The evolution of labour markets in India : 1857-1947

THE EVOLUTION OF LABOUR MARKETS IN INDIA 1857-1947 By Dharma Kumar and J. Krishnamurthy (Consultants) Series: Studies in Employment and Rural Development No. 72 Division: Employment and Rural Development Department: Development Economics Development Policy Staff International Bank for Reconstruction and Development This paper surveys the historical literature on the Indian economy between 1857-1947, with particular reference to aspects of labour in the economy. It also, provides a summary of the latest data on historical trends in national income, on the concentration of land, composition of the labour force, real wages and migration. The paper was written as part of the India cLmponent of RPO 671-84 on "Wage and Employment Trends and Structures in Developing Countries." Washington, D.C., September 1981 CONTENTS Page No. I. INTRODUCTION . . . . . ... . . . . . . . . . . . . . . . . . 1 II. THE HISTORICAL BACKGROUND . . . . . . . . . . . . . . . . . . 3 III. NATIONAL INCOME.E.................. . 5 IV. THE CONCENTRATION OF LAND . . . . . . . . . . . . . . . . . . 7 V. POPULATION: URBANIZATION, LITERACY AND OCCUPATIONAL STRUCTURE 9 VI. SPECIAL LABOUR INSTITUTIONS . . . . . . . . . . . . . . . . . 12 VII. MIGRATION AND THE LABOUR MARKET . . . . . . . . . . . . . . . 21 VIII. WAGES IN THE INDUSTRIAL CENTRES AND LABOUR MARKET INTEGRATION: PRICES AND WAGES IN INDIA.................. 32 LIST OF TABLES Table No. Title Page No. 1 Selected Features of the Indian Population . . . . . . . . . . . 39 2 Trends in Literacy, Urbanisation Infant Mortality 1901-1951 40 3 Structure of the Male Working Force 1881-1911 (Percentage) . . . 41 4 The Industrial Distribution of the Working Force .in Undivided India, 1901 - 1951 (Percentages) . . . . . . . . . . . . . . 42 5 Trends in Per Capita Income in India (Rupees). . . . . . . . . . 43 6 Trends in Net Domestic Product since 1868 for Selected Years . 44 7 Growth of NDP at 1946-47 Prices for Selected Sectors . . . . . . 45 8 Landownership in Madras Presidency: Lorenz Ratios, 1853-4 to 1945-6 . . . . . . . . . . . . . . . . . . . . . . . . . . .. 46 9 Estimates of Overseas Migration from India (000). . . . . . . . . 47 10 Estimates of Emigration To Assam 1882-1904 . . . . . . . . . . . 48 11 Estimated Annual Average Real Wage (1938/39 prices). 49 12 Real Wage Indices for India 1857-1950 . . . . . . .. . . . . . . 50 13 Average Daily Real Wage of Male Daily Labourers Nellore District 51 14 A. Real Wages (in Rs.) in Eastern India . . . . . . . . . . . . . 52 14 B. Coefficients of Variation Percentage . . . . . . . . . . . . . . 3 I. INTRODUCTION Even in advanced capitalist economies it is agreed that "the imperfections which characterise almost any actual market seem particularly pronounced in the purchase and sale of labour." These imperfections are generally worse in less advanced economies, almost by definition: "the worker sells himself with his services, . . . his immediate financial need may place him at a disadvantage in negotiating with employers, he is influenced by non-pecuniary motives, and . . . there are numerous objective barriers to free movement of labour" (Reynolds, 1951, p. 2). The relative weakness of the wage earner is much greater in less developed countries, he knows much less about alternative employment opportunities, and the physical, social and indeed sometimes legal barriers to movement are greater. Moreover, there may sometimes be no market at all; labour may be organized by various forms of custom and command (Hicks, 1969, pp. 9-25). Even where market relations do operate, those in less developed countries may exhibit special features in addition to those discussed above. The various propositions regarding these peculair characteristics include; (1) A very large portion of the labour force is self employed, in agriculture, small industries, and in shops and other services. The pure wage element in the income of the self employed is difficult to compute, and besides decisions to move will not be determined purely by relative wage rates, but also by the earnings on other assets for which there may not be perfect markets; the small farmer is likely to be less mobile than the landless labourer. -2- (2) In densely populated countries there is considerable 1/ disguised or open unemployment in agriculture, and modern industry can obtain all the labour it requires at a constant wage equal to the average product in agriculture, not the marginal product (which may be 0); (Lewis, Sen). (3) Wages in the modern industrial sector will tend to be higher than the average product in agriculture, plus costs of movement etc., so there will be a dual wage structure. The reason given for this discrepancy is generally institutional, in particular trade union activity. But Reynold's analysis of high-aage companies - that companies will pay high wages if efficiency or monopoly or just plain luck, enables them to, since high wages have various economic (and political) benefits, is applicable to the modern sector of less developed countries too (Reynolds, 1951, p. 232). The dualism will end when unemployment in agriculture, or more generally, the non-modern sector, comes to an end, as the modern sector expands, or as a result of population decline or increasing demand for the products by the non-modern sector. (Ranis and Fei, Jorgensen). I/ The argument is usually confined to agriculture but,is often true of small scale industry and services, even with known techniques. In recent years the discussion has extended to the whole of what is known as the "informal sector". -3- II. THE HISTORICAL BACKGROUND ,n the middle of the neneteenth century, the population of undivided India may have been around 223 million, sparse by modern standards, but not in comparison with contemporary populations in countries such as Burma and Thailand. The population was overwhelmingly rural: in 1872 over 90% lived in villages (with populations below 5000); indeed it may have been more rural than at the beginning of the century - D. R. Gadgil has argued that the decay of the old cities, such as the capitals of the pre-British kingdoms, was not matched by the rise of new cities like Bombay and Calcutta, but this is a point on which much research remains to be done. If nearly all Indians lived in villages in the 1850's they were not all occupied in agriculture - there was a very wide range of rural handicrafts and weavers, potters and smiths were numbered in the millions. The importance of other services is frequently overlooked - accountants and other revenue officials, priests, village teachers and doctors, musicians and beggars formed a sizeable portion of :the village population. The next 100 years saw the process of the integration of these myriad villages into the national economy, by the spread of transport and a uniform monetary system, and of the incorporation of the Indian economy into the world economy, by increasing the flows of goods, men and capital across India's borders. What impact did these developments have on the labour force and on labour markets? On one widely held view this period saw the emergence of a national proletariat. The commercialization of agriculture led to growing inequalities in landownership and to the dispossession of the small farmer. Since industry did not expand sufficiently to absorb -4- these dispossessed they were forced to stay on in the villages as landless labourers. Indeed, since the imperialist government opened India's markets to British manufacturers, refusing to protect domestic manufactures or give positive incentives, millions of weavers and other artisans lost their means of livelihood, and joined the ranks of landless labour. Agriculture too was stagnant, so unemployment in agriculture grew, and the real incomes of the poor fell (Palme Dutt, S. J. Patel). We will argue that this view greatly exaggerates the growth of landless labour, is wrong as to some causes (such as growing concentration of landownership), and ignofes others (such as the growth of population after 1921). But first we describe briefly certain trends in national income. -5- III. NATIONAL INCOME Our knowledge of trends in national income has made a rapid advance in the last 15 years, though it is still not on very firm ground. Two studies (Blyn, 1966; Sivasubramoniam, 1965) have estimated long term series on a consistent basis, in place of point estimates. There are still very important areas of uncertainty - Heston, criticises both Blyn and Sivasubramoniam, and does not accept the official estimates which show declining yields per acre between 1920 and 1947, and so has revised Sivasubramoniam's estimates; (Heston, forthcoming). His estimates, which seem to us the most persuasive till further evidence on yields and some other points is forthcoming, are set out in Tables 6 and 7. These show that income per head went up from Rs.120 in 1868-69 (at 1946-47 prices) l/ to Rs.171 in 1930 and then declined to Rs.166 in 1945; that is per capita income increased by over 40% in 60 years from 1868-69. This is of course an extremely low rate, compared with the rates achieved not only by Japan, the U.S., or Germany during this period, but also by many tropical producers of agricultural exports, such as Burma, Thailand, Ceylon and Malaya, where unlike India, arable land was readily available (Lewis, 1970; pp. 30-31). But it is also important to note that, contrary to one widely held view, per capita income did not decline under British rule. The question of yields per acre is very important from the point of view of labour absorption in agriculture. Other germane points include: 1/ Heston's per capita income figures are substantially the same as Sivasubramoniam's (see Tables 6 and 7), though there are important sectoral differences. -6- (1) The fastest growing sector was the tertiary sector, and in particular government salaries. This includes comm2rcial undertakings such as the railways, irrigation and electricity, but also other bureaucrats, whose contribution to real income is more questionable., (2) The growth of some of the other services, in particular commerce and banking is also significant as a sign of the growing commercialisation and monetisation of the economy. (3) The second fastest area of growth was modern manufacturing, but even in 1945. it accounted for only 9% of national income. (4) Industry as a whole accounted for 17% which reflects the very slow estimated rate of increase in the output of small scale industries. Output in this sector is very difficult to estimate - Sivasubramoniam takes Census figures of employment (themselves unreliable), and multiplies by real wages which he esitmates to have increase,d by 69% between 1900 and 1946-47; (see Table 11). Heston argues that this increase is too high, although he too feels some increase in labour productivity probably took place. This is an area in which more u!ssearch is urgently needed. -7- IV. THE CONCENTRATION OF LAND It was once very widely held that before the British land was fairly evennly distributed amongst self-sufficient peasant proprietors; hence there were relatively few landless labourers. British rule brought monetisation, the growth of commercial agriculture, and legal changes which facilitated the alienation of land. Mainly as a consequence of these (population changes were given minor importance in this version), the ownership of land became concentrated in the hands of moneylenders and large farmers, and millions of peasants had to live by working for them. In the last 10 years or so, several blows have been struck at this view, and from several quarters. In the first place, land was already unevenly distributed before the British came, and secondly, in many areas the evidence shows that the concentration of ownership hardly increased under the British. The most complete data are available for Madras Presidency where the revenue figures make it possible to calculate Gini co-efficients for landholdings district by district. (Table 8). These show that there has been remarkably little change in concentration of ownership between 1853-54 and 1950-51; and indeed concentration was even reduced in a few districts, such as Tiruchirapalli, Tirunelveli, Coimbatore and Malabar. There are, of course, errors and biases in the data, but in the absence of land reform there was no reason for deliberate underestimction of large holdings. Moreover, these findings are supported by individual village surveys. The main reason is that both large and small holdings have been subdivided as population grew, and in some districts the holdings of the rich appear to have been broken up faster, especially after 1900. -8- Again, a calculation of the Gini co-effic-snts for Gujarat shows the same trends. In Bengal, as in Madras, land was very unequally distributed in Madras as early as the beginning of the nineteenth century. In Dinajpur, for instance 10% of the population owned 50% of the land in 1808, and agricultural labourers formed as much as 18% of the population (Ray and Ray). Similarly, Stokes' analysis of agrarian change in the U.P. shows that population pressure led to the fragmentation of holdings but to no great growth in pure landlessness. There was a tremendous increase in the numbers of small landowners; the great zamindars at the top of the pyramid remained stable, partly because they were protected by primogeniture, but the smaller non-cultivating landlord lost ground, partly as a result of tenancy legislation. The area under occupancy tenancy increased sharply; the rise in prices and tenancy legislation kept down real rents, so the small proprietor suffered. The middle peasant, able to make use of the new opportunities for commercial cultivation, rose to political and economic power - Stokes estimates that nearly 3/4 of the total area was under peasants cultivating some 8 to 10 acres, and adds that after the abolition of zamindari (one of the leaders of the campaign was Mr. Charan Singh) this peasant elite dominated the countryside. Tenures were complex before the British, and in the legal and economic changes of the nineteenth and twentieth centuries complicated the picture even further. It is not possible even to summarize these changes here; the point to be made is that recent research has shown that increasing inequality in landownership was not the major cause of the growth in agricultural labour. -9- V. POPULATION: URBANIZATION, LITERACY AND OCCUPATIONAL STRUCTURE For the period before 1881, population estimates indicate, at best, rough orders of magnitude. For undivided India, in 1800, the population was roughly 179 million. (See Table 1). By 1881 it had risen to 257 million, but grew rather slowly thereafter to about 306 million in 1921. The trends after 1921 were markedly different and hence this year is often described as the "great divide". Population growth accelerated and between 1921 and 1951 the population of present day India rose from 251 to 361 million. Behind this phenomenon lay high birth and death rates of the order of 40-50 per thousand upto 1921; the death rate fell rather sharply to 27 per thousand in 1951 while the fall in the birth rate was much smaller. .Up to 1921, area under cultivation outpaced population but the situation was reversed after 1921. According to Blyn, between 1891 and 1921, population grew at 0.44 per cent per annum while all crop acreage grew at 0.84 per cent; from 1921 to 1947, population grew at 1.12 per cent per annum while acreage grew at 0.35 per cent per annum. In 1901, the urban sector was still rather small; about 11 per cent of the population lived in urban areas, but urbanization was much more rapid over the succeeding decades and by 1951 about 17 per cent of the population was in urban areas. Literacy levels were also rather low in 1901, about 5 per cent of the population was literate; but by 1951, the proportion was nearly 17 per cent. So in terms of literacy and urbanization there was progress, but the terminal levels were still low. (See Table 2). - 10 - Another indicator of the state of development of the economy is the structure of the work force. Between 1881 and 1911, the share of agriculture in the male work force rose from 70.7 per cent to 74.5 per cent and that of manufacturing fell from 10.6 per cent to 9.1 per cent. There are however deficiencies of coverage etc., in these data. For the period 1911 to 1951, we have better data relating to undivided India and covering both sexes. The share of agriculture rose from 74.8 per cent to 75.7 per cent, the manufacturing fell from 9.6 to 8.7 per cent and services from 15.4 to 15.2 per cent. The changes are less striking if one concentrates on the estimates for males which are more reliable. (See Table 3). The most acceptable assertion consistent with the data is that no large change in the structure of the work force took place over this period. Also trends in per capita real output over the whole period from 1860 indicate a slow increase up to about 1920 and then no significant change. The all-India picture of relative stagnancy, however, conceals important regional trends. In Kerala, Maharashtra, Madras and West Bengal there was a relative shift in the work force away from agriculture, while in Orissa and Rajasthan there was a shift towards agriculture. Also within the manufacturing sector, employment in factories grew while employment in the non-factory sector declined between 1901 and 1951. As there is no evidence of declining per capita income, one cannot accept the view that deindustrialization was occurring. But there certainly was a decline in the handicraft sector in terms of employment, though this was more than offset by the expansion of the modern sector. - 11 - For the period 1807 to 1901, for Gengetic Bihar, Bagchi has argued that there was a sharp decline in employment in manufacturing. His conclusion is based on his use of Buchanan-11amilton's evidence for the early nineteenth century. Apart from doubts about the quality of the data, the estimates of spinners and artisans in 1807 appear to be unduly high. After making corrections we find that in 1807 there were about 902 thousand people supported by industry; the corresponding figure for 1901 is 1231 thousand. Their percentage to the total population however, declines from 12.4 to 10.5 over the period. So, the evidence for Bihar suggest some relative decline in manufacturing, but an absolute increase. One cannot however, assume that this reflects the all-India situation in any way. - 12 - VI. SPECIAL LABOUR INSTITUTIONS In pre-British India there were a host of institutions, customs and laws which impeded the freedom of "every man ... both to choose what occupation he thought proper, and to change it as often, as he 1/ thought proper," (Adam Smith, 1937, p. 99) and of employers to hire the workers they thought most efficient and some of these survived to the present day despite laws seeking to abolish them. These restrictions have however been greatly reduced in force and extent, but this is much more the result of economic forces than of law or social reform. The three main institutions we shall discuss are bondage, the jajamani system and caste, as we shall see, they are closely interrelated. Slavery: Both Hindu and Muslim law permit slavery: people -could legally be bought and sold; agrestic serfdom was also recognized. There were various ways in which slavery might arise - parents sold their children during famines, children were kidnapped and sold as slaves, conquered tribes could be enslaved. Since slavery was generally hereditary, by 1800 there were large numbers of people all over India who were legal slaves; they could be bought, sold, employed and more or less treated as their owners or master willed. But the power of the master over the slave has rarely been complete, even in the classical slave societies of Greece 2/ and Rome, perhaps the closest attempt to make the slave into a mere chattel was in the American South. Certainly in India both formal law and 1/ Adam Smith, "An Inquiry into the Nature and Causes of the Wealth of Nations" (Modern Library edition, New York, 1937) p. 99. 2/ M.I. Finley. - 13 - custom prohibited certain acts for certain categories of slaves - for instance, some forms of punishment were prohibited, or the sale of 1/ a slave apart from his family. There was indeed in India a continuum of degrees of unfreedom; and people could also sell themselves as "slaves" for life, or for a term. It was common to contract to work for a master for a subsistence wage to work off a debt; frequently the debt could not be repaid during the debtor's life time and his sons or brothers or other relatives continued to work for the master till the debt was repaid. From the earliest times servile labour was generally found in two occupations: domestic service and agricultural labour (Chanana, 1960; p. 110). We have hardly any figure of the numbers involved in pre-British times, and even the various estimates made in the first half of the nineteenth century are very dubious. It seems very probable though, that the numbers of agricultural labourers in conditions of bondage, in certain cases near slavery were significantly large in certain parts of the country. For instance, in parts of Malabar (in Kerala) they were estimated at 13-17% of the total population, and similar, indeed even higher, percentages were reported of certain districts in Orissa, Bihar 2/ and Bengal. But these appear to have been exceptional; even in neighboring districts the proportion was much lower; in many other parts of India, such as the Deccan, while the existence of "agrestic slaves" was reported, the numbers were very small, and in still others, there appears to be no mention 1/ On the legal texts, e.g. Dev Raj Chanana, Slavery in Ancient India (Peoples Publishing House, New Delhi, 1960) whether these texts were operative is another question; it was reported of Kerala that the master could punish his slave by death and that cutting slaves noses was practiced in "olden days" (Saradamoni, p. 58.) 2/ Dharma Kumar, Land and Caste in South India (Cambridge University Press, Cambridge, 1965) p. 50; Hiranmay Dhar; "Agricultural Servitude in Bengal Presidency around 1800", Economic and Political Weekly, July 28, 1973, pp. 1351-1357. - 14 - 1/ in the records of agricultural servitude - the Punjab is one such case.- These agrestic slaves were almost always drawn from a few low, generally untouchable castes, such as the Cheruman and Pulaiyan of Kerala, The Pariahs, Pallis and Pallans of Tamil Nadu, the Dheds of Gujarat, the Kyats and Chandak of Eastern India, and so on. In some areas, such as Kerala, nearly all members of these castes were agricultural labourers, and largely servile., in others some members of these "agricultural labour" castes followed other occupations, and even the agricultural labourers were not always unfree. The correspondence between ca3te and occupation was often fairly close; since caste data are more plentiful for the pre-census period than occupational figures, the percentage of agricultural labour has been estimated, on the basis of caste figures, at 10-15% of the total population of Madras Presidency around 1800, and perhaps 17-25% of the agricultural population.- 1/ The slave population of British India was, on one estimate, 8-9 million in 1841. On the basis of the various figures sent in by officials in answer to Parliamentary enquiries on slavery, Banaji puts the total number of slaves (of all types) in India at double this or at least 16 million in 1840, or 12% of the estimated population of India in 1841, but the estimates of slaves are particularly unreliable; D.R. Banaji, Slavery in British India (2nd and revised ed., Taraporevala & Sons, Bombay, 1933), pp. 202-3; Kingley Davis. The Population of India and Pakistan (Princeton University Press, Princeton, 1951) p. 25. 2/ Caste disabilityes were particularly severe in Kerala, where the untouchables were also practically unseekables: when walking along the edge of public roads (all they could walk on) they had to cry around to warn the higher castes of their polluting presence, and to take to the fields as they approached. The right to walk on public roads was bitterly fought for in south India until 'Independence (indeed even in 1948 in Cochin). 3/ D. Kumar, Land & Caste, p. 191 - 15 - Even in the Punjab, as all over India, government officials were traditionally entitled to impress labour for public purposes particularly the repair and maintenance of irrigation works, public buildings and roads and frequently private landlords also exacted unpaid labour for these and other purposes.-/ The British prohibited forced labour for private purposes in the first half of the nineteenth century; they themselves used forced labour for public works, but paid for it. The enforcement of labour for public works was only abolished in the second half of the nineteenth century. As one would expect it was generally the lowest castes who were impressed into unpaid service, but if the need was great poorer members of the higher castes were not always spared.- Certainly nearly every caste had a traditional occupation or occupations-from the priesthood and teaching for Brahmins down to skinning dead cattle for Chamars or scavenging for Bhangis. The question is, what prevented a person from taking up a non-traditional occupation? Practically every caste had sanctions against certain occupations, generally those of lower castes. Thus if a Brahmin took to farming, other Brahmins might penalise him to protect the purity of the caste: there were sanctions on the labour supply side in this system against downward mobility. Upward mobility was restricted on the demand side - members of low castes would not be eligible to take up certain occupations. 1/ The most common word for this system is begar, but there are various regional variants, such as the Malayalam uriyam. 2/ As in Travancore in the nineteenth century; Robin Jeffrey, The Decline of Navar Dominance, (Sussex University Press, London, 1976) p. 56; of North India in the Mughal period, Habib writes "Caste seems to have usually determined whether a person was or was not liable to render begar; but cites the case of a man who refused to render begar because his caste was too high, and was consequently beaten to death; Irfan Habib, The Agrarian System of Mughal India (Asia Publishing House, Bombay, 1963), p. 150. 3/ "Castes in Hindu India are ranked, birth-status groups. The caste, or a subsection of it, is usually endogamous, it tends to be associated with an occupation... .The formal ranking of castes is defined in terms of the belief in ritual purity and pollution...." E. Kathleen Gough, "Caste in a Tanjore Village," in E.R. Leach, ed., Aslects of Caste in South India, Ceylon and Nort- West Pakistan. (CUP. Cambridge, 1960),p. 11. - 16 - In other words, was the labour market segmented, certain jobs being open only to certain castes, and certain castes being permitted to take up only certain occupations? In this case, regardless of skill or other differences inthe quality of labour, one would have to consider the supply and demand for the labour of each caste separately. There were certainly some tasks which only certain castes could perform for ritual reasons. Thus only the higher castes could work in the house and for this reason fetched higher prices as slaves. And the lowest castes were forbidden a whole range of occupations: in Kerala even in 1871 certain untouchables could not enter towns so could only work in rice fields; even if hr wanted to buy and sell, he had to place things on the hand and take whatever the buyer offered; even one somehow managed to get capital or education he could not become a merchant or join a profession (Saradamoni, 62). Certain castes were reported to be specially proficient at their traditional occupations; thus the low caste Gokhas fetched high prices in the Cuttack slave market because of their skill as fishermen.- All castes could perform agricultural labour. The Jajmani System: Yet another institutional fetter on labour mobility was the system of prescribed hereditary obligations and duties linking the land- owners in the village to the artisans, labourers, and village officials. The landowning families (Jajman or "lord") received prescribed services for which they paid a prescribed amount. Thus each landowning family would be entitled to so many pots from the village potter, so many ploughshares and hours of repairs to agricultural implements from the village blacksmith, and so on, and in return would make payments, generally of grain gifts at festivals, and so on. The services the kamins had to render were related to their caste 1/ Hiranmay Dhar, "Agricultural Servitude in Bengal Presidency around 1800", EPW July 28, 1973; p. 1853. - 17 - occupations, and payments were traditionally fixed, generally in kind. Variants of the jajmani system were found over most of India (Beidelman, 1959, p. 7), though apparently not in Bengal. The jajmani system was thus linked to caste as well as to land- ownership: the jajmans were landownees, generally of the upper castes. It was, as the terms imply, a system of unequal relationships. The jajman was the master, or patron, controlling access to the land and able to supply credit and that the fact that he made payments as well as received services should not conceal this. The system has been changing under the pressure of-growing population and increasing comercialisation. Thus payments in kind have been increasingly replaced by payments in cash. It has disappeared in many regions. But it has not completely died out, as some recent anthropological research shows. However, even where it survives it covers a smaller portion of the village and especially a smaller number of castes (Kolenda) - many traditional artisans such as potters and leather makers have lost their traditional clientele. But even where the formal jajmani system has vanished, the patron- client relationship between the landed and the landless persists. Thus at the beginning of the nineteenth century caste restrictions on changes of occupation, legal restrictions on the freedom of the labourer, and the jajmani system of prescribed duties and payments, all fettered the free movement of labour, breaking up the labour market into territorial caste and legal status segments. Changes in the law, the integration of the Indian economy, commercialisation and monetisation, all broke down these barriers but none of these institutions has completely died out. - 18 - Indeed legal history reveals the weakness of the law in eradicating deeprooted social institutions. The government 'was moved to act against bondage by the anti-slavery movement in Britain, and its first steps were halting. The government itself owned slaves in certain areas, and it freed its own slaves in Malabar only in 1836. In 1843 an Act was passed barring the sale of slaves (by Government) for arrears of revenue; and forbidding the courts from enforcing rights arising out of the alleged possession of slaves. The Indian Penal Code of 1862 made the possession of slaves and traffic in them a punishable offence; buying and selling slaves could attract upto seven years of imprisonment, and any one who "unlawfully compels any person to labour against the will of that person" could be punished by upto one year of imprisonment. Nevertheless bondage persisted. Slaves were bought, sold and hired out in Malabar even at the beginning of the twentieth century. But this was comparatively rare, and hereditary slavery disappeared over most of India in the twentieth century. But this was brought about not so much by legal change as by social reform and the growing unwillingness of the lowest castes to accept their caste status.- Ritual caste prohibitions as to occupation are now of little significance except for the priesthood and domestic service; for instance, many if not most, Hindu households will not employ the scheduled castes as cooks. But in general all castes try to enter lucrative occupations - as Brahmins run tanneries a most highly polluting activity - and their success depends on barriers to entry in the shape of education qualifications or skills which a particular group is unwilling to impart to others. And if sweepers are still mainly Bhangis it is because their work is still very poorly paid. 1/ On these social reform movements see in particular Jeffrey, Beteille, - 19 - "Debt bondage", i.e. working at a subsistence wage to work off a debt, did not contravene the slavery provisions of the Indian Penal Code. It flourished all over India throughout the nineteenth century, and cases are reported of people working for years to wipe out inherited debt. The abuses of the system were notorious - the wages were sometimes so low that the labourer could never hope to wipe out his debt; the original sum was fradulently inflated, the labourer being unable to read his bond, and exhorbitant rates of interest were charged. In a'sense these abuses are implicit in the term "debt bondage" - the bonded labour is forced in effect to accept lower wages than a free labourer who could borrow from some one other than his employer. Occasionally laws were passed to control the terms of the bond, such as the Bihar and Orissa Kamiuta Agreement of 1920. Moreover, there is a long history of attempts to control money lending, going back for centuries, and extending to the present. Successive governments have prescribed maximum rates of interest and endeavoured to prevent fraud, and the voluntary or compulsory liquidation of rural debt. But the very number of these laws shows how short-lived their effects are. Debt bondage responds not to legal changes, but to changes in the supply and demand for labour credit. Employers give loans to secure labour services; labourers enter into labour contracts for debt when they need credit for marriages and illnesses, and cannot obtain it in any other way. In the latter part of the nineteenth century, the growing markets at home and abroad increased the demand for agricultural labour; the supply of labour was checked by frequent famines. In some areas such as parts of south India real wages rose, and a few labourers may have managed to free themselves of debt, but wages were not high enough for this to be a general phenomenon, and in many areas employers found it profitable to extend credit, and debt bondage rose (on Bihar see Mundle, 1979: 96). With the growing pressure of population on land from 1921 onwards, employers no longer feel the need to offer loans - 20 - to secure their labour supply - and this has been the main factor reducing the extent of debt bondage (Bremen; Atchi Reddy). Statistics on this are of course difficult to come by, and unreliable when they are found particularly since debt bondage has been illegal since 1976. The Bonded Labour System (Abolition) Act of 1976 made labour bonds illegal, and also extinguished all past bonds but from various reports it would appear that debt bondage is not now wide- spread, (Bardhan and Srinivasan) though it has by no means been eradicated, even by the 1976 ordinance. Thus a field survey made of Palamau district in south Bihar in 1977 revealed that many tribal and lowcaste labourers still were bonded by debt; their wages were considerably lower than those of free labourers and their hours of work longer, so that the implicit rate of interest on their loans was % (Mundle, 1979 - 21 - VII. MIGRATION AND THE LABOUR MARKET Overseas Migration Net migration (emigrants less returned migrants) has been estimated by Kingsley Davis* using census statistics and other sources of migration data for the period 1834-1937. His results show that the areas receiving the bulk of the overseas migrants from India were British colonies in tropical regions accessible by water viz., Burman, Ceylon, British Malaya, Mauritius, Fiji, the Caribbean and East Africa. Between 1834 and 1920 about 21 million Indians migrated overseas, about 16 million returned and about 5 million were net out-migrants. The demand for this labour arose from the abolition of slavery on British colonies and the consequent need for cheep labour suited to tropical conditions-on the plantations. Technology on the plantations** was labour- intensive and did not change with the expansion in demand for their products stemming from the industrial revolution and the move towards free trade. Plantation products were priced low and profits kept high by expanding the volume of exports. This strategy with respect to profits made cheap labour essential for labour costs accounted for a large part of total cost.***. Until 1920 labour for plantations overseas was recruited by an indenture system which usually involved a five year contract with no personal freedom, low pay and harsh living conditions. This system was abandoned in 1920 and * K. Davis, The Population of India and Pakistan, Princeton, 1957, pp. 99. ** Tea, coffee, sugar and later rubber, plantations were the principal sources of demand for labour. *** H. Tinker, A New System of Slavery, London 1974, p. 21. Tinker has an extensive discussion of overseas migration from India upto 1920. - 22 - replaced by the Kangani system which had developed even earlier. For migration to overseas, but relatively nearby destinations, the Kangani system took the form of relatively short term contracts. Generally verbal agreements were concluded with a headman who was often both recruiter and field foreman. This system was freer than the indenture system but did have elements of duress. The Kangani system was progressively replaced by free labour migration which was usually-seasonal and characterised by a high rates of return.* The first world war checked migration to some extent but a revival occured over the 1920's followed thereafter by a slackening due to the depression. The direction of the labour flow also changed with an increasing focus on the Asiatic region in particular to countries near India.** The major labour catchment districts for overseas (and incidentally for internal migration too) were a set of districts in the north: Santhal Parganas, Hazaribagh, Ranchi, Manbhum, Birbhum, Singhbhum, Palamau, Shahabad, Patna, Gaya, Muzaffarpur, Champaran, Saran, Darbhanga, Monghyr, Ballia; Gazipur, Azamgarh, Fyzabad, Basti, Gonda, Gorakhpur, Banaras, Mirazapur and Jaunpur; and in the south, the Tamil districts of Trichinopoly, Mandura, Ramnad Salem, Tanjore,Chingelput, N. Arcot, S. Arcot, Malabar, Tinnevelly and the Telegu districts of Vizagapatnam, Ganjam, East Godavari, West Godavari, Guntur and Nellore***. An important question in this context relates to the caste composition of the emigrants. Grierson in 1883 studied 1200 migrants from north Bihar and found 264 Muslims, 231 high caste people, 454 from middle level castes - agricultural castes like Kurmis etc. -- and 277 from the lowest castes. * Cf. K. Davis, Population ** Cf. K. Davis, Population of India and Pakistan, p. 99. *** Cf. H. Tinker, A New System, p. 40. - 23 - For the same year the protector of Emigrants at Calcutta reported that out of 7695 Hindus, 1995 were from high castes, 2454 from middle level of agricultural castes, 456 artisans and 2790 from the lowest strata.* These data do not confirm the common view that migration was heaviest among those at the base of the social pyramid. However, the reliability of these statistics is rather questionable: there may have been a systematic tendency to over- state one's caste status when such investigations were made. Internal Migration: Internal migration has always been important in India, but the bulk of it probably took the form of rural to rural migration which was never systematically recorded. It is however clear that there were important streams of internal migration in search of employment. There were movements to the coal industry in Bihar and Bengal, the jute textile industry in and around Calcutta, the tea plantations and virgin farm lands in Assam, the canal colonies of Punjab, the plantations (tea, coffee and rubber) of the south west and the cotton textile industry in Bombay. Very little information is available about migration to the south west strip. Probably the majority of immigrants into this region were from nearby overpopulated areas, including the Tamil districts of the Madras Presidency which as we have earlier pointed out were important sources of international labour migration. It is important to notice that the volume of labour going to the modern sector was small in relation to the Indian work force. Even in 1951 only about 2.1 percent of the work force was in modern manufacturing industry. If one were to add in all the employment in mining and quarrying and plantation, the share would not exceed 3.5 percent. So, looked at in relation to the * Cf. H. Tinker, A New System of Slavery, p. 56. -24- vast size of the Indian work force the modern sector, however defined, was extremely small and did not really draw away huge volumes of labour from the countryside as a whole. However, the bulk of this labour came from certain catchment areas,and in these areas, the effects of labour migration to the Indian modern sector may have been more significant, but this has not been analysed so far. Further, these same catchment areas also provided labour for international migration which was quite substantial. After looking at some of the important flows of labour to specific Indian destinations, we examine in a subsequent section, the complex of forces in operation in the catchment areas "pushing" labour out to these and other destinations. Assam: Import of labour into Assam started around the middle of the 19th century. Experimental tea gardens started by the government in 1836 were generally handed over to private operators by the middle of the century. By 1840 two-thirds of the experimental gardens were handed over to the Assam Company free of rent for at least the first ten years.* Given the highly labour-intensive nature of work on tea plantations large amounts of labour were permanently needed. The indigenous population was small and, given plentiful land, there was virtually no landless labour. The consequent shortage of labour is cogently explained by A.T. Campbell (Govt. of Bengal Report) "The wants of the local peasants are few and easily supplied .... There is no labouring class distinct from the cultivator of the soil and dependent for existence on the wage of labour.... Owing to the lightness of the government demand, and the fertility of the soil, the cultiv?ator can, without much exertion, raise from the land sufficient to pay his rent, and * See A Guha "Colonization of Assam: Years of Transitional Crisis 1826-40" Indian Economic and Social History Review, 1968. - 25 - maintain himself and his family."* Labour had to be produced from outside and by 1919, the population on the tea plantations had crossed the million mark. Labour was procured from distant sources, with the Chota Nagpur region being an important source accounting for possibly a half of the total migrants into Assam between 1870 and 1920.** Pure aboriginals or "junglees" fetched the highest rates paid to labour recruiters at the rail and bridge heads for Assam and the Surma valley. Rs. 115 was the maximum rate of procuring such labour, Rs. 100 for "good hardy coolees" and Rs. 90 for "Coolees suitable to the Surma valley and for healthy gardens in the Brahamputra valley."*** Migrants into Assam going into the new farm lands were, however, largely Bengalis, a high proportion coming from Mymensingh district. This migration became significant after 1800 and tapered off after 1931. The Coal Industry: Until about 1920 this industry dependent largely on labour from the coal districts and continuous districts viz from the Chota Nagpur region . Towards the end of the first World War, however, a wider catchment area became necessary and private recruitment centres, and those of the Directorate of Unskilled Labour Supply, were established in the United and the Central Provinces. The labour from these new sources appears to have been mainly from the depressed classes. * Quoted in Wolf Mersch "The Formation of a Rural Proletariat on the Tea Plantations of Eastern India" 1860-1921, paper presented at the conference on Indian Economic and Social History, 1975, Cambridge (mimeo). ** For figures of migration into Assam, see Tinker, New System of Slavery and Mersch, op. cit. * See Mersch, op. cit. @ For a detailed account of labour recruitment in the coal Industry see C.P. Simmons, "Recruiting and Organizing on Industrial Labour Force in Colonial India Case of the Coal Mining Industry 1880-1939, IESHR, Oct.-Dec., 1976. -26- A novel feature of the recruitment system for the coal industry was that the earlier and bigger collieries entered the land system as intermediary tenure holders. Labour households were granted tenancies by the colliery in return for a guaranteed labour supply. The Bengal Coal Company for example had Zamindari rights over an area of 400 square miles. The later and smaller collier.ies could not take recourse to this system due to inadequate resources. Several collieries also had arrangements with local zamindars whereby a certain amount of labour Zor the collieries was guaranteed by the zamindar for a consideration. In addition, contractors were employed who not only supplied labour but also often exercised some managerial functions in mining, raising and loading of coal. This method ensured a certain number of disciplined, controlled and organized gangs of labour and also freed the management from the task of handling the labour. This system was abandoned when the price of coal fell during the Depression and costs had to be pruned. All managerial functions were then taken over from the "raising contractor" by the concerned departments of the coal companies. The Jute Industry: This industry rapidly expanded after 1891 when a major inflow of migrant workers into the industry began. Migrant labour met the needs of the expanding industry and perhaps also displaced existing local Bengali labour in the mills. It is significant that tribal labour was unimportant in this industry. The 1911 census revealed that the majority of the migrants were from a rural background, a sizeable proportion from a craft background and a high proportion, about 25%, from the lowest in the social hierarchy*. * R. Dasgupta, "Factory labour in Eastern India - sources of supply 1855- 1946: IESFa. July-Sept., 1976. - 27 - Cotton textile industry in Bombay: The industry began in the 1850's and it drew labour from other parts of the Bombay Presidency especially the Konkan district of Ratnagiri and from the Deccan districts. Also some amount of labour came from farther field. The Bombay Presidency was not, however, a significant source for international migration and the major catchment areas did not get tied to the Bombay labour market to any substantial extent. The labour was generally procured by foreman-jobbers who also managed the labour in the mill and collected "dasturi" from the workers they hired. As in coal, the labour management function tended to be contracted out. As the industry upto 1920 required a fairly large quantity of unskilled labour and as they were more anxious to see that capital installed was fully utilized, they found that the foreman-jobber did the job very well provided certain norms of output per machine were prescribed. However, after 1920 as the labour force became more militant and organized, the jobber system declined and the mills began to manage their labour directly. Push Factors behind Migration: In this section we examine somewhat more closely the regions which were major labour catchment areas to find out what were the forces operating tending to "push" labour to seek employment outside the area. Chota Nagpur region: The population of this region was growing at about 1.2% per annum between 1881-1931 and population pressure on the net sown area was rising. At the same time large tracts of land were cultivable but remained uncultivated. Small holdings, very little irrigation, rural indebtedness and very low rates for agricultural labour were salient features of the agrarian economy.* Uhile these factors wculd explain the tendency for large * See D. Schwerin in D. Rothermund and D.C. Wadhwa (ed.) Zamindars, Mines and Peasants, New Delhi, 1980, pp. 25-33. - 28 - amounts of labour to migrate from the region in search of employment, the continued existence of a large extensive margin on land needs to be explained. The explanation is best sought in the agrarian structure of the region. The Permanent Settlement granted proprietory rights to the petty Rajas in the region and transformed their subjects from small landholders into dependent tenants. The intermediary tenure holders, who were the effective landlords, characteristically siphoned off most of the meagre surplus accruing to the cultivators by means of high rents and other exactions including "begar" or forced labour. In addition there was a steady encroachment on the land held by cultivators, by the landlords who then cultivated with the help of hired labour. Under this system as rights were very insecure the cultivators had neither the will nor the resources to extend cultivation or improve land. The superior landholders could not invest in land improvement on a sufficiently large scale because they dissipated their incomes in a fashion that kept them in debt.* Thus large tracts of cultivable land remained uncultivated and out-migration of labour persisted. Emigration declined only with the expansion of the coal industry in the region and the emergence of the Jamshedpur steel complex. Eastern Uttar Pradesh and Bihar: We consider here eastern UP and Bihar districts other than Chota Nagpur. Districts from which emigration was important seem to have been characterised by high and rising population density over the second half of the nineteenth century.** This was accompanied by an exhaustion of * See Schwerin "The Control of Land and Labour" in Rothermund and Wadhwa (ed.), Zamindars, Mines and Peasants, pp. 36-49. ** B.N. Ganguli, Trends of Agriculture and Population in the Ganges Valley, London, 1938. - 29 - both the extensive and the intensive margin on land. In Jaunpur district for example, the average density increased from 659 persons per sq. mile in 1872 to 816 persons per sq. mile in 1891. Pasture land decreased as population pressure grew and the extension of area under cultivation reached the limit by the early twentieth century. The incidence of double cropping tended to be higher in areas of higher population density. In the east U.P. districts the crop-mi:x shifted over the late 19th and early 20th centuries towards heavy yielding subsistence crops, some of which were better suited to double cropping.* The Bihar districts serving as sources of migrant labour were also characterized by high and rising population density and exhaustion of margins on land**. The increased demands of the rentier classes upon tenants at will and declines in the real reward to agricultural labourers as custom gave way to market forces probably acted in concert with rising population pressure and exhaustion of margins on land, to 'push' labour into the net of recruiters.*** * Reclaimed land was largely devoted to early rice which made double cropping possible. Area under cheaper grains increased, e.g. in Jaunpur the area under gram was 4 times as large in 1908 as in 1841. Area under peas also rose significantly. Banaras district showed a similar trend. Area under maize and barley - heavy yielding subsistence crops - also rose. Area under sugarcane, which requires long fallowing, fell, and can partly be attributed to need for more intensive use of land for subsistence crops. ** In Muzaffarpur population density rose from 765 persons per square mile in 1872 to 906 persons per square mile in 1891. In 1891 60% of total area was under cultivation and encroachment on pastures and mango groves had occurred. In Gaya, unfavourable natural conditions hampered extension of cultivation. *** See also L. Chakravarty, "Emergence of an Industrial Labour Force in a Dual Economy - British India 1880-1920", IESHR July-Sept., 1978. 30 - Madras Presidency: This region too was characterized by a high and rising population pressure on land accompanied by falling real wages in Ganjam,- Vizagapatam, Bellary, Tinnevelly, Salem and Coimbatore over the last quarter of the 19th century. Cultivation was extended to inferior lands. Forests were cut down to extend cultivation and this could well have adversely affected rainfall. Agricultural practices with respect to dry lands were unsatisfactory and led to rapid exhaustion of the soil. Evidence of increases in area under subsistence crops is available for the Madras Presidency too - 1884-1900 saw a 34% increase in area under cumbu, 20% increase under rice, 17% under cholum and 14% under ragi. Famines combined with population pressure to push labour out of the agrarian economy. Some General Conclusions The argument in brief, therefore, is that despite adjustments in the agrarian economy to rising population density, the soil could not, given technology and water resources, support the large and rising agricultural population. For the poor and asset-less, emigration was the only avenue left for survival. In Chota Nagpur, of course, it was not the non-availability of cultivable land but the insecure structure of rights that acted as a barrier to extension of cultivation. @ Dharma Kumar, Land and Cast, in S.)uth India, Cambridge, 1965, pp. 165, see also pp. 117-119. - 31 - The presence of a significant number of ex-craftsmen in the jute textile mills is consistent with deindustrialization thesis of Bagchi based on Buchanan-Hamilton's survey. According to Bagchi between 1809-1813 and 1901, in the districts of Patna, Gaya, Bhagalpur, Purnea, Shahabad and Monghyr the proportion of industrial to total population fell from 18.6% to 8.5%. While these estimates are far from satisfactory in terms of the assumptions made, even on revisions made by others of Bagchi's estimates*, it does appear that the relative proportion engaged in industry declined even if the absolute numbers did not. The relative stagnation of traditional industry, and the decline of specific lines of production would still account for the craft background of many migrants to the Calcutta Jute Mills. As regards overseas migration, significant income differentials or "pull factors" worked together with 'push' factors to stimulate migration. Wages in Ceylon and Burma were relatively very high and migrants frequently returned with savings which they invested in many cases in land. Remittances too were very important: Ganguli contends that in Eastern U.P. the survival of many poor agricultural families depended on remittances from members who had migrated.. As far as "pull" factors in domestic markets for migrant labour are concerned, comparisons of agricultural wages and conditions of work in catchment districts with wages and conditions of work in the receiving industries and plantation are required. Such comparisons cannot really be made with existing data sources. * See Marika Vicziani, 'The De-industrialization of India in the 19th century", and Bagchi's Reply', IESHR, vol. XVI; No. 2, 1979. Some revisions of Bagchi's estimates by Krishnamurty (as yet unpublished) suggest that the absolute size of industrial population may have risen while its relative proportion may have declined. - 32 - VIII. WAGES IN THE INDUSTRIAL CENTRES AND LABOUR MARKET INTEGRATION: PRICES AND WAGES IN INDIA -- a serial publication -- constitutes the principal source of wage and price data for the 19th and early 20th centuries. This source is not reliable as regards to agricultural wages because (1) untrained reporting agencies were used; (2) cash wages for labourers in towns and their neighbourhood were commonly returned and these are unlikely to be typical of the rates in rural areas prevailing in the concerned districts. We have therefore not attempted to compile an All-India series but there is however a careful study for one district (Nellore) for 1871-1971, which is quoted in the Appendfx* Wage data for selected industrial establishments are also available in the Prices and Wages. But there are many problems with the data, some of which we indicate below. First, in most cases the data relate to one unit in each area and hence they may be unrepresentative. In particular, as a unit grows older the average length of service of workers would increase and this would tend to push up the average wage in that unit, though average wages in the industry as a whole may not have risen. However, wage data are provided for a number of skill categories and separately for unskilled labour. Secondly, wage rates for men, women and children are not separately recorded. Thirdly, the wage period used differs between units and over time for the same unit and it is difficult to know how to convert daily wages into weekly or monthly wages and vice versa**. Finally in the case of the paper industry, * Archi Reddy's pioneering study uses private farm accounts for the first time; he has used this data to correct the official wages series. Another new feature of the thesis is the description of the actual process of wage bargaining - see Appendix. ** For collieries of a Bengal coal company monthly wages are recorded upto 1912 and daily wages from 1913 through 1922. To obtain comparability, judgements about the number of days worked per month have to be made. - 33 - one unit was covered from 1867-1906 and a different unit was covered there- after. This again creates non-comparability. A possible consequence of the growth of an industrial labour force in the Bengal region would be the development of an integrated labour market in the area with a tendency for inter-unit wage rate differentials for the same sort of labour to decline over time. We test this hypothesis for Bengal over the period 1887-1920. We constructed real wage series for certain selected industrial establishments and estimated the coefficient of variation of real wages between the concerned establishments for each year. A clear downward trend in the coefficient would be indicative of the emergence over time of an integrated labour market. To make comparison meaningful we selected those categories of labour which could be broadly classified under unskilled labour. -Money wage series were obtained from Prices and Wages in India and to arrive at real wages the relevant retail price of rice (common variety) in the appropriate administrative division (also from Prices and Wages in India) was used as a deflator. The real wage series used related to: (1) miners at a colliery at Raniganj. Here the Burdwan rice price was used as deflator. C2) Coolies at "a paper mill in Bengal." Until 1906 the wages relate to the Bally Paper Mills and hence Hoogly rice prices were used as the deflator. From 1907-1920 the data relate to the Titagarh Mill and hence 24 Parganas rice prices were used as the deflator. (3) Miners at a colliery in Sodipur, for which 24 Parganas prices were used for deflation. The coefficients of variation of the real wage rate of the three units showed no trend over time. Coefficients of variation of real wages of miners in Raniganj and Sodipur collieries were also computed to see if any integration in the market for unskilled labour in mining occurring. Again -34- there was no trend over time in the coefficients of variation. Last it should appear that our procedures of deflation are responsible for this result, we also compared wages within Burdwan division, in the collieries at Raniganj and Sankatoria. Again the coefficients of variations of money wages in Raniganj and Sankatoria for unskilled labour showed no tendency to decline over time. Thus the thesis that an increasingly integrated labour market was emerging in Bengal over the period 1887-1920 is not supported by the data. Even within a district, the wages for the same kind of labour in the same type of activity (coal mining) show no tendency to converge overtime. The real wage series also indicate that real wages were not rising over the period studied. Fluctuations in real wages do occur due to fluctuations in rice prices: money wages tended to be sticky for some years and then they rise marginally. The shifts in money wages were all upward but price rises nullified these increases after a lag. These findings suggest that the labour market in Bengal in the period under study did not get increasingly integrated. It should be noticed that the area covered was fairly well integrated over time by the modern railway network and this makes the conclusion counter-intuitive. Moreover, the data used are perhaps the best available from the Prices and Wages. Using price data from the Prices and Wages, relating to district headquarter towns, John Hurd has shown that price variation across space diminished between 1870 and 1920. Also Mukul Mukherjee has demonstrated the same phenomenon for Bengal proper for the same period using sub-divisional prices. It is significant therefore that the same tendency is not present in the case of wages for unskilled labour in the organised sector in the same area. [See John Hurd II, "Railways and Expansion of Markets in India", Explorations in Economic History, vol. XII, No. 1, July 1975 and Mukul Mukherjee, "Railways and Their impact on - 35 - Bengal's Economy, 1870-1920", IESHR, vol. XVII, No. 2, 1980. The system of labour recruitment via contractors and the virtual tying of labour to particular middlemen working for particular industrial establishments could perhaps be an important explanation of the lack of integration. Detailed industry-wise studies are needed to understand the working of the labour-market given conditions of destitution in the catchment areas and oligopsonistic practices in the labour markets. At the same time, the evidence certainly shows that the trends in wages in a particular chosen unit in an industry are not at all representative of even that industry in that area. One's confidence in aggregative real wage series based on these limited observations is therefore seriously undermined.* It is imperative that wage data for many more industrial units over long periods of time should be obtained before real wage trends for even the organised sector can be relied upon for this period. * A number of scholars here used the Prices and Wages to obtain real wage trends for such industries as the cotton textile industry in Bombay and the jute industry in Calcutta. Often these are based on one observation per year for unskilled labour and could be extremely unrepresentative. See e.g. M.D. Morris, Emergence of an Industrial Labour Force in India, Bombay, 1965 and K. Mukherjee, "Real Wages in Cotton Textile Mills in Bombay City and Island, 1900-1951", Artha Vijnana, 1960, and Radhakamal Mukherjee: The Economic History of India 1600-1900, Allahabad 1967, Ch. 3. - 36 - Hours of work and intensity of employment A subject which has been almost totally neglected in the historical literature is changes in the length. of the working day and in the intensity of agricultural labour. The sole exception is M. Atchi Reddy's study of Nellore district (Atchi Reddy, 1976), and most of his findings probably apply to many other parts of India as well. There are, of course, no systematic data on the hours of work of agricultural labourers. The hours must vary enormously with type of soil, crop, season, rainfall, and other natural factors. The availability of inputs, particularly irrigation water, is another cause of variability. Moreover hours do not have to be regularly reported to Government, since there is no legislation regulating hours, as there is for factory labour, although the desirability of doing so was debated at least as early as 1927 (Report of Royal Commission on Agriculture in India, 1927). Official reports on agricultural labour assume a 'normal' working day from 7 to 8 hours, in the case of the monthly wage censuses, to 10 hours (Second Agricultural Labour Enquiry). Atchi Reddy has used a variety of sources, including private farm accounts, going back to the late nineteenth century, and considerable ingenuity, to show that there has almost certainly been a fall in the average length of the working day of agricultural labour over the past century in Nellore district. This is indeed what one would expect given the growth of population and the lack of employment opportunities outside agriculture - factors operating in many parts of India. Between 1871 and 1921 there was hardly any increase in the population of Nellore district, and the numbers of male agricultural labourers actually fell, whether due to famine mortalities or - 37 - emigration. But the net sown area increased by 7%. There was a complete change in the next 50 years. The death rate fell, the avenues for emigration were closed off, the population increased by 68% and the numbers of agricultural workers by 150%. But gross sown area could only be increased by 21%. Some forms of non-agricultural employment within the district may have declined - Atchi Reddy mentions in particular the effects of deforestation on hunters and gatherers of forest produce. The pressure of population was manifested in various ways. The size of the operational holding fell sharply. Partly as a result of this, and partly because the employer no longer feared labour scarcity, there was an enormous fall in attached labour. Attached labourers (i.e. those contracted to work for a year or more) may have been around 70% of the work force in 1921; by 1971 they were under 10%, even in rich irrigated villages. These figures are supported by the fact that the wages of the annual labourers fell much faster than the daily wage rates. Annual labourers had always worked longer hours than daily labourers; in addition some of them (the labour headmen, talaimanishi in Telugu) played an importane supervisory role. Thus the decline of annual labour is a pointer to the fall in the length of the working day. But direct evidence is supplied by private farm accounts, which show that employers provided far fewer evening meals to their workers, and by the recollections of the villagers, who blamed the slackness of the current generation, masters and employers, the corrupting influence of the radios and cinemas and so on, as elders will. (Atchi Reddy, 1976; pp. 508-523). - 38 - An interesting question for the economist is the relationship between the length of the working day, and the efficiency per hour. Paddison, the Commissioner for Labour in Madras, said in 1927 that the Madras Labourer worked longer hours than his counterparts in the rest of the country, but "works fairly leisurely". (Atchi Reddy, 1976, p. 512). Employers, when interviewed, stated that efficiency fell off towards the end of the working day, so that beyond a point they preferred to have a larger number of workers, each working shorter hours. As against this, one must take into account costs of supervision. (In Nellore the question is further complicated by the fact that the wages, and hence almost certainly the health and strength of the worker, have declined over the last 100 years.) Controlled experiments on the relationship between length of working day and efficiency in agriculture would be well worth-while. These findings about hours of work have obvious implications for the evaluation of the wages series. On the one hand, since the length of the working day has probably declined, the effective wage rate per hour has risen to that extent, all other things remaining the same. On the other hand, perquisites have fallen, partly as a result of the shortening of the working day, so to that extent the offical wages series conceal a fall in real income. - 39 - Table 1: SELECTED FEATURES OF THE INDIAN POPULATION Population (millions) Year Undivided Indian Birth Rate % Death Rate % India Union UI I U UI I U 1 2 3 4 5 6 7 1800 179 154 1850 223 189 1871 255.2 209.1 1881 257.4 210.9 49 .. 41 1891 282.4 231.4 46 .. 41 1901 285.3 238.4 49 51 43 43 1911 303.0 252.2 48 49 47 49 1921 305.7 251.3 46 46 36 36 1931 338.2 278.9 45 45 31 31 1941 389.0 318.7 40 .. 27 1951 361.1 Sources: Columns 2 and 3: Figures for 1800 and 1850 are from Suranjan Sen Gupta et. al.: "Estimates of 19th Century Population of India", and "Estimates of 19th Century Population of Pakistan", Bulletin of the Socio-Economic Research Institue, Vol. 3 No. 1 (1969) and Vol. 4 Nos. 1 and 2 (1970). Columns 4, 6 and the rest of 2 are from K. Davis: Population of India and Pakistan, Princeton 1951. The figures in columns 5, 7 and most of 3 are from M. Nanavati and J.J. Anjaria, Indian Rural Problem, (Sixth Revised Edition). - 40 - Table 2: TRENDS IN LITERACY, URBANISATION INFANT MORTALITY 1901-1951 Year Per cent Literate Per cent Urban Infant mortality Expectation in total population population total % live births of life Birth 1901 5.4 10.8 295 23.5 1911 5.9 10.3 282. 23.1 1921 7.2 11.2 247 24.8 1931 9.5 12.0 227 29.3 1941 16.1 13.9 199 32.6 1951 16.7 17.3 Sources: Literacy rates are from Census of India 1971, Series 1, Paper I of 1971 Supplement, Provisional Population Totals, (Delhi, 1971), p. 37 Urbanisation figures are from Ashish Bose "Six Decades of Urbanization in India", Indian Economic and Social History Review, Vol. II No. 1, January 1965. Estimates of infant mortality and expectation of life are from P. Dasgupta "Estimation of Demographic Measures for India, 1881. Based on Census Age Distribution". Population Studies Vol. 25 November 1971. - 41 - Table 3: STRUCTURE OF THE MALE WORKING FORCE 1881-1911 (Percentage) Sector 1881 1911 1. Cultivators 51.7 53.5 2. Agricultural Labourers 10.7 13.4 3. General Labour 8.3 2.7 4. Plantations, Forestry, Fishing etc. 1.7 4.9 5. Mining and Quarrying 0.1 0.1 6. Manufacturing 10.6 9.1 7. Construction 0.5 1.1 8. Trade and Commerce 4.9 5.5 9. Transport, Storage and Communication 1.8 1.7 10. Other Services 9.8 7.7 Total Workers 100 100 Source: Unpublished estimates of J. Krishnamurty. -42 - Table 4: THE INDUSTRIAL DISTR BUTION OF THE WORKING FORCE IN UNDIVIDED INDIA, 1901 - 1951 (Percentages) No. Activity Males/ 1901 1911 1921 1931 1951 Females/ Persons (1) (2) (3) (4) (5) (6) (7) (8) 1. Cultivators M 53.2 53.5 55.9 49.8 54.4 F 43.6 41.0 48.1 30.4 45.7 P 50.3 49.6 53.5 44.3 52.2 2. Agricultural Labourers M 14.3 15.4 14.4 19.5 16.4 F 30.2 32.5 28.0 43.8 34,.5 P 19.1 20.8 18.6 26.3 21.1 3. Livestock, Forestry, M 4.2 4.9 4.3 4.9 2.9 Fishing, Hunting and F 2.9 3.2 3.2 3.8 2.3 Plantations, Orchards P 3.8 4.4 4.0 4,6 2.4 & Applied Activities 4. Mining and Quarrying M 0.1 0.1 0.2 0.2 0.4 F 0.1 0.2 0.3 0.2 0.3 P 0.1 0.2 0.2 0.2 0.4 5. Manufacturing M 9.5 9.1 9.0 8.4 9.1 F 11.4 10.9 8.4 8.8 7.7 P 10.1 9.6 8.8 8.5 8.7 6. Construction M 1.1 1.3 1.1 1.2 1.4 F 0.8 0.8 0.8 0.9 0.9 P 1.0 1.2 1.0 1.1 1.3 7. Trade and Commerce M 5.8 5.5 5.9 5.9 6.1 F 3.5 5.3 5.2 4.8 2.8 P 5.1 5.4 5.7 5.6 5.2 8. Transport, Storage and M 1.5 1.7 1.3 1.4 1.9 Communications F 0.2 0.2 0.2 0.1 0.3 P 1.1 1.2 1.0 1.1 1.5 9. Other Service- M 10.2 8.3 7.8 8.6 7.8 F 7.2 5.9 5.7 7.1 5.6 P 9.3 7.6 7.2 8.2 7.2 Total Workers M 100 100 100 100 100 F 100 100 100 100 100 P 100 100 100 100 100 Source: Unpublished estimates of J. Krishnamurty. -43- Table 5: TRENDS IN PER CAPITA INCOME IN INDIA (Rupees) Per Capita Income (at 1946-47 prices) Year Sivasubramoniam Heston (1) (2) (3) 1868-69 - 120 (73) 1872-73 - 125 (76) 1890 - 134 (82) 1900 - 145 (88) 1902 147 (90) 147 (90) 1910 157 (96) 154 (94) 1920 164 (100) 164 (100) 1930 171 (104) 171 (104) 1935 166 (101) 166 (101) 1940 169 (100) 169 (100) 1945 163 (99) 166 (101) 1950 Notes: (1) The figures in brackets in columns 2 and 3 are index numbers of per capita real income (1920=100). (ii) In columns 2 and 3, the 1902 estimates are three year averages; for all other years after 1900 they are 9 year averages. -44- Table 6: TRENDS IN NET DOMESTIC PRODUCT SINCE 1868 FOR SELECTED YEARS (Rs. million at 1946-47 prices) Year Heston Sivasubramoniam (1) (2) (3) 1868-69 30353 1890-91 31843 1898-99 42400 - 1900-01 39929 40792 1910-11 49254 49669 1920-21 44925 46754 1930-31 58388 58291 1940-41 63008 62821 1946-47 66899 65907 - 45- Table 7: GROWTH OF NDP AT 1946-47 PRICES FOR SELECTED SECTORS (Rs. million) Average of Average of Increase Sector 1900-01 to 1904-05 1940-41 to 1946-47 over the period (1) (2) (3) (4) 1. Agriculture (S) 22190 24970 12.6% 2. Agriculture (H) 19737 25966 31.6% 3. Manufacturing (S) 1009 5524 447.5% 4. Small-scale Industry (S) 4200 2783 13.9% 5. Government (S) 1125 5521 390.8% 6. Government (H) 2139 5375 151.3% 7. NDP (S) 43277 64572 49.2% 8. NDP (H) 42244 66422 57.2% 9. Population (millions) (S) 288.0 405.6 40.8% Notef (S) = Sivasubramoniam (H) = Heston. -46- Table 8: LANDOWNERSHIP IN MADRAS PRESIDENCY: LORENZ RATIOS, 1853-4 to 1945-6 Registered Period Holdings Owners* (Pattas) 1853-4 0.63 - 1890-1 0.59 0.54 1900-1 0.62 0.59 1930-1 0.63 0.63 1945-6 0.62 0.60 * Where holdings were jointly held separate calculations have been made to take into account inequalities in individual ownership. Source: Dharma Kumar: Landownership and Inequality in Madras Presidency, 1853-54 to 1946-47 in The Indian Economic and Social History Review, Vol. XII No. 3 (July-September, 1975) pp. 242. -47- Table 9: ESTLMATES OF OVERSEAS MIGRATION FROM INDIA (000) 1834/35 - 1936/37 Year Emigrants Returned Migrants Net out migrants 1834-35 62 52 10 36-40 188 142 46 41-45 240 167 72 46-50 247 189 58 51-55 357 249 108 56-6- 618 431 187 61-65 793 594 199 66-70 976 778 197 71-75 1235 958 277 76-80 1505 1244 272 81-85 1545 1208 337 86-90 1461 1204 256 91-95 2362 1536 790 96-1900 1962 1268 694 1901-05 1428 957 471 06-10 1864 1482 383 11-15 2483 1868 615 16-20 2087 1867 220 21-25 2762 2216 547 26-30 3298 2857 441 31-35 1940 2093 (-) 162 36-37 815 755 59 Total 30191 2394-1 6250 Source: K. Davis: Population of India and Paksitan (Princeton - 48 - Table 10: ESTIMATES OF EMIGRATION TO ASSAM 1882-1904 Year Number of emigrants 1882 22,559 1885 29,398 1890 36,080 1892 56,050 1895 72,837 1897 95,931 1898 49,169 1900 62,733 1902 26,684 1904 24,259 Source: H. Tinker: A New System of Slavery. -49- Table 11: ESTIMATED ANNUAL AVERAGE REAL WAGE (1938/39 prices) Year Real Wage 1900-01 93.96 1905-06 105.57 1910-11 118.92 1915-16 128.66 1920-21 118.80* 1925-26 174.64 1930-31 215.15 1935-36 189.23 1940-41 159.85 1945-46 154.84 1946-47 158.34 Source: S. Sivasubramonian: "Income from the Secondary Sector in India 1900-47" IESHR, 1977, Table 18 pp. 488-489. - 50 - Table 12: REAL WAGE INDICES FOR INDIA 1857-1950 (Base 1916/20 = 100) Year R. Mukherjee K. Mukherjee 1857 134 n.a. 1860 n.a. 145 1870 112 111 1880 121 107 1890 97 107 1905 116 103 1916 100 n.a. 1920 n.a. 100 1930 101 125 1940 165 151 1950 n.a. 171 Source: R. Mukherjee, Economic History of India, p. 58 K. Mukherjee in V.B. Singh (ed.) Economic History of India, 1857-1956, pp. 657-660. Note: R. Mukherjee takes a simple average for skilled and unskilled workers. K. Mukherjee's figures are nine year averages except 1860 which is a seven year average. -51- Table 13: AVERAGE DAILY REAL WAGE OF MALE DAILY LABOURERS NELLORE DISTRICT 1893-1971 Year Official Wages Adjusted Wage Rate (in totals of 2nd sort rise) 1893 165 285 1895 185 305 1901 128 248 1906 121 241 1911 241 241 1916 284 284 1921 270 270 1926 282 282 1931 256 256 1936 169 169 1941 143 143 1946 203 203 1951 252 252 1956 177 141 .1961 185 157 1966 270 216 1971 166 133 Source: M. Atchi Reddy: "Wages Data from the Private Agricultural Accounts, Nellore District 1893-1974" I.E.S.H.R. 1979. The adjusted figures have been obtained by adjusting official figures in the light of information obtained from private agricultural accounts and other sources. - 52 - Table 14: (A) REAL WAGES (IN RS.) IN EASTERN INDIA I II III IV Year Real wages Real wages Real wages Real Wages of miners of miners for coolies of miners in Raniganj in Sodepur in a paper in Sankatoria mill 1887 2.13 2.56 2.84 3.49 88 2.87 2.54 2.63 2.99 89 1.69 2.08 2.01 2.21 90 2.32 2.15 2.14 2.33 91 2.16 2.20 2.19 2.36 92 1.96 1.73 1.76 1.83 93 2.11 1.67 1.64 1.78 94 2.46 1.80 1.83 2.36 95 2.46 2.33 2.21 2.75 96 2.24 1.99 1.78 2.23 97 1.64 1.21 1.36 1.54 98 2.24 1.54 1.78 2.23 99 2.63 1.83 2.33 2.79 1900 2.25 1.58 1.94 2.31 01 1.93 1.70 1.66 2.05 02 1.74 2.25 1.57 2.05 03 1.84 2.47 1.69 2.28 04 2.19 2.60 2.28 2.56 05 2.05 2.40 2.44 2.61 06 1.48 1.91 2.13 1.79 07 1.24 1.67 1.87 1.51 08 1.06 1.49 2.08 1.68 09 1.94 1.78 2.53 3.46 10 2.98 3.17 3.09 3.60 11 2.63 2.50 2.79 2.77 12 2.24 2.40 2.54 2.40 13 2.37 2.10 2.14 2.41 14 2.33 2.08 2.12 2.08 15 2.33 2.03 2.00 2.03 16 2.33 1.71 1.95 1.75 17 3.25 2.18 2.31 2.60 18 4.29 2.44 2.41 3.21 19 1.99 1.47 1.51 1.60 20 1.80 1.49 1.35 1.67 Sourcd: Prices and Wages in India. - 53 - Table 14 (B): COEFFICIENTS OF VARIATION PERCENTAGE Year 1,2 & 3 1 & 2 1 & 4 1887 12.7 9.2 24.2 88 5.2 .6.1 2.1 89 8.8 10.4 13.3 90 3.8 3.8 0.0 91 0.0 0.9 4.4 92 5.6 6.2 3.4 93 11.9 11.6. 6.5 94 15.0 15.5 2.1 95 4.4 2.7 5.6 96 9.4 5.9 0.0 97 12.7 15.1 3.2 98 15.7 18.5 3.7 99 14.6 17.9 2.9 1900 14.3 17.5 1.3 01 6.8 6.3 3.0 02 15.6 12.8 8.2 03 16.9 14.6 10.7 04 7.5 41.8 7.8 05 7.6 7.9 12.0 06 14.7 12.7 9.5 07 16.5 14.8 9.8 08 27.2 16.9 22.6 09 15.5 4.3 0.0 10 2.5 3.1 9.4 11 3.5 1.2 2.6 12 5.1 3.5 3.5 13 5.4 6.0 0.0 14 5.0 5.7 5.7 15 7.0 6.9 6.9 16 12.8 15.4 14.2 17 18.5 19.7 11.1 18 28.8 27.5 14,4 19 14.2 15.0 10.9 20 12.1 9.4 3.8 APPENDIX ON CASUAL DAILY LABOURERS Casual daily labourers supplied not less than 70% of the hired labour on fields between 1951-71 in the district of Nellore.- The day-to-day labour market is the most active in their case. In busy seasons, e.g., harvest, these labourers are in great demand. Most of them, male and female, generally gather together into a temporary union in one place of the village and begin to bargain with the employers. Sometimes, the employers form into another group and settle a rate of wage in consultation with the leading members of the labourer's groups. The rate depends upon the urgency of the agricultural operations. For instance, the expectation of early rains which could spoil the harvest raise the rates to enormous proportions, though for a short while of a day or two. This procedure has been followed in many of the delta villages from 1960 if not earlier and in some of the dry villages since the year 1969 and is more frequent at harvest times than the rest of the year. In such cases where the labourers enter into such bargains ususally no perquisites like meals are provided. The links between the employees and the employers are very slender and last only for that day of work. In the delta villages of Nellore and Kovur taluks this method has been resorted to in the case of all agricultural operations other than harvest expecially when labour scarcity has been felt in that region. The daily bargaining for wage rates which was rare before 1969 began to become common since that year due mainly to the prevalence of this system of temporary unionism. The consequent result was not only the suspension of all perquisites, but the employer was freed from his moral 1/ Village surveys revealed this fact. obligation of supporting such labour families in times of stress or giving advance wages and low interest loans and gifts. There is another method said to be followed in most of the dry villages of the district at present (1971-75) and in most of the delta villages since 1958, with regard to the casual daily labourers. In this case the daily rates of wage are fixed for the season by a joint meeting of the prominent employers and workers; the nature and amount of the perquisites to be given are left to the discretion of the employer, such rates are being revised from season to season in the recent period, i.e., since 1958. In abnormal cases such as a sudden change in the prices of grain or a sudden change in the volume of supply or demand, they may be revised once in a few weeks or days, but not daily as in the first method. Here employers meet changes in their demand for labour in different volumes by varying the perquisites. For example an employer who needs a large number of labourers may offer today as an extra perquisite while another who wants to reduce the applicants for work may ask the workers to provide their own meals. Another method said to be of recent origin, followed in some delta villages after 1950, is the middlemen system. Where the operations involve more than 50 labourers per day the farm owner enters into a contract with a middlemen called 'maistry'. The maistry enters into a contract with the landlord for a fixed sum (grain or money) and employs his own labourers. In the maistry system generally no perquisites or meals are provided by the employer, nor is he concerned with the rates of wage or the hours of work put in by the labourers. The maistry is responsible for maintaining the quality and quantity of work, and he keeps for himself the difference between the wages he paid to workers and the sum he receives from the farm owner as his profits. - iii - It is said that this method has been largely followed since 1960 mainly in cases where the labourers are willing to work for a longer day in order to get higher wages than the market rates and also where the employer fails to attract the labourers in required numbers. Presidency and Nellore Large and influential landowners in the Presidency having a command over the labourers could get their services for lower rates than the prevailing market rates. The Board of Revenuefound in 1908, "Special rates paid by landholders and other persons who have special influence over the labour market and are able to command labour at less than the usual rates, were found to prevail... .pecial rates as low as 6 pies in cash or 1 - seers of grain have been reported 2 o by the Revenue Inspectors." 1/ In the district of Nellore there is no evidence that the big landlords did deliberately underpay the daily labourers, but the presence of a large number seasonal and vruthi daily labourers in most of the villages of the district might have checked the scope of competition among the employers in hiring agricultural labourers. On the other hand, in some villages whenever there was competition among the employers it was common to bid up the rates of wage. Butterworth found in the village of North Mopur (Nellore taluk) the ordinary wage per day in 1900 was 1 Munta (7 lbs.) of paddy. But, "sometimes, there is keen competition for these labourers [Weeders] and the wages rise to a munta and a half".2' 1/ 1908 Qnl. wages Census, Form B. 2/ Butterworth Report, para 15. Source: Achi Reddy U - iv - Even after 1900 no reference is made in any official report to such underpayment of wages in this district. According to the oral evidence obtained in 1974, the influential landlords can only employ a large numoer of seasonal labourers but cannot pay lower wages to their worker than what their counterparts receive from other employers.