Historiography

The foreman in the plantations and the mills has received much attention from historians of labour.Footnote ³ In analytical narratives, agency has been discussed from the perspectives of labour control and social inequality. In particular, two theories of agency are more prevalent: needs of capital and needs of labour.

According to Morris, agents were needed inside the Bombay mills to reduce supervision costs.Footnote ⁴ Agents were also needed to meet temporary shortfalls in supply of labour. Even when direct recruitment was practiced or preferred, the sudden need for extra hands could not be met by direct recruitment, which involved time. Chandavarkar suggests that such shortfalls arose in the 1920s and the 1930s from fluctuations in demand.Footnote ⁵ According to Mazumdar, shortfalls arose because the supply of stable labour was inadequate at the going wage.Footnote ⁶ Shortfall could also arise for technological reasons. For example, casual labour was used mainly in certain low-skilled tasks inside the mills. Ginning, blowroom, spinning and carding relied on casual labour more than the weaving sheds.

On the other side, in the plantations and docks, where labour agents were both powerful and pivotal, needs of capital were of a different kind. Plantation labour was anything but casual. And, while agents performed supervision to some extent, supervision costs could not have been high, given that the work was not highly skilled, payments were on piece-rate basis and it was not easy to shirk because surveillance systems were both efficient and harsh. It was not easy to quit a tea garden job either given their remote location and laws against desertion. In this case, quite clearly the main reason behind agency was search cost. Agents sprang from missing markets, meaning a context in which the employers did not know where labour was available at what wages. Needs of capital could also stem from a persistent cultural failure that vitiated all forms of modern enterprise of that time. The owner of a large business, whether European or Indian, was unable or unwilling to communicate with the workers. For both these reasons, direct recruitment with paid labour officers was not considered a serious option. Assam plantations illustrate this syndrome particularly well.Footnote ⁷

While these theories of the agent do not consider why the workers might want this system to continue, Chakrabarty proposes a different kind of cultural explanation based on what systems of authority the workers themselves wanted or were used to. In the jute mills in Eastern India, he writes, ‘the respect that the sardar commanded had roots in certain other aspects of culture to which both the sardar and the worker belonged. It derived from a precapitalist culture with a strong emphasis on religion, community, kinship, language, and other similar loyalties’.Footnote ⁸ The workers had economic needs too that the sardar met. Search costs were high for the workers, in the nineteenth century because they had too little knowledge about the mill or the plantations world; and in the twentieth century because by then it was difficult to get an entry into the workplace without the agent.

‘Needs’-based explanations of the labour contractor, however, are not entirely satisfactory. One problem is the absence of a theory of where needs sprang from. Some of these needs, for interpreters for example, should be universal to many early factory sites, but did not necessarily give rise to a labour-lord of the South Asian kind. Furthermore, needs change and vary, and the fit between an institution and demands that arise in particular contexts is not necessarily close. For example, the agent persisted in Bombay despite numerous complaints about the adverse effects of agency on the workplace. In 1940, 450,000 workers in India did the work (processing 3.5 million bales of cotton) that took mills in Japan 190,000 persons.Footnote ⁹ Comparisons of a similar nature were made between Dundee and Calcutta on the efficiency of the jute labourer, to Calcutta's disadvantage. These concerns of contemporary observers connect with a view advanced recently by Clark, that there was an irreducible labour efficiency difference between India and other industrial economies.Footnote ¹⁰ It has been argued, in particular, that the competitive weakness of the Bombay mills in the late-interwar period derived not from management failure, but from the fact that labour was of a quality poorer than that of manufacturing labour in Japan, Europe and North America.Footnote ¹¹ The agent was at best ineffective in mitigating, and at worst the cause of this syndrome. What, therefore, we have to explain is why agency was so pervasive and persistent despite being against the needs of capital at times. For another example, needs-of-labour arguments face the criticism that agents frequently misused authority. Scholarships on both plantation labour and mill labour acknowledge that agents extracted rents from the workers, although different authors see these transactions somewhat differently.

Chandavarkar's now classic critique of linear models of industrialization suggests further difficulties with viewing the labour-lord as a precapitalist relic in capitalist workplace.Footnote ¹² First, the latter approach amounts to accepting that industrial capitalism followed one trajectory everywhere, and that those features that did not fall in that universal pattern were survivals of some kind. Instead, we need to recognize that capitalists could choose between different models, drawing upon locally available resources and experiences, giving rise to different trajectories along which modes of labour organization could evolve. Entrepreneurial choices were ‘ceaseless improvisations’, suggesting that labour-management systems were neither a ‘technologically determined process’ nor a survival. Second, the representation of preindustrial society in these approaches is often one-dimensional, proceeds by opposing the pre-industrial to the industrial, and thus does not allow for the possibility of congruent and inter-dependent development of both. I find this critique a very productive one.

There is indeed a case to suggest that the insertion of an agent in modern enterprise was not a totally new system, but an extension of contemporary models of labour organization, such as the institution of village headman and the putting-out master-artisan. Historical sources do suggest examples of the headman being absorbed in supervisory roles in plantations, whereas the handloom weaver elite enjoyed elitehood inside the textile mills. Agency in the plantations and mills can be seen as incorporation within the so-called modern enterprise of particular forms of a larger and established practice of putting out. The labourer in this situation was ‘more in the service of the Maistry than of the firm’, in the words of one contributor to the migration literature.Footnote ¹³

That being said, this incorporation must be understood also in relation to the dynamics of late industrialization. A labour-lord rather like the South Asian jobber was indeed present in many regions experiencing late-industrialization.Footnote ¹⁴ Significant examples from other contexts suggest that late industrialization might make a convergence of agency roles such as those of creditor, negotiator, recruiter, banker, and foreman, more likely. The superimposition of a modern sector in the backdrop of largely traditional agriculture and handicrafts, in which the household was the main organization, the consequently weak, missing, or context-specific labour markets and quite strong social barriers to labour mobility, would have needed institutional response of some kind. That particular environment raised the costs of searching for labourers suited to mass production sites, and also raised costs of training, supervision, and welfare. These high search, retention, and training costs required different types of authorities to join together, or for one kind of authority to spill over into another sphere. The jobber was an example of the convergence. The sardar in the Assam tea estates was an example of the spillover.

Spillovers and convergences of a similar nature seem to come up in different contexts in a late-industrializing society, and it would perhaps be appropriate to ask whether or not these examples arise from a propensity that is far more general than labour historians have so far considered. Business history shows that groups that deliberately interlocked enterprises and moved away from Adam Smith's model of specialization and division of labour were common in both India and Japan, common enough to lead to an attempt to theorize the formation of groups less in culturalist terms and more in terms of the economic need for interlocked operations.Footnote ¹⁵ In the quite independent context of rural markets, imperfections and missing markets are seen to give rise to individuals who integrate transactions of different types.Footnote ¹⁶ To a large extent, these examples of interlocking, integration, convergence, and spillovers can be explained by the fact that commercialization in a traditional society typically occurs in a fragmented manner. Not everything commercializes at the same time at the same pace, and input markets are often undeveloped or absent even as commodities become marketable. With labour, we have a counterpart situation wherein work becomes marketable, but the worker remains outside exchange relations. These anomalies of rapid and late commercialization are well-known. What is perhaps in need of emphasis is the corollary that these anomalies create space for individuals already enjoying economic or political power of some kind to try to become monopolistic suppliers of those key inputs in which exchange relations lag behind. Such systems were as much a gain for the buyers of goods or inputs as it was for the sellers. Much as interlocked rural markets conferred special monopolistic privileges on the merchant-creditor, interlocked authorities in the labour market were, on the one hand, a gain for the monopolistic labour-controllers, and on the other hand a gain for the employers by cementing labour to the workplace more firmly than either a pure foreman or a pure recruiting agent or a pure headman might have done.

And finally, the incorporation of traditional authority in this fashion into modern work sites carried costs in that the intermediary could take advantage of the inadequate knowledge of the employer about the qualities of individual workers. In other contexts of early industrialization, employers tried to replace labour-lords with foremen and managers as industrialization matured. The transition was not an easy one, indeed a ‘protracted battle’.Footnote ¹⁷ In Bombay, these attempts coincided with a stressful adaptation to increasing competition. Precisely because recruitment and supervision via a worker-intermediary represented the choice of an institution, rather than the decision of isolated capitalists to hire isolated contractors, it was hard to get rid of. An isolated agent serving the needs of capital could be fired, but replacing an institution that had acquired legitimacy, that represented a set of rules followed by workers, managers, and employers for a considerable time, was a much harder task.

The three short studies that follow illustrate these general points. The first deals with a wide range of nineteenth-century contexts, and suggests that the labour agent could be re-created out of older forms of economic and social relationships.

The Labour Agent in Nineteenth-Century Contexts

It is useful to go back to the emigrant indentured workers of the early nineteenth century; for in this sphere we can glimpse an ‘origin’ of the sardari system. The phrase ‘Hill Coolies’ appears in the very first official documents prepared by the British Indian governments on emigration. Who were they? A statement describes them as follows:

Forest-dwellers living on semi-arid hill slopes appeared to be more sensitive to new wage-earning opportunities than the settled peasants who were averse to leaving their land. By hill coolies the shippers and their agents most commonly understood the particular caste or profession known as Dhangars (later called Oraon). Dhangars indeed were prominent among those who left for overseas work, though they were possibly outnumbered by all the other castes together. They were also prominent among the migrants headed for the tea estates in Assam and Dooars. The uncertain cultivation prospects in Chota Nagpur drove many to labour. But there was more to the connection between Dhangars and wage-work than this push factor. By 1838, the indigo boom was already several decades old. In Bihar, the Dhangars were well-represented in the indigo factories, and known to be good workmen. The indigo planters preferred them, and this preference carried over, by channels we can only guess, to the European shippers of migrant workers in Calcutta and onward to the planters in Mauritius.

Groups of hill tribes, it seems, already had headmen when they moved. In later reports on cultivation in the region, there is an explanation why this was so:

In some ways, therefore, headship could be a carry-over from communal land organization. The worker groups that came from villages that carried a sense of communal ownership and control over resources, including labour, perhaps formed cooperative communities more easily. The institution of a headman was inseparable from that of the community. A cooperative community was a non-market institution, but precisely for that reason, it needed a representative in dealing with the market outside it. The headman was this channel of negotiation between the community and the market. Even when the headman himself did not join migrant groups, the concept travelled in a modified form.

A second, and quite different, example of endogenous root of headship occurs in South India. Almost all recruitment to Ceylon tea estates was done without indenture by the kangany system, which was seen as recruitment of a ‘purely patriarchal character’, giving rise to a labour force ‘subdivided into a number of smaller groups, each under its patriarch. . .; and the family principle is manifested in the groups which are under these minor headmen’.Footnote ²⁰ The group was more a social unit, even a work team. It held its earnings together, and had joint liability for advances made to them. The headman-cum-group principle gave financial security to individual members. The head kangany in an estate was in charge of the wage-bill of the workers in an estate. The fact that the kanganies received from their employers or at least some of them, a commission for each day their recruits worked, gave ‘the introducer of the labourer . . . a financial interest in the welfare of his protégés.’Footnote ²¹ The financial relation between the kangany and the workers under him or her has received much attention in the literature. It appears that money cemented these bonds in perhaps a more systematic way than in the other prominent examples of the foreman. It cannot be doubted that workers needed the sardari authority to protect themselves from predators. At the same time, shades of debt bondage have also been found in the relationship.Footnote ²²

Not all sardars were headmen. Hill coolies, first of all, did not form the majority among emigrant groups. Kanganies could be found only among Tamil labourers who moved in specific destinations, principally, Ceylon, Malaya and the Straits settlements. Many, indeed, were individual migrants without a team. Reports of raids on illegal labour depots in Calcutta in the 1830s suggest the picture of a chaotic human mass, rather than an organized and hierarchical one with a clear leader among them. It is likely that in many situations the institution was re-created out of the needs of capital and the needs of labour with reference to a notional tradition. What were these needs?

In some cases from the 1840s, the groups appear to have elected leaders. Of groups that were found travelling in Arcot in search of work, it was written

The journey needed one person to receive and distribute food rations, and was again conducive to the emergence of headmen. In the return journey from Mauritius, sardars performed an important function, to keep charge of the savings brought home by the individuals. If the onward journey reported many individuals who were on their own, almost all individuals belonged in a group on the return journey, and followed a headman. The logic of this organization had to do with keeping money in safe custody. Of the 200-odd returnees who were interviewed by the police and port officers at the Calcutta port in the winter of 1841–42, almost every individual carried a sum of money. In view of reports of frequent thefts and the operation of a thriving underground economy in the lower deck, a variable mixture of sardari authority and teamwork would have been crucial to the safe-keeping of money. The sardars kept the money or kept accounts. Sardars had reason to keep trust. They were potential returnees, and one tainted by allegations of fraud had less chances of returning. It needs to be added that aboard a ship the sardars had reasons to be fearful of their wards if something went wrong with the money.

Increasingly as the nineteenth century moved on, we see sardars being selected by the employers and their subordinates. The employers, their agents or subagents needed several qualities at once: an interpreter, one who could keep charge of money and make payments, and one who had enough influence on the group to make them work well and to minimize chances of desertion. The communication angle is illustrated in many later examples. Knowledge of English, which usually meant previous service under a European or in the army, made some migrants suitable for the job.Footnote ²⁴ Even in groups that did not necessarily have sardars, the employers chose one man to do the job. In a linguistic melting pot, communication was an important function for the workers too. It was for this reason that the original batches often continued intact under a sardar.

In the sugar plantations, the sardari system settled down to a supervisory one, though retaining through the transition its connection with the function of keeping accounts.

Keeping accounts is one function that recurs with remarkable consistency in nearly all reports of the sardar-cum-supervisor from the nineteenth and the twentieth centuries, whether in mines, plantations, small-scale industry, and in a modified manner, in the mills.

Sometime about the middle of the nineteenth century, the historical process that led to the formation of sardari authority changed in Mauritius. Net immigration reached a peak in 1854, and from 1861, Indo-Mauritian population began to outpace immigrants. Increasingly, the sardar was an ‘old immigrant’, rather than a headman among a fresh batch of recruits. Sugar milling being a seasonal activity, and there being a ready pool of settled workers now available, the mode of hiring shifted towards casual labour by the late nineteenth century, with the result that some old immigrants now became labour contractors. In other examples of migrant communities too, we observe this transition: from headman to contractor.

Thus, any tendency towards strategic formation of collectives must be seen together with a tendency towards formation of hierarchies. Migrants were, and are, always divided and hierarchized. Sardars held the team together and by doing so rose in rank. Among early immigrants, through the 5 or 6 years they lived in Mauritius, there were instances of individual mobility. People set up small businesses, moved from the sugarcane estate to household jobs, some older immigrants were given land, which eventually became the foundation for a large small-planter society, and some were promoted to headship. Strong patronage relations between the sardar and the European overseer could become an avenue of mobility.Footnote ²⁶ These tendencies loosened the team. In formation of hierarchies among the migrant communities, old immigrants who could speak creole, and returnees were placed in an advantageous position.Footnote ²⁷ Their superior knowledge of working conditions and where jobs were available were indispensable information for potential migrants, whereas on their power of persuasion the employers relied for fresh recruitment of good workers. As Carter has shown in the context of Mauritius, returnees played a significant historical role in increasing information in the public domain, and in transforming the overseas destinations from an unknown to a known world to potential migrants by the last quarter of the nineteenth century.Footnote ²⁸

The authority of the sardar also derived from a persistent worry about quality of workers. Ironically at the same time, questions of efficiency of the sardari recruitment system were raised. In some colonies the law on the infringement of the indenture had to be invoked too often. Some quite vocal returnees ‘gave the colony a bad name’.Footnote ²⁹ The problem partly was that from the more successful settlements, in the Caribbean, Fiji and East Africa especially, where a section of the settlers progressed from unskilled manual labour to trade and skilled manual labour during the nineteenth century, the scale of return was smaller compared to Mauritius, where avenues of mobility were limited. But these points do not explain away the persistent and long-standing complaints about quality.

When we gather together these complaints across different temporal and spatial contexts, the definition of quality can be shown to shift. There were four senses in which quality of worker or similar terms were used, and these were vagrancy, shirking, industriousness, and mobilization. The mid-nineteenth-century employer in the tea districts was clearly worried about vagrancy. His or her counterpart in the sugar plantations elsewhere in the world was concerned about shirking and industriousness. The mill managers in Bombay of the 1930s were more concerned about hidden trade union elements. In the official view, different classes of people migrated in the nineteenth century, and some were better able than others to withstand the discipline and rigour of agricultural labour in the plantations. Some were escaping law or the harshness of life and livelihood in their places of origin, and these were more prone to shirking and complaining about the routine. A certain number of ‘priests, barbers, jewellers, and men of that kind’ did voluntarily enlist with the agents, and they were, some official accounts believed, the complaining type.

The agent was closely connected with the perceived problem of quality. Throughout the nineteenth century, any returnee sardar could earn handsome amounts by sending fresh recruits. In the views of the employers, some took unfair advantage of the offer. ‘They [the returnees] also send back before their indentures are over persons of the wastrel and loafer class of whom the colony wishes to rid itself’.Footnote ³⁰ To get a handle on the problem, which would mean being better-informed than the sardar, employers tried to reduce industriousness to a few social markers. This was a longstanding preoccupation in colonial records, though the constant grappling with social markers of good work did not result in a better system. In the early nineteenth century, the ‘junglies’ or forest-dwellers were preferred to peasants, and the commission in recruitment reflected these preferences. In the second half of the nineteenth century, when the tribal element among migrants had fallen in response to high mortality among this class, ‘Madrassies’ and North Indians were similarly classified, in a manner that clearly reflected a preference for the upper castes.Footnote ³¹ Yet, Brahmins were frequently clubbed with undesirables when it came to industriousness.

Problems of selection were persistent with recruitment in the Assam tea gardens as well. The Assam example illustrates how different functions, leader and recruiter, could be made to join together in response to perceived problems of adverse selection.

Tea Estate Labourers in Assam Tea Plantations

Almost contemporaneous with the major burst of recruitment for Mauritius and the Caribbean, there began systematic recruitment of labour by the tea plantations in Assam. The number of workers in the tea estates increased from less than a hundred thousand in 1870 to six times that figure in 1900. Recruitment systems differed between the three major plantation clusters, the Dooars in northern Bengal, Surma Valley in easternmost Bengal, and the Upper Assam. The two former regions were more accessible than Upper Assam, and faced much less difficulty in recruitment. Both these regions consequently could afford to experiment with the system of recruitment. In both, free or non-indentured labour played a larger role. For Upper Assam, however, problems of supply were more serious.

In 1905, Sir Bampfylde Fuller, chief commissioner of Assam, in a circular addressed to the Indian Tea Association, asked why tea garden labour had become ‘markedly unpopular’, and the labour force recruited was ‘both intractable and inefficient’.Footnote ³² Two connected factors seemed responsible. First, the labourers were engaged on 4-year contracts, running away from which invited arrest and trial. And second, most labourers were recruited by agents who did not fully and correctly inform them of the working conditions in the plantations or the terms of employment. This view was far from exception at the turn of the century. Not only was it widely held, it was held across class and political boundaries. Influential voices among planters in Assam, Dooars and Surma Valley acknowledged the ‘serious abuses’ that labour recruitment systems entailed, and at the same time saw themselves unable to live without these systems.

Sir Bampfylde spoke at a time of transition in the gardens of Upper Assam. About the middle of the nineteenth century when plantations in Assam and Dooars began to grow, the region was sparsely populated, had bad communications, and no local labour pool at all. Indeed, the difficulty of communication was legendary, as the following description of transporting tea in Assam about 1845 shows:

Not surprisingly, finding labour for such destinations as these would pose a challenge. Large flows of overseas migration from Chota Nagpur had by then begun. And, it was inevitable that the planters would consider a system very similar to the one already in place for supply of labour to Mauritius.Footnote ³⁴

In the 1840s, recruitment was done by professional labour agents or arkatis. Many signed an indenture on a wrong understanding of what the job entailed, where they were going, and what their chances were of returning. A general and acute labour scarcity, met with workers who felt cheated even before arrival, gave rise to demands for securing labour more firmly to land. The solution was a long (5 years to start with) indenture contract at the plantations end. The distinctive feature of the Assam contract was that it was drawn between individual workers and the planter, rather than between the contractor and the planter, as in the case of Madras and Ceylon. Liability of fulfilling the contract fell on the individual worker, and to secure this obligation an act was passed that made it possible for employers to penalize and bring back the deserting worker, the Workmen's Breach of Contract Act, better-known as the Act XIII of 1859. On the other side, the Inland Immigration Act of 1863 made it mandatory for all contractors to be licensed. The duration of the contract was shortened too.

Shortage, however, continued. By the third quarter of the nineteenth century, increasing competition for large-scale recruitment had made this traffic more difficult. Development of railway construction and mining in the ‘catchment’ areas raised the difficulties of getting workers at earlier wages. The contractors had to travel further to obtain workers. Landowners were generally against emigration. Some charged unauthorized commission for allowing recruiting agents into their territories. Others spread rumours with the help of the police.Footnote ³⁵

Desertion and abuse continued too. To allow an alternative to indenture, an act passed in 1882 enabled ‘free’ recruitment by unlicensed contractors. In practice, most free labourers were also put on a contract directly when they reached Assam. The so-called free labourers did not often see the difference, and the recruiters did not try too hard to explain it to them. These labourers were free until they reached depots in Dhubri or Dibrugarh, river-ports where the contracts were signed after considerable haggling over commission between the garden agent and the contractor's men, and before they were taken to the gardens.

In 1900, going to Assam had acquired an adverse image in lower Bengal and Bihar. The recruiters needed to be both tough and wily to do their business. The business attracted such people.Footnote ³⁶ For example, the free recruiters told the prospective worker that he or she would be employed in the coal mines located in the same larger region. On such promises, the party came to Raniganj, a mining town in the Burdwan district of Bengal, which contained a ‘depot’ or waiting station for prospective emigrants. At this stage, the party had to appear before a Magistrate and the worker was required to state that he or she was going to Assam willingly. Upon learning the truth, sometimes the labourer tried running away. The manager and the subagent tried to force the person back to the courtroom. If the labourer gave evidence at this stage that he or she had been forced, the manager could be prosecuted. However, prosecution was rare because the labourer's will gave way at some stage in these proceedings, especially in a situation in which the underworld was patently interested.Footnote ³⁷

And yet, reports of recruitment probably exaggerated the involuntary nature of the migration. During famines and epidemics, Assam received a large number of migrants who wanted to escape starvation. Neither wage nor the quality of life was actually worse in the plantations.Footnote ³⁸ The evidence of desertion does not suggest that desertion was ever a serious problem. In 1877, deserters as proportion of the mean labour force formed 5.7%; in 1901, the percentage was 3 (the proportion rose again later, in anticipation of the end of the Act XIII).Footnote ³⁹ By comparison in the mills, the incidence of ‘absenteeism’ was far worse, at 10%–15% of the workforce on an average day.Footnote ⁴⁰ At the end of the nineteenth century, actual instances of punishments under the specific provisions of the 1859 Act were as low as 0.1%–0.2% of total garden labour force.Footnote ⁴¹

The adverse reputation of Assam had built on two connected problems, the notion of the region as a land-of-no-return, and scarcity of women in the plantation society. The problem was not that a large number made a serious attempt to escape. The problem rather was that very few did. Many renewed their contracts voluntarily; to their people back home, these individuals vanished permanently. Upper Assam in this way ‘tied one down’.Footnote ⁴²

The scarcity of single women added to the perceived horrors of plantations society. The gardens employed a large number of women, nearly all being married when they arrived. Few among the single male migrants could find a wife in the local society. Few could return home easily to marry. Few had the bride-price in hand. Single women rarely volunteered for enlistment. After free recruitment began in 1882, single women needed to furnish a witness certifying that they were going of their own volition. Women accompanied by male relatives were hired, but upon paying a commission to those relatives. Single women migrants were often desperate individuals who ‘only wanted to get away’.Footnote ⁴³ Others were cajoled or tricked.Footnote ⁴⁴

About 1880, the feeling among planters was that, whereas the professional contractors were good at supplying large numbers quickly, their methods ended up gathering ‘the sweepings of the bazaars’.Footnote ⁴⁵ With its bad reputation behind, Assam had become the ‘bankruptcy court, the divorce court and the poor-house of the down-country districts’.Footnote ⁴⁶ Planters were routinely left with individuals who had no experience with cultivation. There was little a planter could do to coerce the contractor to minimize the risk of shirking. ‘My chief objection to the contractor system was that the contractors recruited for Assam as a whole and not for individual concerns, the result being that . . . unless we took the bad with the good, they used to refuse to supply us.’Footnote ⁴⁷

The conditions of the contractor-recruited workers of Upper Assam appeared especially unsatisfactory compared with ‘sardari coolies’ or workers hired by foremen. By contrast with the unhappy single male migrant homesick to an extraordinary degree, they lived in a society that had a larger number of friends, relatives and women. This had already been the practice in the Surma Valley. By 1880, an ideology of the ‘headman’ took roots in the European planter circles. ‘The best way of working with the natives of the coolie class is to deal with them through headmen who understand their likes and dislikes in a way which no European could’.Footnote ⁴⁸ Sardari recruitment, in this way, solved many problems at once, that of getting labour, getting stable labour, the challenge of creating a new society, and facilitating communication. The example of the kangany system also influenced this perception. Based on somewhat idealized descriptions of the sardar, the Assam Labour Enquiry ‘made a recommendation—which runs quite contrary to all received thought on the subject—that estates should work through headmen and gangers’.Footnote ⁴⁹

Who were the garden sardars? They were senior workers, who sometimes rose from the ranks, sometimes were established headmen. In all cases they were ‘the working head of his own settlement of coolies and he would see to their welfare’.Footnote ⁵⁰ The expectation from the new rules was that a sardar, sent back home with a license and monetary incentives, would recruit from immediate relations and friends, and thus would serve two purposes at once, training and social stability. There were other advantages besides this combination of supervision, training, and recruitment. Sardari workers were cheaper to obtain. The contractor spent money on search, but the sardar did not have to.

In 1915, the sardari system was the only legalized system of recruitment for Upper Assam. The change was such a thorough one that even the employer ‘cannot himself legally engage labour in a recruiting district; he must do so through the agency of his garden sardars’.Footnote ⁵¹ Equally, a labourer who wanted to voluntarily join the garden had to go through a sardar. A labourer who left home could not return to Assam without a sardar. With all competition blocked, and labour demand still brisk, the sardar inevitably turned into a contractor. The triumph of the new sardari system, thus, brought on the collapse of the ‘true sardari system’ as it existed in the late nineteenth century. ‘Such men’ as the new sardars ‘are obviously professional recruiters and not genuine labourers. Not infrequently they are sent to districts other than their home districts’. In one case of inspection in Basti, United Provinces, it was found that a large number of recruitments had taken place at the railway station. Search costs were still very high, and so was the commission offered, so that the sardars had incentive to switch effort from training relatives to recruiting from far and wide. The new law that prohibited entry of anyone into an estate unaccompanied by a sardar virtually killed the labour market in the gardens of Upper Assam.

The ‘true sardari system’ died because it was particularly incapable of reducing search costs. Sardars could recruit only on a very limited scale. All said and done, ‘[T]he sardar is an amateur whose operations are confined to the neighbourhood of his own village. The recruiter is a professional whose field is the whole district’.Footnote ⁵² The usual practice was to call in a worker who was about to retire, and send the person home in search of workers. Many treated the occasion as a paid holiday, which was a luxury in the tea garden world. Equally, a planter often treated a demand for a holiday as a chance to recruit new people. Some never returned. Some returned to join another garden on a false identity. The sardars thus sent home did not try too hard to search for people. Interviews of retired sardars at the Royal Commission verified what planters and officers often complained, that the old-style sardar succeeded in getting not more than two to three people on average in a lifetime consisting of three visits home. The old-style sardari had by then exhausted its potentials. ‘My sardars failed because they had recruited all their relations and friends and so had exhausted their connections’.Footnote ⁵³

While the old system died, the new sardari system brought back the old problems of selection. Once again, recruitment began to ‘result in a considerable number of undesirables being brought right up to the gardens’.Footnote ⁵⁴ But by the early interwar period, the connotations of quality had changed. Desertion was still a problem, but on a much smaller scale. Assam did not evoke the fears that it had two generations before, at least partly because returnees had increased in number, and as with Mauritius, ‘every migrant on his return became an apostle of migration’.Footnote ⁵⁵ The emerging problem, rather, was mobilization. Each sardari worker was bound by social ties to one sardar, and loyalties were not easily transferable. The wages of ordinary workers, therefore, barely rose between 1910 and 1930, even as sardari commission remained high. Workers who appeared before the Royal Commission on Labour in India (1929–31) testified to the harsh conditions of employment inside the gardens and persistence of low wages.Footnote ⁵⁶ Managers admitted that the low wages worked as disincentive to work.Footnote ⁵⁷ Beatings were common when a worker turned out to be ‘lazy’ or recalcitrant. Debts secured the sardar's control over the workers. The sardar discriminated between ‘his own men’ and the others he was asked to supervise, and cultivated the belief that runaways could be beaten and arrested.

Assam, in other words, moved from indenture to an internal putting-out of effort, by merging supervision and recruitment. While a response to the adverse reputation that Assam indentures had acquired, the joining of recruitment and leadership functions solved one set of problems to raise another. The contradictory and dual character of the labour agent, the persistent trade-off between getting numbers and getting quality, is again illustrated in the example of Bombay cotton textile mills.

The Jobber in the Bombay Cotton Textile Mills

The counterpart of the sardar inside the Bombay cotton textile mills was the jobber. Significantly, the headman-cum-foreman inside the Calcutta jute textile mills was called sardar, in a milieu that had a considerable presence of workers from Chota Nagpur, where the term originated. In both these contexts, the labour-agent was rather more indispensable as supervisor than in the plantations context.

The most influential recent study of the jobber deals with a time when the power of the jobber was in decline, late in the interwar period. This was a time when, ‘subject to the pressures of both the management and the labour force, jobbers frequently found themselves in the jaws of a nutcracker.’Footnote ⁵⁸ The elements that made it so were a competitive environment in which Bombay was under severe pressure, a legislative environment in which the dispute settlement process depended less than before upon informal negotiators, and diffusion of centres of power within the worker ‘neighbourhood’, the key concept in this work. These elements, however, were becoming significant variables in the interwar period. Historical accounts dealing with an earlier period would suggest that the jobber's power was growing and less challenged from inside and outside in the late nineteenth century.

The early history of the Bombay jobber remains unknown. While mill construction began in the 1850s, the growth of the industry was shaken by the 1860s cotton famine and subsequent crash of the world cotton market at the end of the civil war in America. It seems likely that the jobber did not play a major function in the first phase of the industry. When older insiders looked back to the nineteenth century, they saw the early jobber rather as a recruiter of labour. They brought fellow villagers into the city. The jobber, in other words, was a particular response to the absence of labour markets in the mid-nineteenth century. The jobber was needed in the early years of mill development, Morris suggests, also because of a linguistic, cultural and communication gap between the managers and the workers. A senior worker who spoke the language of the ordinary worker and came from the same social background, and yet could communicate with the managers, bridged the gap.Footnote ⁵⁹ Morris also suggests that the jobber became transformed from an agent of communication to an agent of command. The institution survived because the jobbers were found to perform a useful managerial role, namely, supervision and enforcement of discipline.

By the prewar decade, the jobber had clearly become a feature of the shop-floor. A ‘Retired Mill Manager’, wrote at the turn of the century,

Despite being in charge of machinery, the jobber never became exactly a foreman, but remained an under-supervisor, a medium between the workers and the technical men. The jobber did maintain the machines, repair small problems, but was still not seen as equivalent of technical men. He was answerable to the ‘Lancashire managers and their Parsee assistants’ for keeping the machines in good condition, performing tasks in time, and supervising work teams under him. The jobber was rather like a putting-out master artisan inside the factory. The jobber, or his female counterpart the naikin, was in charge of completing tasks, and had a fair degree of freedom in hiring and firing the work team that he or she deployed to complete tasks. He or she was also in charge of the training of these workers. The recruitment function, however, carried on, and employers depended on the jobber heavily during episodes of labour shortage. For example, cotton textile mills expanded rapidly in Bombay after the 1890s; there was expansion particularly in weaving as the China market for cotton yarn was lost to Japan. This move, together with a plague epidemic, caused sudden shortages of labour. The jobbers toured the neighbourhoods and tried to gather the needed extra hands. Young adults from the nearby workers quarters were extensively employed on ‘half-time’, and formed a pool from which sudden needs could be met.

Early in the interwar period, according to Chandavarkar, the jobber performed multiple roles, ‘recruitment, technical and disciplinary functions’. But in many ways, the figure had evolved away from the image we have of the jobber based on nineteenth-century descriptions. In particular, in the 1920s the jobber had acquired a rather unsavoury reputation for exploiting the hire-and-fire privileges that the mill managers conferred on him. The leftist trade unions were vehemently critical of the jobber. To interpret this syndrome, we need to look more closely at the difference in recruitment pattern between the prewar and the interwar periods. At 1920, the jobber was no longer supplying ‘villagemen’, nor supplying particular segments such as young adults, but selecting casual labourers already available in the working-class neighbourhoods of the city. By 1920, children's work had been effectively banned, but the availability of unemployed adult male hands right outside the mill gate had increased. The early-interwar period saw increasing rural crisis and unemployment in the cities. Real wages in Bombay mills were more than double that an artisan would earn in a small town, and significantly above the annual average incomes in farming. The cost of moving to the city had decreased, and the knowledge of living conditions increased. Workers in search of casual work would present themselves at the mill gate every morning. The head jobber selected who would be employed for the day, and the machines would be started, well before the department heads entered the mill premises. The jobber was still the main recruiter.Footnote ⁶¹

Chandavarkar explains the jobber's particular value as a recruiter by variation in the demand for casual labour, in turn attributed to fluctuations in market and availability of raw cotton.Footnote ⁶² The interwar period, however, saw a secular shift in the jobber–worker bargaining position. The workers collectively had lost their scarcity value as the casual labour pool grew enormously in the city, whereas the jobber's recruitment role was still formally intact in most mills.

This interpretation of a shift raises one question rather neglected in labour history scholarship. Why was the jobber needed for recruitment any more if search costs had fallen and a labour market was finally in existence? Presumably, the jobber knew the men individually and understood the difference in quality. This was the answer advanced by millowners until the 1920s.Footnote ⁶³ But this is hardly a convincing answer. Bombay mills were clearly not models of work effort. That the jobber was better at sifting quality and was retained on the strength of this particular skill is contradicted also by the fact that many new cotton textile mills in the interior, competing fiercely with Bombay, dispensed with the jobber easily. In these new units, personnel managers recruited workers, both casual and permanent, by means of advertisement, interview, and recommendation, and the system worked well.Footnote ⁶⁴ In Bombay too, managers and unions often made statements against the jobber. The former saw the jobber as a necessary evil, the latter saw the jobber as an exploitative agent.

One plausible theory of the persistence of the jobber despite the erosion of any real need for a labour agent at the work site is path dependence. The jobber was an institution with presence in the neighbourhood, among mill management, unions, and work teams. He, or his female counterpart, was always part of an internal putting-out, responsible for punctuality, quality, training and discipline, and in return was allowed to exercise considerable freedom to hire and fire workers under him or her. The point of this system was not meeting labour shortage. The point was rather the employers' unwillingness to assume responsibility for punctuality, quality, training, and discipline. Having persisted with this system for long, the mills could not disrupt it quickly. For one thing, an alternative system would be far too costly to put in place. And, for another, the jobbers commanded legitimacy and loyalty from workers, cemented by their power of dismissal.

The putting-out aspect of the jobber–worker relationship is illustrated by the fact that in some situations, when the jobber left the job, the team lost theirs too, or followed their captain.Footnote ⁶⁵ Part of this custom was re-created. But there was also involved here a notion of work team that can be seen at work in its clearest form in artisanal textile factories, and that also recur in sardari recruitment in the plantations. Merchants established factories producing carpets or cloth in northern India, but did not recruit workers. They did not even know who the workers were. They hired masters to do the work by piece rate, which the latter did by bringing in their own apprentice teams.Footnote ⁶⁶ The jobber and the sardar partially belonged in the same tradition. They were neither workers nor managers. Rather, they were masters in command over apprentices.

The other side of indirect labour management was ‘an utter lack of supervision in the Bombay mills’. At 1908, when the Indian Factory Labour Commission investigated the question why the workday was very long in the Indian mills, it was apparent that the millowners had hardly any serious thought on the point of organization of work. This was attributed to ‘the fact that the owners themselves had not been brought up to the practical working of a mill’.Footnote ⁶⁷ In 1903, the retired mill manager cited above suggested that this work organization where the jobber was supervisor-cum-boss not only was a long-standing feature but also carried hazards especially when scarcity appeared,. . . ‘the jobber can baffle the foreman or manager who is utterly unacquainted with the class of hands working in Bombay mills.’Footnote ⁶⁸

The system of work characterized by cultural distance, indirect management, and shades of internal putting-out continued without a significant problem until late in the interwar period. From the first major strikes in 1924 and 1925 to the time when the Great Depression struck, the work organization in Bombay came under serious scrutiny for generating too much inefficiency. There were clearly far too many people inside the Bombay mills to do a job that it took a Japanese mill less than half that number to perform. The situation did not demand a remedy when competition was mainly from Lancashire, where wages were more than seven times that of the Bombay wages in 1920. But a threat developed with Japan's entry, wages in Japan being only 15%–20% higher. Why were the mills overmanned? Morris suggests that, given that the jobber received commission on the heads hired, it was in the interest of the jobber to hire more heads rather than less. This is plausible, but we need to qualify. The commission was not entirely a matter of choice for the employers. It was an ingredient of a system that combined the putting-out principle inside the factory with large-scale hiring outside.

Even as internal putting-out continued out of inertia, the specific needs of the employers or the workers that the jobber once met were not so critical any more. In this changed environment, managers, workers and the labour agent all tried to change strategies, and the agent came under increasing pressure, a story too well-known to be repeated.Footnote ⁶⁹

Summary

The labour contractor is not just a phenomenon that is past, but remains pervasive in informal labour markets in South Asia today. Persistence of indirect management on a very large scale invites the historian of labour to suggest origins of this institution, and its significance for industrialization. This study of agency originated in such a problematic. Four points suggested by the material above are potentially useful towards meeting these larger objectives.

First, the labour agent in many standard Indian examples represented a convergence of three distinct intermediary roles – supervisor in the production site (the foreman), recruiter or labour contractor, and headman of communities. Second, this particular kind of agent, who represented incorporation of the putting-out principle in large-scale hiring and employment, had counterparts in the traditional economy. The nineteenth-century origins of the labour agent can be understood better by relating the agent to two contexts: agency in recruitment of labour for overseas migration where the term sardar was already entrenched before it resurfaced in the mills and plantations, and established practices of agency in traditional agriculture, among indigenous tribes, and the crafts, from which background the majority of the migrant workers came. The foreman as headman embodied the hierarchical element that characterized any master-apprentice and patron–client relationships intrinsic to training and supervision in traditional contexts.

Third, this incorporation was shaped by specific imperatives of the modern economy. The pure recruiting agent was an entrepreneur who arose in response to missing markets for labour for large-scale recruitment purposes. From the workers' side, migration and lack of literacy demanded protection of earnings and accounts from predators. A collective achieved this by means of the sardar, who was the point of negotiation and the financial custodian. That being said, the emergence of headman-foreman-recruiter was a flexible process rather than being driven by linear developments within either the traditional society or the modern one. In overseas migration, it happened early. In Assam, it happened late. In Bombay, the mode of recruitment changed. The blend of authorities, thus, was both variable and contingent.

Lastly, this incorporation of traditional authority in modern work sites was not without costs. It gave rise to two types of trade-off. There was a trade-off between quantity and quality. The Assam case showed that the recruiter-agent solved the numbers problem to give rise to problems of stability, industriousness and peace. There was also a trade-off between employers' control on labour organization and their control on numbers. The joining of headman and foreman roles gave the intermediary rent-seeking opportunities. Bombay's tortuous history of mobilization and industrial relations reflected the contradictory presence of the jobber. Rents could be absorbed as production costs in an expanding industrial system with rising wages and profits, but posed a problem when competition intensified. The constant worry over worker quality and efficiency in the 1920s reflected this syndrome. On the other side, as the labour pool expanded, the labour agent began losing rationale to both workers and managers, and yet institutional inertia and depressed markets made changing the rules difficult.