losses.84
Equally, while the old production regime had necessitated certain
back-breaking forms of labour, and the exploitation of female and
child labour, it had provided long months of non-labour and alterna-
tive means for women and children to earn their millet-porridge out-
side the fields. Millet and cotton roughly shared the same eight-month
growing cycle, leaving the other four months free; and, before the
invasion of Indian markets by industrially-spun yarn, the hand-spin-
ning of cotton thread had provided an important by-employment for
all classes of the rural poor.
The new production regime, supported by the new famine codes,
certainly offered better protection from the extreme vagaries of the
climate. Direct loss of life due to sudden starvation was never to be the
same problem again. However, famine relief works supported life only
at the most minimal levels of subsistence and labourers-cum-poor-
peasants, who were obliged to return to them again and again-in
1892, I896, I900-I, 1917-19, 1922-24-were scarcely likely ever to
become 'sleek' and to rise above their allotted station in life.
Moreover, at famine relief camps, they faced death by other means
rather than escaped it altogether. As Elizabeth Whitcombe has
argued, famine camps were notorious centres of disease and may have
killed with microbes as many lives as they saved with food.85One of
the major reasons for Bellary's demographic stagnation, which saw its
I88os recovery from the Great Famine peter out and remain incom-
plete fifty years later, was the recurrent outbreak of plague, which
frequently centred on the famine relief camps. Instead of a roller-
84 Censusof India, I88I, Madras, vol. I (Madras, I883), p. 227; i89g, vol. XIII
(Madras, I893), p. 2.
85 E. Whitcombe, 'Disease and
Mortality in Indian Famines', presentation at
workshop on Famine in India, SOAS, October 1989.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
152 DAVID WASHBROOK

coaster ride between starvation and glut, the Bellary poor now faced a
long death of attrition between a more constant but inadequate sup-
ply of food and murderous diseases.
Further, to sustain even this supply required greater and harder
labour than ever before. Put most simply, with a labour force of the
same size, the district was cultivating 20 per cent more land and 60
per cent more labour-intensive cotton in the late I920o than it had
done in the late I88os. It was also doing so at lower levels of profit (at
least to cotton) and at no increase in piece-related rates of wage. The
process was accomplished, via the casualization of wage labour, in
part by lengthening the working year and, in part, by tapping new
sources of unpaid family labour. One of the 'advantages' of groundnut
cultivation was that its work schedule made demands for labour at
previously slack times of the year, increasing the effective work-load of
agriculture.86The shift of cotton into small farm production, as we
have seen, was principally made possible by the harnessing to it of the
unpaid family labour 'stored' inside the peasant family. In fact, cotton
made especially heavy demands for female and child labour, whose
'nimble fingers' were essential to successful picking.
In spite of the extra burden of work, however, there is precious little
evidence that labourers-cum-small-farmers ever did better than to
achieve a precarious subsistence-and that only with regular 'help'
from the famine department. Their small scale of debt-loaded produc-
tion and casual earnings from wage labour can have left them with
little surplus in good years and, come bad years, their trailing in large
numbers to risk the diseases of the relief camps hardly bespeaks much
in the way of 'progress'. A life that had once been a gamble on the
monsoon became one dependent on the whims of the market-place
and the medical profession; and, over nearly fifty years, the greater
'securities' of the market-place failed to provide it with the means to
ensure even its own regular reproduction.
But if the new regime brought few clear benefits to labourers-cum-
small-peasants, who did gain from it? Most obviously, the purchasing
companies and the ultimate consumers of Bellary cotton, grain and
groundnut obtained these commodities at lower cost-of-production
prices, certainly in terms of labour. But perhaps the greatest benefici-
aries were the local magnate farmers themselves. They increased the
profitability of their farming by sloughing off labour's costs of

86 Groundnut was harvested in November and

December, which previously had
been a lull in the agricultural season.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA I870-I930 I53

reproduction onto the state; by cutting their wage bills; by speculating

on rising grain markets; and by adopting groundnut cultivation.
They were also in a position to gain directly from small-farm pro-
duction. In part, these gains came from renting out scarce factors of
production to small farmers: land at five-times the revenue rate; bull-
ocks at Rs I per acre of work. But, in perhaps most part, they came
from the implications of the credit system. Kolliner has estimated that
the volume of credit in Ceded Districts' agriculture may have quadru-
pled between the late I88os and late i92os-alongside the expansion
of cash cropping.87Robert argued that this credit was supplied by
many sources: purchasing houses and banks, Komatis (the main pro-
fessional moneylending caste of the region) and farmers with sur-
pluses to spare.88
Of these, however, there cannot be much doubt that the latter
greatly predominated. The Cotton Commission explicitly noted that,
in the depressed condition of the market for 'Northerns and
Westerns', purchasing houses made very few advances against the
crop.89Groundnut purchasing houses certainly did make advances
but, as this was a large-farmercrop, only to large farmers. Banks, too,
concentrated on large-farmer clients who, alone, could provide the
kind of security which they required. Komatis, in fact, were very
scarce in Bellary and grouped only around the main towns: in 1921
they numbered scarcely 7,000 or 0.075 per cent of the population.90
The great bulk, perhaps go per cent, of credit was supplied by farmers
themselves.
And, among farmers, who had the greatest surpluses to lend?
Robert argued that the evidence of the credit surveys of the I920S
showed lending to be a promiscuous activity among all sections of
farming society and not to be concentrated in the hands of the
magnates.91Of course, there was indeed much promiscuous lending-
and lending on of money itself borrowed. However, Table 3, largely
taken from Robert's own evidence, hardly reveals a lack of concentra-
tion. The first three columns are given as by Robert to show that all
farmer classes participated in the credit market. In column 4, I add a
series which Robert omitted to offer, showing the proportion of
revenue payers represented by each of the revenue-paying categories.

87
Kolliner, 'Structure of Credit', pp. 24-41,
88Robert, 'Economic Change', p. 68.
89 Cotton
Committee, pp. 24, 29.
Censusof India, 1921, vol. XIV, pt 2, p. 120.
9 Robert, 'Economic Change', pp. 69-70.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
I54 DAVID WASHBROOK

TABLE 3
Distributionof Creditors
by RevenuePayingCategory

Pattadars % Creditors % Loaned % Pattadars

Rso-io 34 i8 72.0
RsIo-30 32 22 20.0
Rs30-50 I4 15 5-?
Rs50+ 20 45 3.0

Source:Robert,'EconomicChange',p. 70;and 'Statementof the Rent-roll',LandRevenue

Report
1925/26.

As can be seen, the addition of column 4 alters the entire meaning of

the Table: it reveals that 45 per cent of rurally generated credit was
provided by a group comprising less than 3 per cent of the landowning
population. If this does not represent significant 'concentration', what
does?
Given the expansion of the credit system, it also represented an
enormous aggregate investment. Kolliner, following the credit surveys
of the period, estimated the debt on Bellary agriculture at between Rs
2.9 and Rs 4.5 crores.92Taking just the lower figure, and allowing 90
per cent of it to have been farmer credit, the 45 per cent of it held by
the 3 per cent of richer farmers comes to Rs 1.2 crores: Rs 12,000,000
held by some 4,00ooopattadars at an average of Rs 3,000 each.
Moreover, it seems highly probable that most of this lending was
directed at the smaller peasantry. As Kolliner's analysis suggested,
the principles ofintra-rural lending depended more on 'personal' than
formal 'institutional' relations.93 The major problem facing small
farmers was that, in institutional terms, they possessed little security.
Who was better placed to develop personal ties with them than the
employers who provided them with their principal sources of off-farm
work and could off-set debts against notional wage payments? For
'medium' farmers, who employed more labour from their own families
than they hired, small-farm debtors were a much greater risk.
Indeed, viewed in this way, the entire shift of cotton production
from large to small farms can be seen as a mechanism whereby,
through the application of usury and 'service' capital, magnate-
creditors sought to respond to the conditions of depression in the
cotton market and to continue to squeeze a healthy profit out of the
crop. By acting as its major financiers and advancing it the factors of
92Kolliner, 'Structure of
93
Credit', pp. 25-6.
Ibid., p. 55.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA I870-1930 I55

production which it lacked, magnate farmers were able to draw

returns from small farming's one supposed advantage-unpaid family
labour. The family now laboured longer and harder and passed most
of the profits of its work to the magnates in interest payments and
rents. Not only did the new economic system 'rationalize' the deploy-
ment of labour, most critically it cheapened it-in this case, literally,
to the price of nothing.

But yet there may still have been a price to pay for capitalism's new
efficiencies and rationalities. As Christopher Baker has argued, dur-
ing the 1940Sand 195os evidence began to accrue of a general decline
in levels of agricultural productivity and fertility across the whole of
South India.94In many ways, this decline could be associated with the
rapid expansion of cultivation, which had taken place over the
previous seventy-five years. In Bellary, the decline would seem to
have started earlier and to have been marked even by the i920s.
During the first great cotton boom in Bellary, in the I86os, average
per acre outturns were reckoned to be in the region of 375 lbs ofkapas,
making about 93.75 lbs of lint.95This was, admittedly, an impression-
istic figure and can only have related to exceptionally good seasons.
Nonetheless, as a 'best season' possibility, it compared vary favour-
ably with other levels of cotton production found in other parts of the
South at this time. This favourability was also reflected in the fact that
Bellary town was chosen as the site of the first spinning factory in the
South.96
By the I920S, such levels of production (and the spinning factory
which closed down in 1915) were but golden memories. The Season
and Crop Reports had slimmed down Bellary's notional cotton yields
to 50 lbs of lint per acre (200 lbs of kapas). But, when the changing
'seasonal factor' is read against this figure, it appears that in not a
single season of the decade was it actually reached: outturn varied
between 52 and 94 per cent of the norm and averaged 75.8 per cent, or
151.2 Ibs of kapas/37.8 lbs of lint per acre.97Even this figure would
have astounded the Cotton Committee (1927), which claimed to find,
94 Baker, Rural Economy,pp. 227-8, 509-13.
95Kelsall, Manual,p. 262.
96 MadrasDistrictGazetteer,BellaryDistrict(Supplement), p. 68.
97 Seasonand CropReports,annual.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
I56 DAVID WASHBROOK

on the Bellary farms which it examined, cotton yields averaging just

86 lbs of kapas per acre-the lowest, by a considerable margin, of any
it found anywhere in British India.98However, it should be said that
the season of its survey was the worst in the decade for cotton produc-
tion: the Season and Crop Reports estimated outturn at just 52 per
cent of the notional norm, or 104 lbs of kapas per acre.99
The detailed village and farm surveys of the Banking Enquiry and
of the ICAR report on production confirm this dismal picture. The
Banking Enquiry, which operated during a cotton season with a rat-
ing of 93 per cent, noted yields varying between 150 and 220 lbs per
acre on the sixteen Bellary farms which it examined.'0? Close crop-
monitoring procedures in three Bellary villages between 1933 and
I936, revealed cotton outturns which varied between 27.2 and 308.8
lbs per acre and which averaged I74.4 lbs in I933/34; 99.7 lbs in
I934/35; and I70.4 in 1935/36.101 As 1933/34 and I935/36 were the
two best-rated seasons in Bellary since the First World War (at 99 per
cent and 96 per cent of norm respectively), these levels of production
suggest a near halving of yields since the i86os.102
Something of a similar case can be made for yields of cholum/jowar,
the principal millet crop, although the fall was rather less steep. What
is, admittedly, only impressionistic evidence from the second half of
the nineteenth century, confidently expected cholum yields, in what
must have been good seasons, to average 5 imperial maunds (4I 7 Ibs)
per acre.'03The Season and Crop Reports throughout the I920s kept a
figure close to this (450 lbs) as the anticipated norm. Once again,
however, the annual seasonal ratings indicate that it was never actu-
ally reached in the whole of the decade. Outturns varied between 61
and 96 per cent and averaged 80.8 per cent or 363.6 lbs per acre.'04In
the villages surveyed by the Banking Enquiry, during a season with a
cholum rating of 93 per cent, outturns varied between 192 and 320 lbs
per acre.'05 The ICAR noted, in 1934-35, district-wide average yields
of 333 lbs.'06
98 Cotton
p. 51.
Committee,
99 Season and CropReports, 1926/27.
100The yields are given in Bellary 'country' maunds of c. 26 Ibs. MPBCE, V, pp.
272-335.
'o0 ICAR, IV, pp. I64-73.
102 and Guha has
Season CropReports,1933/34-35/36. noted similar problems in the
Western Deccan. Guha, Agrarian Economy,pp. I Io-12.
103 C. Benson, An Account
of theKurnoolDistrict(Madras, I889), p. 25.
104Seasonand CropReports,annual.
105The yields are given in Bellary 'kadavas' of c. 63 lbs. MPBEC, V, pp. 272-335.
106ICAR, IV, p. I I1.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA I870-1930 I57

Against these statistics of decline and stagnation, of course, should

be set those relating to groundnut, which rose from virtually nothing
before the First World War to cover close to 15 per cent of the cropped
acreage by the early I930S. Where it displaced cholum, it offered a
crop capable of generating cash returns three times higher per acre. It
also survived the dry conditions of Ceded Districts' agriculture par-
ticularly well (three times, in the decade after 192I, its seasonal rating
topped 0oo per cent and it averaged 91 per cent of a norm of 1, 20 lbs
per acre).107 Further, it was a leguminous crop and thus soil-
replenishing in its effects. In many ways, the coming of groundnut
saved the Ceded Districts' economy although, and very particularly,
saved it only for the larger farmers. In the Banking Enquiry survey,
only one small farmer was found to produce it-and he was a cattle
owner and leaser-out who happened to farm on the side.108In the
ICAR survey, it was a noticeable absentee from the cropping mixes of
the few smaller farms examined.109
In seeking explanations of this decline, an immediate cause might
be found in deteriorating rates of productive investment in agri-
culture. Well-irrigation, for example, declined over the period:
whereas about I8,ooo acres of cultivation were well-watered in the
I89os, the number had fallen to barely 9,000 by the early 930os.110
Perhaps more seriously, cattle and plough to acreage ratios also
deteriorated: from about I plough per 22 acres of cultivation and one
bullock-pair per o acres in the late I88os to I: 30 and I : 12 by the late
1920S.lll Cattle became extremely sparsely used in Bellary agri-
culture. The ICAR noted average rates of work of only 4.9 days per
acre--among the lowest in British India and about I/8th of those used
on comparable farms in Coimbatore district in the much more suc-
cessful South Indian cotton belt in Tamilnadu.112 Further, as both the
Cotton Commission and Banking Enquiry reported, there was little
investment either in the improved varieties of'Hagari' cotton avail-
able after the First World War. Although these varieties offered better
107 Seasonand
CropReports,annual.
108MPBEC, V, p. 278.
109 The smallest farm
(I3.98 acres) in the ICAR survey grew no groundnut. The
two others below 25 acres grew it only once in the three years of the survey. ICAR, IV,
pp. 88-12I.
110Seasonand CropReports,I902/o3 and 1933/34.
"' In i890/9I, 85,000 ploughs and 209,000 bulls and bullocks were held to be
working 2. I million acres of cultivation; in I925/26, 80,ooo ploughs and 208,000 bulls
and bullocks were held to be working 2.4 million acres. AgriculturalStatistics189o/9g
and Season and Crop Reports 1925/26.
112ICAR,
IV, p. I4.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
158 DAVID WASHBROOK

yields and quality premiums at market, their take up was extremely

slow and, even by the late i920S, they covered barely a quarter of
Bellary's cotton acreage.13
Behind the deteriorating production conditions in Bellary agri-
culture, it is difficult not to see several factors directly related to the
nature of its economic 'expansion'. Magnate-centred production
regimes up to the i86os and I87os had concentrated production on
best quality inam lands, using high levels of animal inputs and allow-
ing for fallowing rotations which never put cotton on the same land
more than once in three years and which left fields unused 2 years in
7.114 Under such production conditions, even in the I920S and I93os,
very high yields were clearly possible. One Bellary magnate farm,
using the old methods of production and surveyed by the ICAR in the
early I930s, obtained cotton outturns of over 300 lbs in the best year
and grain yields of over 400 lbs in two years out of three.115But the
production conditions of the new small-farm agriculture were very
different.
In the first place, small-farm proliferation took agriculture towards
low productivity marginal land. Under the Bellary revenue system,
land was tax-rated according to estimates of its fertility in bands
varying from 4 annas to Rs 3-4-0 per acre. The massive expansion of
small farming was very much onto the lower qualities of land: in
1929/30, the average per acre assessment of land in pattas worth less
than Rs o1 was slightly under 8 annas (Rs 0.5) compared to an
average of about Rs i across the full scale of patta-holding.16
Equally, small farming had to cope with minimal animal inputs.
This brought two problems, which several agronomic experts of the
I920S and 1930S claimed to be acute in the district. First, there was a
tendency not to 'deep-turn' black soil as often as was optimal to limit
the progress of nath grass and to retain the fertility. Small farming
relied heavily on the 'guntaka', a kind of hoe, which, while adequate
for day-to-day operations was not a long-term replacement for the
plough."7 Second, of course, lack of cattle meant a shortage of manure
and 'undermanuring', particularly for soil-exhausting crops such as
cotton, was held directly to be a major cause of declining yields."8
113
Madras District Gazetteer,Bellary District (Supplement), pp. 68-9; Cotton Committee,
p. 51-
114 Kelsall,
Manual, pp. 262-7.
115
ICAR, IV, pp. I64-73.
116
Calculated from 'Statement of the Rent-roll' in Land RevenueReport 1925/26.
117 Francis, Bellary District, p. 85.
118
Ibid., p. 86; MPBEC, V, p. 272.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA I870--I930 I59

Such yields were further depressed by the inability of small farmers to

allow adequately for fallowing. To overcome the difficulty, they usu-
ally interplanted cotton with korra millets. This practice, however,
while partly protecting the land, caused cotton yields to drop drasti-
cally. It also was, in the end, no substitute for fallowing and eventu-
ally would still ruin the soil.
After furrowing and (under-)manuring his fields, the small farmer
then had to find seed. The reason generally given for the failure of
Hagari cotton to spread more widely was that small farmers could not
afford to pay the cash prices required for its seed; nor to retain any
seed from season to season since its sale (mainly to magnates as cattle
fodder) provided an important supplement to per acre earnings.19
Small-farming practices, then, tended both to put heavy pressures
on the soil and to be unable to utilize the new technological advances
of the period. This left large-farming to develop the forces of produc-
tion-or at least to sustain the productive base. To some extent it did,
through the introduction of Hagari cottons and groundnut and by
maintaining fallowing, deep-ploughing and manuring procedures.
However, there is little evidence of large farmers significantly increas-
ing their productive investments on the land in relation to the
improvements in profitability which they were enjoying in the grain
and groundnut markets. Dwindling cattle-acreage ratios indicate that
they were not extending their investments in cattle. Equally, the
decline of well-irrigation shows their unwillingness even to maintain,
let alone increase, investments in irrigation. Indeed, in many ways
they were 'withdrawing' from their once central role in agriculture-
producing less cotton, letting out lands, farming a progressively smal-
ler proportion of total cultivation.
But what, then, were they doing with their profits? There is some
evidence of their becoming involved in the marketing and crop-pro-
cessing industries: several magnates acquired cotton gins and ground-
nut decorticating machines.120However, and most obviously from the
statistics on the expansion of credit, they were investing in
moneylending-and on a huge scale. In the late I92OS,wells were
estimated to cost about Rs I,ooo to dig and agricultural costs of
production averaged around Rs 13 across all crops.'2' The Rs 1.2
crores of credit issued by the 3 per cent of large farmers in the district
represented the equivalent of I2,000 new wells or 900,000 acres of
119MadrasDistrict
Gazetteer,BellaryDistrict(Supplement), pp. 68-9.
120
See my 'Economic Stratification'.
121
MPBEC, III, p. 807.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
i6o DAVID WASHBROOK

cultivation. Plainly, the short-term 'profitability' of capital invested in

agriculture now turned more on its ability to exploit the subsistence
needs and unpaid family labour of the small peasant workforce than it
did on its ability to raise levels of production.

VI

But what caused this situation: why should the increasing penetration
of capital into production, in the circumstances of Bellary, have
resulted in ecological devastation and poor peasant exploitation?
Superficially, the history of the market and of a variety of geographi-
cal and ecological factors might seem to supply the answers.
Obviously, the declining market for Deccani short-stapled cotton
seriously reduced the profitability of cotton production, the principal
cash crop. Climatic uncertainties and water shortage made it imposs-
ible, at least under criteria of competitive profitability, for Bellary
farming to respond by taking up the new long-stapled and 'watered'
cottons that now dominated the market.
There was some response in terms of the development of groundnut
as a new major cash crop. However, Bellary's great distance from the
nearest cattle breeding grounds (in Nellore district), and lack of
adequate pasturage, limited the possibilities of this crop. Further,
neither groundnut nor grain demanded the same levels of labour
input as cotton. This left labour as Bellary's most abundant and
progressively cheapening factor of production, which capital came
most naturally to exploit.
Yet it is never satisfactory to treat society simply as the passive
victim of intangible forces of nature and the market. Clearly, different
patterns of human intervention could have brought about different
results. One such pattern can be traced to the activities of the state,
which, through acts of both omission and commission, bore a heavy
responsibility for these consequences. In terms of omission, of course,
the colonial state did remarkably little to find technological solutions
to Bellary's ecological and agronomic problems. As Christopher
Baker has seen, its obsessions with riverine irrigation meant that it left
the 'dry' uplands without adequate support or investment.'22Nor did
its concerns with crop improvements extend beyond crops with export
potential to the millet grains on which most of society lived.
122
Baker, Rural Economy, ch. 5.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA I870-1930 I6i
In terms of commission, however, the colonial state's responsibili-
ties are even more direct and open up questions on a second pattern of
human intervention-that represented by class relations. In many
ways, it was the actions, intended and unintended, of the early Com-
pany state that set up Bellary's social problems into the twentieth
century. Penal levels of revenue assessment, together with the careless
granting of inam rights which could be confiscated by a small elite,
fundamentally altered the distribution of wealth and power within the
district. As a new and more penetrative age of capitalism dawned in
the second half of the nineteenth century, Bellary entered it with most
of its potential sources of 'capital' (good quality land and supra-
subsistence surpluses) concentrated in the lands of a very small group
of landed magnates. Subsequent state intervention further helped
them to turn their wealth into capital. The new famine codes enabled
them to withdraw from responsibility for the reproduction of their
own labour forces. And more effective administrative and policing
systems helped to guarantee their 'private' rights of property against
the moral and material demands for 'shares' still emanating from
society. The state made the particular capitalist class which domin-
ated Bellary farming.
And having made it, and made it in a way which opened a vast gap
between the resources of the magnate elite, on one side, and the 70+
per cent of indigent labourers-cum-small-farmers, on the other, it
stood back to allow the logic of capital to work itself out. Hardly
surprisingly, that logic saw capital attaching itself to and exploiting
the very indigence of the labourer-cum-small-peasant, whose relative
share in a social product, expanded by the extension of cultivation
and cash-cropping, declined in proportion to capital's own advance.
And profitable though this pattern of exploitation may have been in
the short-term, its effect on the long-term development of the forces of
production was largely negative.
The wide variety of different histories of the market, ecology, state
intervention and class relations in colonial Indian farming, of course,
make it impossible to generalize directly from Bellary's experience to
that of other regions. But if this experience does have something to
tell, it must be of the importance of factors of'distribution' (and hence
of class) in determining the way in which the deepening penetration of
capitalism, from the later nineteenth century, affected the means and
relations of production. Prior distributions of rights and resources,
and the ability or inability to defend them, structured capital's pos-
sibilities of 'progress'. In Bellary, they gave capital a near-rightless

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
i62 DAVID WASHBROOK

and -resourceless mass of labour, broken further by the Great Famine,

which became its greatest, and ultimately almost only, source of pro-
fit. Had magnate success in seizing the grain heap, and selling it for
profit, been resisted more adequately, capital's trajectory of 'develop-
ment' must necessarily have been different.
Besides pointing to the importance of 'distribution' and class
'resistance', Bellary's experience, relatedly, can also be seen to pose
economic history a teasing set of questions, which, as yet, have been
inadequately addressed. These concern the rising price of grain,
which 'outperformed' the prices of both commercial export crops and
of manufactured goods. Grain, however, represents food and, immedi-
ately, opens up the issue of how far colonial India's economic growth
may generally have been retarded by a structure of prices which
raised the cost of subsistence faster than that of anything else.
Several other issues flow from this. It has been conventional to look
to export-orientated commercial crops as the principal source of 'pro-
fitability' in the economic development of the I880o-930 period. Yet,
in Bellary, it was the profitability of grain production which was in the
ascendant and which virtually doubled in relation to that of cotton
over these years. Through the operations of the grain: cotton price
ratio, large farmers, who produced most of the grain, can be seen to
have transferredto themselves much of the notional surplus earned by
consumption-deficit small peasant cotton producers. We may need to
examine much more closely intra-rural relations of production,
marketing and subsistence before assessing the 'benefits' of the com-
mercialization of the epoch.
And, indeed, there are questions concerning the reasons for the
massive hike in grain prices, particularly 'dry' grain prices, in this
period. As noted earlier, standard explanations have focused on 'the
transport revolution' and, clearly, this was an important element of
the context. However, it must be counted as, at least, passing strange
that, of all food grains, the highest rises in price should be found in the
markets for millet grains. These, being extremely bulky to their nutri-
tional content and having a low 'quality' preference, circulated the
least widely of any of the standard food grains. In fact, it is even quite
difficult to work out who, on a regular basis, would have been
involved in this market. Urban populations, in the Ceded Districts as
elsewhere, rarely consumed the crop since they could import a wide
variety of rices (premium for the rich and 'broken' for the poor) and
much preferred to do so. Outside times of dearth and famine it must

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
 AGRICULTURE IN COLONIAL INDIA I870-1930 163

be doubted that millets were much consumed outside their locales of

production.123
But perhaps that is the answer, or part of it. As discussed earlier,
millets were associated with climatically dangerous 'dry' zones in
which crop failure, in one locale or another, was a near certainty:
millet circulated largely in response to local dearths which spread
their high prices through its market. However, if this were all that
there were to the matter, it would not explain, of itself, why the
underlying floor price of millets should have risen continuously and
been sustained, as in the later I920S, through years when there were
no serious crop failures in the millet belt. Nor why annual oscillations
in price should noticeably have diminished.124
To function adequately, the model must presuppose the long term
and progressive structuring of 'dearth' conditions into the grain
market. But what could have caused this? McAlpin had argued
vigorously against the once-held notion that rising food prices reflec-
ted declining food production as acreages under commercial crops
displaced those under grain.125She pointed to the general expansion
of all cultivation as having kept food acreages in line with population.
In Bellary, where there was no population increase, this was certainly
the case: actual acreage under food crops remained about the same
through the 20 per cent expansion in total cultivation, which took
place between I890 and 1930. But what of yields? On the evidence we
have presented (which parallels that presented by Baker for
Tamilnadu), there is a strong prima facie case for a decline in grain
yields, which, if true, would certainly account for a steady rise in price
as a consequence of progressively deficient supplies.l26
And our evidence suggests a further possibility too. How far did
changes in the relations of labour, themselves, generate an increased
market demand for food stuffs and thus contribute to their price rise?
The evidence which we have seen of the casualization of wages, of the
shifting of the wage medium towards cash and of the proliferation of a
food-deficit small peasantry all suggest an increasing 'market'
123S.
Krishnaswami, who was concerned with food-deficit problems during World
War II, took the entire millet crop to be rurally, and locally, consumed. S. Krish-
naswami, RuralProblemsin Madras(Madras, 1947).
124
M. McAlpin, 'Price Movements and Economic Fluctuations' in D. Kumar
(ed.), CambridgeEconomicHistoryof India,II (Cambridge, 1983).
125
McAlpin, 'Railraods'.
126
Baker, RuralEconomy,pp. 227-8, 509-I3. See also, G. Blyn, AgriculturalTrendsin
India i891-1947 (Philadelphia, 1966).

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions
I64 DAVID WASHBROOK

demand for grain-and a demand inside the countryside itself-

which would help to explain the apparent lack of dry grain movement
along the railways. In effect, the class transformation of rural Bellary,
which we have been discussing, must be given a prime responsibility
for having raised food prices as part, and only one, of the strategy
whereby capital increased its profitability at the direct expense of
labour's subsistence, family relations, leisure-time and, ultimately,
even ability to reproduce its own life.

This content downloaded from 14.139.235.3 on Wed, 01 Apr 2015 08:52:44 UTC
All use subject to JSTOR Terms and Conditions