De-industrialisation in India Reconsidered

Author(s): Raghabendra Chattopadhyay

Reviewed work(s):
Source: Economic and Political Weekly, Vol. 10, No. 12 (Mar. 22, 1975), pp. 523+525-531
Published by: Economic and Political Weekly
Stable URL: http://www.jstor.org/stable/4536967 .
Accessed: 28/03/2012 11:24

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].

Economic and Political Weekly is collaborating with JSTOR to digitize, preserve and extend access to
Economic and Political Weekly.

http://www.jstor.org
 SPECIAL ARTICLES

in India
De-industrialisation Reconsidered
RaghabendraChattopadhyay
This article discusses the findings of Daniel Thorner and I Krishnamurthny on the question of
de-industrialisation in India during the early years of this century. The author also attempts an alter-
native measurement of changes in industrial employment and argues that the process of de-inrdustrialisa-
tion continued right upto 1931.
I of the Working Force in Manufactu- of capital intensive technique. But, "at
Introduction ring, 1905 to 1951: A Theoretical and present, data on output in non-factory
Empirical Analysis",8 challenged the establishments is obtained largely
MANY economic and social thinkers in thesis on theoretical and empirical through assumptions regarding output
India in late nineteenth century were
grounds. The aim of the present article per worker in this section. As a conse-
often critical about British economic is to show, first, that though Krishna- quence, neither the output nor the out-
policy in India. The prevailing opinion murthy argued from a wrong plane, put per worker figures can be depend-
among the Indian (and sometimes also Thomer did suffer from a faulty ed upon".''2 Krishnamurthy anyway
British) thinkers was that British policy
methodology and, secondly, to provide evolves a highly "indirect measure of
in India was resulting in gradually the reader with an alternative test of capital intensity". Using this technique
worsening the economic condition of the hypothesis of "de-industrialisation" he finds that " ... there was a shift to-
the 'natives'. William Digby' and in British India. wNardshigher capital intensity in Indian
Romesh Dutt2 furnished statistical in- manufacturing, though the size of the
formation to prove their contention.
manufacturing sector in India did not
This view was, of course, opposed by II change significantly"." But after so
most of the British officials in India.
But the issue is yet to be decided.3
Krishnamurthy and His Theory much ado, he comes to an odd judg-
Daniel Thorner discussed the case
an(I Empiricism ment "the de-industrialisation hypo-
thesis cannot really be tested",14 and
of "de-industrialisation" in India.4 Any discussion on "de-industrialisa-
to him "the use of working force
The purpose of his study was to evaluate tion" must define industrialisation. In
ratios to test this hypothesis is not a
the notion of de-industrialisation in Krishnamurthy's opinion, "Industrialisa-
solution to the problem".'5
India under British Rule during the tion means a rise in the share of manu-
last phase of 19th Century. Thorner facturing output per capita. This is As we suggested at the beginning
formalised, in the following manner, historically associated with a rise in the of this section, the problem is one
the opinion of economists, mostly Indian, share of manufacturing in the total of definition and - we may also add -
who argued that India had suffered working force. But a rise in the ratio of fidelity in the experience of econo-
from the process of de-industrialisation of workers in manufacturing to mic development. It is true that the
at the hand of the British Rulers: "It total workers is neither a necessary nor manufacturing working force data alone
is a three-fold proposition to the effect a sufficient condition for industrialisa- do not give us a total picture of the
that the decine of handicrafts continu- tion in any theoretical sense. By the level or degree of industrialisation. Any
ed well on into the twentieth century, same token, a decline in the share of concept of industrialisation should con-
that it was not compensated by a manufacturing in total output or a de- sider the level and rate of change in
sufficient rise of modern industry and cline in manufacturing output per output also. But what Krishnamurthy
that, in consequence, the Indian eco- capita, constitutes "de-industrialisation". overlooks is that the output data -
nomy became more and more agricul- A decline in the share of manufacturing like the working force data - alone
tural".5 To test this hypothesis, Thor- in total working force is by no means cannot give us any measure of (de-)in-
ner based himself on the information a necessary or a sufficient condition."9 dustrialisation.
available in the Census reports of With this definition or rather this "ne- Let us first try to understand "indus-
1881-1931. As an index he considered cessary and sufficient condition" of (de-) trialisation". The "necessary and suffi-
the male industrial working force as a industrialisation, Krishnamurthy argu^es cient condition" of Krishnamurthy is
proportion of the total working force against Thorner that, "to show that the per se nothing but a symbol denoting
and reached the conclusion that "the number or the proportion of working the ratio between the manufacturing
census data for males do not support force engaged in manufacturing has re- output and thbe total output. Here
the case either for absolute 'de-indus- mained roughly constant does not dis- Krishnamurthy suffers from a severe
trialisation' of the working force or prove the 'de-industrialisation'hypothe- eclecticism. Because, we can hardly
even for relative 'de-industrialisation'".6 sis".10But, "the hypothesis can only be understand anything from this ratio
As an after-thought he suggested that tested with data on output in manu- without considering the institutional-
"if indeed a major shift from Industry facturing and total output in the eco- structural framework of the economy.
to Agriculture ever occurred during nomy".1"Moreover, he suggests that it Apart from the usual index num-
the British Rule in India, it might have is also necessary to know the capital ber problem associated with the mea-
happened some time between 1815 and intensity in the different branches of surement of the change in such a ra-
188O."7 the industrial sector. A decline in the tio,'6 there also exists the problem of
Thorner's conclusion did not go un- magnitude of the working force is ownership of the output. When (or if)
challenged. J Krishnamurthyin his quite compatibl-e with a rise in output the manufacturing sector in a colonised
article, "Changes in the Composition in the presence of increasing adoption economy is in the hand of the impe-

528
ECONOMIC AND POLITICAL WEEKLY March 22, 1975

rialists then the output may be gene- as a concltusionto our discussion above and when it happened "Agriculture"
rated for the outside market. In such I suiggest that we can evolve the fol- showed a decline and "General La-
a situation any increase in output may lowing criteria for the study of chan- bour" an increase. Hence he suggest-
not increase the net national income. ges in an economy vis-a-vis industriali- ed that most of the workers defined as
The profit accrued from the sector may sation: (i) change in the NNP or Na- general labourers could not be agricul-
not be ploughed back into the economy. tional Income; (ii) change in the ratio tural workers only. He then, agglome-
The other remaining source of income of manufacturing output to the total rated "Trade" with "Manufacture",
generation within the economy in such output (or NNP) and (iii) change in the etc, and regarded this agglomeration as
a case is the increase in the number manufacturing working force with res- the Industrial Sector.28
of working force involved or an in- pect to total population. If all these Thomer had suspicion about the data
crease in the real wage of the working three factors increase over the years, returned in the census tables on female
force. The real wage (or even money we can strictly define a process of in- workers and hence he separated the
wage for that matter) may not rise in dustrialisation. Similarly, a decline in data on male workers from those on
such an economy due to the pressure all the three factors will strictly indi- females. "We shall be well advised to
of unemployment and lack of working cate a process of de-industrialisation. consider separately the data for males
class organisation. So we will have to On the other hand, a decline in the and for females in each census year
consider all these problems together second or the third factor above, toge- and to disregard completely the 1881
with the "necessary and sufficient con- ther with a constant national income (or figures for women".29
dition" of Krishnamurthy which is ob- NNP) will also indicate, a process of He then considered the male indus-
vioUsly not suifficient. de-industrialisation.23 trial working force' as a proportion of
The argument put forward above is As far as the output parts [(i) and the total working force and showed
not hypothetical. In fact the studies of (ii) above] are concerned the informa- that this ratio remained more or less
Gunnar Myrdal,'7 Celso Furtadol8 or tion available for India is scattered constant over the period 1881-1931.30
A G Frank19 amply describe the pro- and needs thorough research. But stu- The absolute number, as he formulat-
blems of "underdeveloped" couitries dies by Daniel Thorner on "long-term ed, also remained stationary. Hence,
where the emergence of "pockets" or trends in output",24 by S J Patel on "The conclusion forced upon us by
"enclaves" of industries like tea, rub- "Long-term changes in net output in the census occupational data is that the
ber, jute, coffee or minerals have re- India 1896-1960",25 and S Sivasubra- industrial distribution of the Indian
sulted in economic retardation of these manian26 show that the national income working force from 1881-1931 stood
countries. We may also note that at in India remained more or less stag- still",31 and there was no de-industria-
the time of their inception these indus- nant over the period of 1900-1940. So, lisation either in the absolute or in the
tries were more capital-intensive in if we can now show that the manufac- relative sense.
these economies than agriculture or turing working force declined over the My objection to Thorner is two-fold.
handicrafts. period then we have a strong case for First, Thorner wvas wrong in consider-
The theoretical (and partly also em- "de-industrialisation", unlike Krishna- ing the ratio between the industrial
pirical) problems connected with the murthy or Thorner. 'working force and the total working
definition or the criteria of industriali- III force as the -index of de-industrialisa-
sation have been thoroughly discussed Therner's Study and Its Critique tion. By taking the total working force
by Amiya Bagchi,20 and R B Sut- As stated in the beginning, the pur- as the denominator of the ratio, Thor-
cliffe.2l For us it is sufficient to note pose of the present article is to recon- ner did not take into account the un-
that we should connect the notion of tider the study of "de-industrialisation". employed population and the depen-
industrialisation with that of economic Let us first discuss, in brief, the me- dants. The correct method would be
development. Erection of a few facto- thodology Thorner followed. To study to consider the industrial working force
ries may not carry any meaning to the "de-industrialisation", Thorner consider- as a ratio of the total population.
economy if it does not lead to econo- ed a period of fifty years - 1881 to Even if the absolute number of the in-
mic progress. Historically the period of 1931. He studied the census reports dustrial working force does not rise
industrialisation in the now developed and tables published from 1881, the over the period, the ratio between the
countries was a period of rising manu- second All-India census, to 1931, the industrial working force and the total
facturing and total output with a ris- Leventh one. Thorner divided the whole population may decrease due to an in-
ing industrial working force which ab- working force into five large groups: crease in population over time. This
sorbed the surplus agricultural popula- (i) Agriculture, Factory and Fishing phenomenon should be considered at
tion. As we have seen above, this (ii) General Labour (iii) Manufacturing, least as a case of relative de-industriali-
is not only historically true but should Mining and Construction (iv) Trade and sation.32 The method is particularly re-
be accepted as the "necessary and suffi- (v) Transport and Other Services.27 levant when the purpose is to study
cient condition" even in a theoretical He included "General Labour" in economic development through indus-
sense. Sutcliffe,22 on the basis of em- the agricultural category as the census trialisation. The important fact in this
pirical study, derived some criteria for tables do not mention the character of context (as I purport to show in the
the measurement of the level of indus- labour in which the i workers in this next section of this article) is that the
trialisation. According to his study, the group were involved. Thorner observ- industrial working force, when taken as
level of industrialisation is showvn by ed that "Agriculture" and "General La- a proportion of the total population, did
(i) the level of industrial output as a bour" usually showed an inverse rela- continue to decline upto 1931.
proportion of the GDP, (ii) the level of tionship and this was due, in his opi- My second objection is regarding
manufacturing output with respect to nion, to incorrect tabulation of the Thorner's method of collation of data
the total industrial output, and (iii) the "General Labour" in the census opera- on the working force. Several objec-
level of the percentage of the popula- tions. Many of the agricultural popu- tions could be raised regarding the me-
tion involved in industrial activities. lation were supposed to have been in- thod of grouping of data by Thorner.
With the help of these criteria and dlude'd in the "<General Labsour" group But we shall limit ourselves to illustrat.

525
 March 22, 1975 ECONOMIC AND POLITICAL WEEKLY

ing only those which may have a direct census reports from 1881-1931, I find available.
bearing on his conclusion. We shall that although the census officers were Thorner considered a period of fifty
discuss only three points here. First, taking increasing care regarding tabu- years, 1881 to 1931. The method of
Thorner's handling of the data on Ge- lation, none of them were sceptical census tabulation regarding occupation
neral Labour is wrong. As we describ- about the female data for censuses other during this period changed from time
ed above, Thorner includes workers than the 1881 one. We do not have to time. The reliability did not also
under this heading in the agricultural any a priori ground (excepting for remain the same. Particularly, the
sector. B R Kalra, as a census officer 1881 and may be for 1891) to refuse earlier census operations suffered from
for Census 1961, calculated that about to accept the data on female workers the lack of experience of the enumiera-
20 per cent of the unspecified labour- returned by the censuses. Asi far as tors. It was only from 1901, the
ers were in the urban areas.33 J Krish- the enumeration practice is concerned, fourth census, that reliable data started
namurthy in his article "The Growth of both males and females might have to be returned. Thorner started from
Agricultural Labour in India - A been affected equally. Although his- 1881 and concluded that the deeline in
Note",34 was also of the opinion tllat torians may have to take resort to industrial occupation upto 1931, shown
"there is certainly an argument against thoughtful conjectures in the absence in the census data as published, is
this [lumping 'unspecified' components of evidence, it is better not to be too rnostly due to the highly "meretricious
with agricultural labour] as we do have assertive regarding the past and neg- 1881 data".39 The implicit assumption
evidence that a significant proportion lect concrete evidence when found. Re- seems to be that if we do not pay heed
of the unspecified belongs to urban garding the nature of female occupa- to the 1881 figures the censuses would
areas".35 Going through the census tion, Thorner seems to suffer from his tell us a different story. As Thorner
tables also we find that the agricultuiral Western view of economic activity. He did not have any faith in 1881 data -
labourers were separately listed and it puts emphasis on the word "economic" and he was perfectly justified in his
has been mentioned clearly that the by putting it into italics. But the role lack of faith - I have thought it bet-
general labourers who were tabulated of women in India, particularly in the ter to leave them out. As far as 1891
in the industrial caetgories were "un- small and household industry, cannot be data were concerned they were not only
skilled non-agricultural labourers" or minimised. In the census tables also unreliable but also non-comparable with
"unskilled labourers connected with we find that it was under these the later period census. Hence, we
earthen and other works". One gets the "orders" rather than in the large in- start from 1901. I shall not consider
idea that these workers might have dustries that comparatively large num- the period after 1931 because of the
been connected with road building, ber of women were returned. We also same reasons that Thorner had: 1941
railway, irrigation and other construc- know from the studies on Indian Cot- census did not return any relevant fi-
tions and services. ton Textile Industry by different scho- gures due tot war exigency and there
Secondly, the idea of including lars that activities like cotton spinning was such a change in the method of
"Trade" in the Industrial Sector is also and finishing of cotton cloth were often classification of information since 1951
done by female labour.37 Moreover, that it would be very difficult to com-
erroneous. This is particularly so when
the object is to studv the process of there is also a trend of decline of fe- pare the last three censuses (1951,
male participation in large-scale indus- 1961 and 1971) with the earlier ones.
(de-)industrialisation. Trading may not
try during the period of our concem.38 The constraints that we impose upon
necessarily be an indicator of the pro-
industrialisation. The rising On the basis of the above argument, I us no doubt narrow down the field of
cess of
numnber of present day Calcutta hawk- oppose Thomer's exclusion of female our operation to only four censuses, viz,
ers and petty traders may, on the other occupation from his consideration. 1901, 1911, 1921 and 1931. ranging over
hand, be taken as an indicator of the Thus far, I have put forward my ob- a period of three decades only. But we
acute crisis of unemployment in West jections regarding (i) Thomer's me- shall see that the trend, are sufficient-
Bengal and the adjoining provinces. thod of measurement and (ii) his me- ly distinct to help us reach some con-
Moreover, the inclusion of Trade in the thod of tabulation. I shall try to pre- clusions.
industrial sector also goes against the sent in the following section an alter- A brief comment on the census re-
almost universal classification of econo- native method of measurement and try port in general and the method of cen-
mic activities into primary, secondary to correct the mistakes of Thorner. This sus tabulation may not be out of con-
and tertiary occupations. We should will lead us to a conclusion different
text at this point of our study. Our
also keep in rnind that even the Clark- from Thorner's regarding the problem sources of information are the "Occu-
Kuznets type of foimulation does not of de-industrialisation in India upto pation" and the "Population" reports
apply in the case of underdeveloped 1931. and tables published in the different
countries. Here, lack of emPloyment census volumes.40 Regarding occupation
due to the "development of under- IV tables the first thing to note is that the
developmnent" may force a portion of the Alternative Methodology method of classification differed from
populationi to the trading occupation. Our point of criticism against Thor- time to time. "In 1891 all occupations
This can be hardly taken as a case of ner's method of measurement was re- were divided into seven main classes
increasing economic activity.36 So we garding his acceptance of the industrial as follows: A) Government, B) Pasture
conclude that Trade should be consi- working force only as a ratio of the and Agriculture, C) Personal Services,
dered separately. total working force. We have also D) The Preparation and Supply of Ma-
Thirdly, we do not find any reason noted why we should consider the terial Substances, E) Commerce, Trans-
not to take the (lata on female workers industrial working force as a propor- port and Storage, F) Profession and C)
into account. A glance at Thomer's tion of the total population and not of Indefinite Occupation and means of
article will show that the foot-notes the total working force only. But be- subsistence independent of occupation.
supporting his suspicion about female- fore going into it we should make the "Subordinate to the seven ab'ove
data illustrate only froam the Census of reader familiar with the nature of (and classes were 24 orders. These were
India, Report, 1881. Going through the thie problems regarding) the information further sub-divided into 77 suborders

526
ECONOMIC AND POLITICAL WEEKLY March 22, 1975

TABLE 1 VAPIATLON IN POPULATION epidemic of 1899 to 1900 took a heavy dants' and 'As Subsidiary to Other Oc-
DURING 50 YEARs: 1881-1931 toll of the Indian population, particu- cupations'. The column 'Actual Total
larly in the rural areas."4The decline Workers'(male and female) include those
1881-1891 +33,418,341
1891-1901 + 7,046,385 in the growth of population can be ob- who were 'partially agriculturists'.These
served from Table 1.451 This decline were also shown separately. Now as ag-
1901-1911 + 20,795,340 might have affected the industrial work- riculture was assumed to be the major
1911-1921 + 3,786,084
ing force also by decreasing their num- alternative source of employment to a
1921-1931 + 33,895,296 ber. In 1918 a virulent influenza epide- person employed in a non-agricultural
mic broke out in Western India. The activity so the term "partially agricul-
epidemic invaded India twice in a very turists" was used in a weak sense to in-
TABLE 2: POPULATION OF INDIA,
1901-1931 short span of time and spread all over clude all persons who accepted any other
India and practically levelled off the activity as a subsidiary occupation.
1901 294,361,056 natural growth of population of the Hence by adding up the figures in the
1911 315,156,396 seven previous years.46 This can be column "As Principal Occupation" and
1921 318,942,480 also noticed from Table 1. The po- the corresponding figures in the column
1931 352,837,778 pulation figures from 1901 to 1931 are "As Working Dependants" in 1931, we
given in Table 2. arrive at the data which are comparable
and 478 groups. In the ensuing cen- The particular feature of this epide- with those of earlier years. The persons
suses the main division into classes, mic can be best discussed by quoting listed "As Working Dependants" in-
order and sub-orders described above, from the Census reports ". . . the first cluded those with any subsidiary occu-
remainedpractically unchanged. [In 1901 epidemic xvas most prevalent in urban pation.49
the number of classes was increased by areas but it was not of a specially vim-
one. The Class G of 1891 was divided To obtain the data on industrial work-
lent type and probably for that reason ing force, I take the data appearing
into two. G retained "unskilled labour it is said to have affected the young
not Agricultural", and a new class H under 'industry' for 1911, 1921 and
children and old people most severe- 1931. But I separate the 'Trade' fi-
was formed of "Means of Subsistence ly". In the second invasion "... the
Independent of Occupation."] In the gures from them. In 1901, the traders
rural areas were more severely affect- were not separated from the makers and
case of groups, however, although the ed . .., mortality was specially high
general arrangement was maintained, so I separate them according to their
among adults (20-40), particularly occupation. I separate all the sellers
there were many alterations in detail. among adult females".48 We should
Some of the old groups were amalga- from the "Industry" table in 1901. We
remember here, that World War I also have figures for 'makers' and
mated or transferred to other sub- did not affect the increase in popula-
orders, while certain new groups were 'makers and sellers'. We retain them as
tion in India to any cognisable extent. actual workers in Industry since it will
created with the object of distinguishing
(a) workers from sellers Iand (b) work- With this short review on occupation be impossible to separate the sellers
ers in factories from these engaged in and population we may get to the main fronm'makers and sellers' figures. As the
hand industrv. The net result is to raise part of our discussion. We shall consi- persons under this heading were appa-
the number of detailed heads or groups der the "industrial working force" as rently producer-sellers, we should con-
to 520'".41But objections were raised only those who were directly and expli- sider them as working people in the ma-
that the same occupations might be citly involved in manufacturing activity. nufacturing sector also (when they are
classified under different heads and the To start with, we shall leave aside 'ge- given under the sub-order 'Industry').
method was not adequate to distinguish neral labour' from our consideration. We add all the 'sellers' with the per-
the handicrafts from the factories.42 Our method of tabulation is outlined sons involved in "Commerce" in 1901
In 1911 a new method of classifica- below. census to form a table of "Traders"
tion was accepted. The new scheme The classification schemes in different comparable to the figures given under
followed that of the famous statistician, census tables were not similar. From the Trader sub-order in the later pe-
J Bartillon of France. In the new 1901 to 1921 the figures were, more or riods. These persons should in fact go
scheme, all the occupations were divid- less, directly comparable. In all these to the services sector (a la Colin Clark).50
ed into fou-r classes and twelve sub- three publications the occupation tables The agricultural working force data are
classes with sub-divisions going upto have figures showing 'Total Supported', obtained directly from the census tables
499. The four classes were A) Produc- 'Actual Total Workers' (male and fe- as the data do not need any correction.
tion of Raw Materials, B) Preparation male separately) and 'Dependants' (both The data are given under the heading
and Supply of Material Substances, C) sexes together). "Agricultural Forestry and Fishing" in
Public Administration and Liberal Arts In 1931, the tabulation method was the census tables. For the population
and D) Miscellaneous.'3 The ensuing changed. There we have columns as figures I have considered the population
censuses upto 1931 followed this new 'Total Following Occupation', 'As Princi- tunderthe age group 10-59 relevant for
scheme with little change. Regarding pal Occupation', 'As Working Depen- our operation. I consider this age-group
our field of study the new scheme was
better in that now traders were group- TABLE 3: INDUSTRIAL WORKING FORCE, 1901-1931
ed separately under the order "Trade". Actual Workers
This group consisted of (i) the "Com- Year Total, including
merce" Sub-Order of 1901 and (ii) the Dependants Male Female Total
"Sellers", who in 1901 were included 1 2 3 4 5
in the order showing "makers" also.
1901 44,075,981 11,467,465 5,957,769 17,425,234
We shall now have a few words on 1911 40,352,019 11,503,467 6,611,763 17,515,230
the growth of population over the pe- 1921 33,167,018 10,658,372 5,040,001 15,725,373
riod of oulr consideration. The plague 1931 12,114,962 3,610,462 15,725,424

527
March 22, 1975 ECONOMIC AND POLITICAL WEEKLY

as showing the able-bodied population TABLE 4: POPULATION (1901-1931), ALL-INDIA

in India. The usual practice is to take
the 15-59 age group. But in the under- Year Total Population Population in the Age Group 10-59
developed countries age 10 is consider- Total Male Female
ed sufficient for the poorer section of
1 2 3 4 5
the population to start working.51 Any-
way, I have observed that the conside-
1901 293,377,300 199,694,721 102,897,683 97,125,782
ration of 15-59 age group also do not 1911 312,643,693 210,046,280 108,868,898 101,177,382
affect our conclusion. Population figures 1921 315,350,442 212,201,356 110,540,657 101,660,699
are also taken from the Census tables.52 1931 349,759,318 236,531,599 122,824,995 113,706,604
With the above knowledge and infor-
mation regarding the occupation and the
population, we shall now discuss the ta- TABLE 5: INDUSTRIAL WORKING FORCE AS A PROPORTION OF THE TOTAL
bles constructed below. These tables, POPULATION AND OF THE POPULATION IN THE AGE-GROUP 10-59

are all constructed, as stated earlier,

from the information available in the Year Total Industrial Total Total Per Cent Per Cent
census publications from 1901 to 1931.5 Working Force Population Population of of
I have already stated my method of ar- (Male + Female) in -fie Age- (2) to (3) (2) to (4)
riving at the figures given in the tables. Group 10-59
1 2 3 4 5 6
In Table 3 the space in the second
column showing total industrial popula-
1901 17,425,234 293,377,300 199,694,721 5.94 8.73
tion including dependants remains 1911 17,515,230 312,643,693 210,046,280 5.60 8.34
vacant for 1931 because 1931 census 1921 15,725,373 315,350,442 212,201,356 4.91 7.41
does not supply us with information on 1931 15,725,424 349,759,318 236,531,599 4.50 6.65
dependants. We find in Table 3 that
the number of total population depen-
dent on "Industry" decreased from 1901 TABLE 6: MALE INDUSTRIAL WORKING FORCE AS A PROPORTION OF THE TOTAL
to 1921 but the figures giving actual POPULATION IN THE AGE-GROUP 10-59 AND AS A PROPORTION OF THE MALE
POPULATION IN THAT AGE-GROUP
total working force after a decline upto
1921 shows a tendency to move upward
in the third decade, though the number Year Male Industrial Population in the Age-Group Per Cent Per Cent
Working Force of (2) of (2)
remains much less than in 1901. The Total Male to (3) to (4)
number of female working force any- 1 2 3 4 5 6
how shows a decreasing tendency upto
1931 (as shown in the 5th column in 1901 11,467,465 199,694,721 102,897,683 5.80 11.15
Table 3). But this fall in the number 1911 11,503,467 210,046,280 108,868,898 5.48 10.57
is compensated by a rise in the male 1921 10,685,372 212,201,356 110,540,657 5.04 9.70
working force in the 20's. This is an 1931 12,114,962 236,531,599 122,824,995 5.12 9.86
interesting phenomenon. One obvious
conclusion is that the female workers 6 will make the case obvious. The reason of the declining tendency shown
were being replaced by the male work- Industrial Working Force as a propor- in the census data. It is true that our
ers in the large industries. At the same tion of the total population (column 5, tables do not show any absolute fall
time as the total number does not Table 5) declined from 5.94 per cent in in the number of female industrial
increase (it is abou-t the same in 1931 1901 to 4.5 per cent in 1931. The same working force, but, as I stated earlier,
as it was in 1921), we should compare working force when taken as a ratio we should not exclude the female work-
it with the growth in population. of the total workers' army declines from ers from our consideration. Anyway,
Table 4 gives us the total population 8.73 per cent in 1901 to 6.65 per cent in the fact that the rate of industrialisa-
of India from 1901 to 1931 in column 1931. tion was not sufficient to absorb the
2; columns 3, 4 and 5 give us the total To strengthen our finding from Table increase in the able-bodied population
population in the age group 10-59 and 5, we consider the male industrial in the economy is clear from the table.
the male-female breakdown of this working force separately in Table 6 Moreover, the rate of industrial ocWupa-
population. I shall consider this part where the same story is repeated. The tion was declining over the years even
of the population as the able-bodied male working force in Industry when upto 1931; this is also evident from our
population who may form the total taken as a proportion of the total table. In fact if we consider our in-
'army' of workers in the economy. (The workers 'army' (population in 10-59 age formation of industrial working force
reason for taking 10-59 age-group has group) declines from 5.80 per cent in as a proportion of the total working
been already discussed earlier.) force as formulated by Daniel Thomer,
1901 to 5.12 per cent in 1931 and when
In Table 5 we consider the Industrial they are taken as a proportion of the we find the same story being repeated
Working Force as a proportion of the male workers army the decline is from once more.
total population and the population in 11.15 per cent in 1901 to 9.86 per cent In Table 7 we tabulate the total
the age-group 10-59. Here we find that in 1931 - shown in columns 5 and working force as given by Thorner5'
whether we consider the total popula- 6 respectively of Table 6. in column 3 against the total
tion or the total 'army' of the popula- The informnation available from Tables male industrial working force (as we
tion who could join the working force, 5 and G 'toaether clearly shows that found) in the second column. We then
the proportion of the total workers contrary to Thomer's opinion the in- consider this male working force as a
occupied in Industry, declines steadilv clusion of'data on female working force proportion of Thorner's total working
over time.0 A glanee at columns 5 and in the calculation could not be the force in column 4 and find that the

528
ECONOMIC AND POLITICAL WEEKLY March 22, 1975

TABLE 7: MALE INDUSTnuAL WORKING FORCE AS A PROPORTION OF wrong inclusion of General Labour was
TOTAL MALE WORKINGFORCE one of the factors behind the conclu-
sions Thorner reached. This can be
Year Total Male Total Male Per Cent corroborated from the 1961 estimates
Industrial WorkingForce of (2) also.57 The Census estimates are made
WorkingForce to (3) on the basis of a pro rata distribution58
1 2 3 4 of workers tabulated under the heading
"general labour" among the different
1901 11,467,465 93,607,000 12.25 sectors. In fact if we consider the tab-
1911 11,503,467 97,333,000 11.82 les given by B R Kalra in the appendix
1921 10,685,372 95,734,000 11.16 to the 1961 census publication we can
1931 12,114,962 100,179,000 12.09
come to the conclusion that all the hy-
potheses that Thorner tried to disprove
TABTix8: TOTAL MALE WORKING FORCE AS A PROPORTION OF TOTAL are strongly proved by the estimates.
POPULATION AND POPULATION UNDER AGE-GROUP 10-59 The method of estimation is as rigorous
as the data would permit and I find no
Year Total Total Population Per Cent Per Cent reason to refuse to accept them.
Male Population in Age- of of For the sake of brevity, I submit the
Working Group (2) to (3) (2) to (4) substance of only one table showing
Force 10-59
1 2 3 5 Kalra's estimates59 (Table 10).
4 6
In Table 9 we find that all the sec-
-1901 93,607,000 294,361,056 199,694,721 31.80 46.88 tors show a more or less steady decline.
1911 97,333,000 315,156,396 210,046,280 30.88 46.34 This is particularly true for the second-
1921 95,734,000 318,942,480 212,201,356 30.02 45:12 ary sector, which is more important in
1931 100,179,000 352,837,778 236,531,599 28.39 42.22 our case. The Non-Workers on the
other hand, showed an in-crease after
TABLE 9: MALE AGRICULTURAL WORKING FORCE AS PROPORTION OF 1911. As I stated earlier, Kalra's esti-
POPULATION
mates corroborate my findings.60

Year Agriculture, Total Population Per Cent Per Cent

Forestry Population in Age- of (2) of (2) v
and Fishing Group to (3) to (4) Conclusion
10-59 In this article we have discussed the
1 2 3 4 5 6 work of Daniel Thorner and J Krishna-
murthy regarding the de-industrialisation
1901 64,148,000 294,361,056 199,694,721 21.72 32.12 in India during the early years of this
1911 70,244,000 315,156,396 210,046,280 22.26 3(3.44
1921 69,646,000 century. In the fourth section of this
318,942,480 212,201,356 21.84 33.82
1931 72,197,000 352,837,778 236,531,594 20.46 30.53 article, we resorted to an alternative
measurement of the change in industrial
employment. On the basis of our find-
percentage declines from 12.25 per cent Thorner supposed. ings we observe that the process of de-
in 1901 to 12.09 per cent in 1931. The One main item we have left out is industrialisation did continue even upto
story remains the same: the ratio shows "general labour". We shall take it up 1931. We have tried to show that Thor-
a steady decline upto 1921 and though in a short while. We have also left ner was mistaken in his inclusion of
it picks up again during 1921-1931 the out traders and persons involved in "General Labour" and "Trade" in the
proportionis still less than that of 1901. Transport, Communication and Stor- Industrial Working Force and his ex-
We can also show that even Thor- age - the tertiary sector. Persons in- clusion of Female Working Force. We
ner's data on male working force volved in mining and quarrying were have also shovn that Thorner was in-
declines steadily as a proportion of so small in number that they could in correct in his method of measurement
the total population of the total 'work- no way change the rates. The census of "relative de-industrialisation". In-
ers army' as we formulated (Table 8). estimates in 1961 give the figures as stead of considering the total popula-
This requires us to observe the nature 116433, 296968, 325867, and 21548 tion he considered only the total work-
of the agricultural working force also. for 1901, 1911, 1921 and 1931, res- ing force which itself was relatively de-
As Table 9 shows, the agricultural pectively.56 About traders in the ter- clining over the period.
working force55 also shows a declining tiary sector we can say that though the We have not measured the changes
tendency (except for 1911-1921) as a absolute number shown by Thorner in manufacturing output. Krishnamur-
proportion of either the total popula- increased from 5044000 to 5659000 in thy tried and failed. We have evidence
tion or the workers army. This observa- 30 years from 1901 to 1931, it can be from the study of a number of econo-
tion gives us an even grimmer picture. found that the percentage of traders mists on national income accounting that
Thus it was not only that the industry to the total population decreased from the net national product or the national
was showing a declining rate of absorp- 1.71 per cent in 1901 to 1.6 per cent income in-India, did not increase in the
tion of pqpulation but agriculture also in 1931. The same is, in fact, true for period. Evidence may also be .r4produc-
showed the same tendency. It may be the whole of the tertiary sector. This ed to show that the real wage of the
observed that in the Tables above, where would give -us an idea of the nature of Indian workers did not increase in the
I considered Thorner's data, I have the unemployment problem during the same period.61 These facts together with
always stuck to only the male working time. the evidence of a declining working
force. This shows that the number of We have not discussed 'general force and a declining xindustrial working
femnalescould not disturb the picture as labour' yet. If w.as my -contention that force (in the relative sense from our

529
March 22, 1975 ECONOMIC AND POLITICAL WEEKLY

measurement but in the absolute sense TABLE 10: PERCENTAGE OF DISTR1BUT1ION OF POPULATION OF EACH SEX INT
from Kalra's estimates) lead us to con- WORKEnRSAND NON-WORKERS (WORKERS BEING FURTHER CLASSIFIED INTO
SECTORS,ALL-INDiA)
clude that personalities like William
Digby or Romesh Dutt were perfectly
justified in holding the British rulers in Year Sex Total Primary Sector Secondary Tertiary Non-
India responsible for the economic Population (a) (b) Sector Sector Workers
plight of the Indian people. We may 1 2 3 4 5 6 7 8
safely argue that the Indian economy
even during the early quarter of this Persons 100 31.48 33.44 5.88 7.29 53.39
century was characterised by a stagnant 1901 Male 40.09 43.00 7.52 10.59 38.89
(if not declining) agriculture and service Female 22.61 23.59 4.20 3.91 68.30
sector and a declining manufacturing Persons ,, 33.82 35.98 5.35 6.74 51.93
1911 Male 42.40 45.69 6.78 9.52 38.10
sector with an increasing population Female 24.93 26.02 3.86 3.85 66.27
and hence declining economic activity'
Persons 33.68 35.65 4.89 6.38 53.08
This perhaps also means that the de- 1921 Male 42.29 45.12 6.36 9.04 39.48
gree of unemployment was ever in- Female 24.67 25.74 3.35 3.58 67.33
creasing. Persons 30.24 32.37 4.42 6.51 56.70
1931 Male ,, 40.05 43.17 6.08 9.02 41.73
Notes Female ,, 19.97 21.06 2.69 3.88 72.37

[I am grateful to Amiya Bagchi for

his guidance and inspiration. I have rate (than before - if there was Susobhan Chandra Sarkar") follows
also drawn much benefit from dis- any -,or) compared to the rate of the same method.
cussions with many belonging to the decline of manufacturing output. 33 Cf, "A Note on Working Force
Centre for Studies in Social Sciences, The ratio of manufacturing output Estimates, 1901-1961", by B R
Calcutta, and my friends in Anya to the total may also be overstated Kalra, Research Officer, Office of
Artha.j in domestic money terms if through the Registrar-General India, Cen-
1 William Digby, "Prosperous Bri- high tariff protection and monopo- sus of India, 1961, Appendix-I.
tish India - A Revelation from ly, industrial prices are inordinate- 34 Published in Indian Economic and
Official Records", London, 1901. ly high. All the above possibilities Social History Review, Vol IX, No
2 R C Dutt, "Economic History of are particularly probable in a co- 13, September 1972.
India in the Victorian Age", Lon- lonised economy. 35 Ibid, p 328.
don, 1906. 17 "Asian Drama", Pelican Books, 36 The 1961 census also helps us in
3 Bipan Chanda and Taru Matsui 1968. this regard. The report suggests
had to oppose M D Morris in as 18 "Development and Under Deve- that shifts from agriculture to non-
late as 1960s. See Tapan Roy lopment", University of Califoomia agriculture may be due to either
Choudhury (ed), "Studies in Indian Press, 1964. (i) population growth or (ii) econo-
Economic History", Delhi, 1972. 19 "Economic Development of Latin mic growth. The authors of the
4 The discussion has been includ- America", Penguin, 1971. report judge that the first kind of
ed as a Chapter in the book "Land 20 "Some International Foundations of shift is reflected in the strengthen-
and Labour in India" by Daniel Capitalist Growth and Under De- ing of the number of workers in
and Alice Thorner, Asia Publish- velopment", Economic and Politi- the traditional sector rather than in
ing House, 1962, pp 70-81. cal Weekly, Special Number, Bom- favour of other sectors. There we
5 Thorner, op cit, p 71. Apparently bay 1972. will have to consider Trade and
Thomer fornalises the opinion of 21 "Industries and Under Develop- Commerce in its association with
the personalities like R P Dutt, ment", London, 1971 (particularly other sectors. If an increase in
A S Silvankar, and Colin Clark. pp 16-18). Trade and Commerce is associated
See Thomer, op cit, pp 70-71 and 22 Sutcliffe, Industries, etc. with an increase in traditional sec-
footnote 1, 2 and 3 of Thomer's 23 Other possibilities are not required tors like cultivation, livestock, fo-
article. in our discussion. restry, fishing, mining and quarry-
6 Ibid, p 76. 24 Thomer, "Land and Labour", p ing, etc, and household industry
7 Ibid, p 77. 82-112. then it will indicate a shift due to
8 Published in Indian Economic and 25 Included as a chapter in S J Patel, growth in population and not due
Social History Review (IESHR), "Essays on Economic Transition" to economic development. On the
Vol IV, No 1, March 1967. Asia Publishing House, New York, other hand, a rise in employment
9 Op cit, p 2. 1965. Patel gives an account of in trade and commerce with an
10 lbid, p 3. the studies by other economists associated rise in non-traditional
11 Ibid. also. A K Bagchi also gives a short sectors like non-household indus-
12 Ibid review in the introduction to his try, etc, indicates a rise in employ-
13 Ibid. book "Private Investment in India, ment of the second type. Cf "Re-
14 Ibid, p 15. 1900-1939", Cambridge University lationship between Patterns of
15 Ibid, p 3. Press, 1971. Shift from Agriculture to Non-
16 Output in the aggregate sense is 26 "National Income in India 1901- Agriculture, 1951-1961 and the
a vector and hence what Krishna- 1947" by S Sivasubramanian, mi- Levels of Development", Census
murthy can at all measuire is the meographed, Delhi School of Eco- of India 1961, Vol I, Part-I, A(i),
ratio between the value-products nomics, 1965. Ch V, p 265.
of the manufacturing sector and 27 Thorner, op cit, p 75. 37 M D Morris. "The Emergence of
the national income. The value- 28 Ibid. an Industrial Labour Force under
product will depend, among other 29 Ibid, p 76. Industrial Capitalism, A Study of
things, on the terms of trade be- 30 Ibid. Bombay Cotton", Bombay, 1965,
tween the agricultural sector and 31 Ibid, p 77. passim.
the non-agricultural sector. Pro- 32 To support my view the works of 38 See Table 3 below.
duction stagnancy or decline in the Amiya Bagehi may be cited here. 39 Thorner, op cit, p 77.
manufacturing sector will not be Bagchi in his study: "De-industri- 40 Census of India 1881 and 1891,
reflected in tlhe ratio or in the alisation in Gangetic Bihar, 1809- Reports and Tables. Census of
change in it if the agricultural sec- 1901 and Its -Implication" (to be India 1901 Vol 1, Part I, (Reports)
tor faces a decline at a more rapid published in "Essays in Honour of Part II (Tables). Census of India

5.30
ECONOMIC AND POLITICAL WVEEKLY March 22, 1975

1911 Vol I, Part I,. (Reports) and conclusion. 59 Constructed from the table given
Part II (Tables). Census of India 51 The Factory Act of 1881 had fixed in page 395 "A Note on Work-
1921 Vol I, Part I (Reports) and the lower limit of employing child- ing Force, etc,". Here "All India"
Part II (Tables). Census of India ren at seven years. The working means Union of India which was
1931 Vol I, Part I, (Reports) and hours for children in the age-group a part of India under British
Part II (Tables). 7-12 wverefixed at nine. The Fac- Rule. Primary Sector (a) con-
tory Act of 1892 increased the sists of (i) cultivators, and (ii)
41 Census of India 1911, Vol I, Part- Agricultural Labourers. Primary
I, Chapter XII, Para 507-509. lower age limit to 10 years which
was retained in the 1912 Act. Cf Sector (b) consists of (i) and (ii)
42 Ibid, para 510. above and (iii) persons involved in
43 Ibid. D R Gadgil, "The Industrial Evo-
lution of India in Recent Times" mining and quarrying, forestry,
44 Thorner, op cit, p 87, Census of fishing, plantation, etc. Second-
India 1901, Vol I, PlartI (Reports). Oxford University Press, 1948.
52 Loc cit. ary Sector: Workers in Manufac-
45 Obtained from Census of India turing anid Construction. Tertiary
1931, Vol I, Part II (Table II). 53 See footnote 40 above. Sector: Trade and Commerce,
46 Census of India 1931, Vol I, Part 54 Op cit, pp 78-79, (Table 1). Transport and Storage and Com-
I, p 10. 55 Taken fronmThorner, op cit, pp munication.
47 Op cit, Table 1. 80-81, (Table 1). fact Kalra's findings mnayappear
48 Loc cit. 56 Census of India, Paper Number 1 60 In to make my own estimates super-
49 This has been suggested in the of 1962, Appendix-I, "A Note on fluous. Against that I can only
Census Report, cf, Census of India Working Force Estimates 1901 to submit that I had set out here
1931, Vol I, Part I, (Reports), page 1961" B R Kalra, Registration mainly to locate the faults in Thor-
273, para 112. The term 'Actual Officer, Office of the Registrar- ner's methodology and this neces-
Workers' is used with respect to General, India. sitated making my own 'estima-
all workers in principal and depen- 57 Loc cit. tion' taking into consideration Thor-
dent occupations excluding the 58 J Krisbnamurthy has an objection ner's data.
number of individuals duplicated to this pro rata distribution. Cf his 61 Apart from the obvious theoreti-
by those appearing also as earn- article "Secular Changes in Occu- cal possibility in an underdevelop-
ing by means of subsidiary occu- pation Structure" in IESHR, ed populous colonial country like
pations". Volume II, Number 1, January India, for the empirical evidence
50 As there has been a controversy 1965. But be also concedes that an interested reader may go
over the inclusion of "Mining and "There is certainly an argument through "A Short History of
Quarrying" in the secondary sec- against this [lumping of general Labour Condition under Industrial
tor (Cf Colin Clark, "Condition and labour with agriculture] as we Capitalism", by Juergen Kuczynski
Economic Progress"), I do not do have evidence that significant Volume I, London 1942, See also,
include them in the manufactur- proportion of the unspecified be- A K- Bagchi, "Some Characteri-
ing sector. But because of the long to urban areas." Cf, "The stics of Industrial Growth in
small size of the sector, 'mining Growth of Agricultural Labour in India", Economic and Political
and quarrying' during the period, India", IESHR, Volume I, Number Weekly, Annual Number, February
its exclusion would not affect the 3, September 1972, p 328. 1975.

Changing the Corporate Tax Base

V D Lall

This paper discusses the- question of changing the corporate tax base. While the author agrees
with the findings of Gulati ancdBagchi (Annual Number, February 1975) on the possibility of extravagant
expenditure by those in control of companies, he suggests that the case for changing the existing tax base
rests on even more fundamental arguments than have been spelt out by them. These are set out in
Section 1; the alternative bases are indicated in Section II; and in Section III a feuw policy issues are
raised.
CORPORATE tax reform in India has no change in the corporate tax base. proved that icorporate (lirectors, who
been mainly centred on changing the Nicholas Kaldor, about two decades manage and control corporate affairs,
statutory tax rates, remodelling conces- ago, suggested simplifications in com- have only a negligible financial stake
sions and incentives, and imposing pany taxation; but again he did not in the operations of their units. This
levies on capital gains, excess profits, advocate abolition of corporate profits as does not bind them to bringing eco-
dividends, inter-corporate investments, the tax base.2 nomies into the corporate operations. In
etc. But the basis of corporate tax fact, their privileged position, with
S-TRONG CASE FOR NEW BASE
structure, which is corporate profits, negligible financial stake - except
has not been touched. As the Bhootha- It is only over the last few years possibly where top managements have
lingam Report pointed out: the deter- that serious consideration is being a percentage share in corporate profits
mination of the tax base is "in a sense g;ven in India to the very idea of - enables themr to camouflage personal
even more important than the rates of clhanging the existing tax base, particu- expenses on company accounts, en-
the taxation themselves. The princi- lirly to replacing it with the value- couraging thereby a cost-push effect in
ples and 'methods of determining the -added tax. I S Gulati and Amaresh product-pricing. Thev suggest that
tax base constitutcs the essence of the Bagchi have recently made out a strong this inherent structuiral lacuna can be
main tax structure. It is this structuire case for a new tax base. In their study overcome by shifting the tax base to
which should be built on a reasonably on 'A Proposal for Reforming Corpo- business expenses net of wages.
stable basis."l However, it suggested ration Tax',3 they have convincingly This paper examines the proposal for

531