1 Introduction

The episode has inspired important contributions in comparative industrialisation. One strand in the scholarship analyses managerial decisions regarding choice of technology [Otsuka, Rains and Saxonhouse 1988; Kiyokawa 1983]. However, this line of reasoning is more applicable to spinning than weaving, whereas capital-labour ratios varied between India and Japan in both spinning and weaving. Another argument, advanced by Gregory Clark in particular, emphasises the level of effort supplied by the average worker, but leaves the effort level to be explained by unspecified cultural factors [Clark 1987; Clark and Feenstra 2001]. Taking the efficiency argument further, Susan Wolcott (1994) attributes the divergent effort level to union resistance to rationalisation, that is, to attempts to make productivity improvements, in India. However, even before the so-called standardisation and rationalisation efforts started in Mumbai, indices of effort level were already relatively low in the latter. The rationalisation move can be seen as a response to the problem, and not the cause. Bishnupriya Gupta (2003) argues that the managers failed to see the existence of a virtuous circle between wage and efficiency, due to a peculiar division of managerial responsibility on the shopfloor. The efficiency wage thesis has antecedents, as we shall see. Wage hikes, however, were rendered difficult by the competitive situation itself.

This literature more or less overlooks a contemporary debate on efficiency in the Indian mills. Contributors to this debate

Economic & Political Weekly January 5, 2008

discussed management, discipline, culture, and wage incentive, themes that reappear in economic history discourse 70 years later. One section among the industry insiders who joined the debate emphasised labour efficiency. In their view, well before the competition from Japan began, labour institutions that had developed in the mills had a built-in bias for waste of effort. The Japanese competition merely exposed the problem. This institutional explanation for the falling-behind has not been explored enough in the economic history scholarship. I develop the hypothesis more fully in the present essay.

Intermediaries

The argument of the essay is a simple one. Most large-scale enterprises in 19th century India did not recruit workers directly, but recruited through labour contractors. Further, almost from their mid-19th century beginning, or possibly from a little later, large-scale employers such as factories, plantations and ports left a great deal of supervision and training to these intermediaries. As against these useful services, the intermediary received a most important privilege, the freedom to hire whoever he/she wanted to hire, and the freedom to fire workers under him/ her with only the minimum formal consent of the management, or often not even that. In other words, an internal putting-out of human effort characterised the worksite. The managers gave the intermediary a job to do, and the latter got that job done with his or her own work-team. I argue that the power of the intermediary, or more generally the practice of workers having the authority to fire other workers, gave rise to moral hazard problems. In effect, it led mills in Mumbai to hire more hands than was necessary and hire poorly trained workers even when better quality of people were available.

Japanese competition exposed this syndrome. Until the 1920s, Indian wages were low enough compared with Lancashire or New England for the mill-owners to overlook the costs of internal putting out. The system surely simplified the managerial role of the employers. Mill-owners could spend their excess energy dealing in shares and commodities, worlds that they understood much better and loved much more than they did the world of labour. But wages in Japan and China were not much higher than those in India. When competition from Japan began to bite, employers became more introspective. All along, sections among mill managers had a better sense of the costs of the internal putting out system. When the crunch came, both found the convention hard to break.

The rest of the paper has four sections. Section 2 places the theme of the paper against a more general discussion on the labour contractor. Section 3 describes the background to the inter war debate and the contemporary viewpoints. Section 4 deals with the institutional argument on efficiency. Section 5, the last section, concludes.

2 The Contractor in Theory and History

The argument that the institution of the contractor could become costly for the employer, especially because it could potentially compromise skills, training, quality, and incentives, faces objections from both economic theory and economic history, and therefore, needs to be qualified and contextualised.

38

Economic theory and economic history recognise that hiring a labour contractor can be a rational employment decision in certain contexts. For example, when a labour market is undeveloped, or the employers do not know the prospective workers well enough, or the market is too risky or too seasonal for continuous direct employment, or labour regulations make direct employment contracts too costly for the employer, engaging a labour contractor may be rational [for a survey of these issues see Abraham and Taylor 1996].

Consistent with this insight, early or late industrialisation relied heavily on an agent for recruitment and for retaining rural labour in urban and unfamiliar workplaces. The agent was ubiquitous in the early 19th century cotton textile mills in England and north America, and in Japan and India in the late 19th century.2 Historically, the drives behind such decisions included a lack of information on labour markets, and sometimes communication gap between the workers and employers. Seasonality and risks were at work too. Regulatory law was rarely an important factor in the early history of contracting. On the other hand, there were two additional drivers behind the importance of the contractor in the early history of the mechanised factory, which have not drawn sufficient attention from labour economists. One of these was the scarcity of managerial talents and specialised managers, which led employers to share some managerial functions with skilled foremen. And the other was the scarcity of technical schools, and in turn the need to rely on the skilled foreman as a master or provider of training. This is the context in which the phrase “inside contractor” was often used in the early 19th century New England mills.

Quality Concerns

Whereas this paper emphasises skill development, several historians of cotton mill labour in 19th century India and Japan have suggested that the work that the average millhand in these countries did was not particularly skilled, and could be easily learned. The prevalence of young women from farming background in the mills in Japan is sometimes cited in illustration of this point about skill. The prevalence of generic or low skills could strengthen intermediary authority in particular ways. Consider the follo wing argument, for example. A classic work on Mumbai textile mill labour proposes that, contrary to presumptions made in some older accounts of the Indian worker, Mumbai mill labourers represented a workforce that carried a great deal of commitment to the worksite.3

In turn, this story is consistent with a notable feature of early recruitment, that it prioritised “reliability” as opposed to skill or formal training. “The term ‘reliability’”, one contributor to the literature wrote, “might be said to include a wide variety of attributes of importance either to an employer or the jobber” [Kannappan 1985]. G Balachandran points out that in a situation when the intermediary had to take both wages and skill levels as given, and skill levels were generally low, wages and productivities could systematically differ, the difference being appropriated by the intermediary.4 This is, of course, a different interpretation of “rent-seeking” by the intermediary from the one that the critics of the Mumbai jobber argued (more on this

January 5, 2008 Economic & Political Weekly

critique below). I find this entirely plausible and broadly consistent with the story that early, even a great deal of late recruitment was driven by reliability rather than quality concerns, so that quality tended to vary somewhat randomly.

All that being said, agency had costs too, and the particular view that I explore in this paper suggests that these costs increased, in relation to Japan, in interwar Mumbai. How do we explain the costs in the context of the global history of the labour contractor that I just outlined? While the agent was used almost universally in the early stages of industrialisation, and the textile mill work was semi-skilled, both these con

more lasting factor, the threat of Japanese and Chinese competition in certain grades of cloth in which Mumbai found it impossible to match east Asian prices. A significant burst of import-substitution during the war had by then ended. And competition was beginning to increase, both inside India and in those overseas markets (especially Africa and west Asia) where Indian textiles had enjoyed a good run in the preceding 20 years. In 1913, Indian mills produced 1,163 million yards of cotton cloth and 683 million lbs of cotton yarn. The figures for imports were, 3,200 million yards and 44 million lbs respectively. And 10 years later, Indian

1907 472 353 808 638 135 983 ver, employers in Europe, America, Japan, 1917 933 357 1614 661 595 1924 after, production stagnated, in Bombay island

and indeed India, realised that training 1925 871 262 1954 686 1031 2073 the stagnation was particularly acute (Table 1).

3 Contemporary Perspectives

In the first half of the 1920s, the cotton textile mills in Mumbai experienced an unprecedented fall in profitability, after enjoying a decade-long boom. Average rate of dividend paid out by Mumbai mills between 1917 and 1921 was 65 per cent, in 1922, 58 per cent, and in the three years 1923-25, an unheard of 19 per cent.6 In Ahmedabad, average dividend in 1917-22 was 49 per cent, and in 1923-25, 35 per cent. These rates were directly proportional to rates of profit. In Sholapur, Indore, Madras, Central Provinces, Bangalore and Bengal likewise, profits and distributed profits fell, if by smaller magnitude than in Mumbai. Almost all large mills in Mumbai skipped paying anything at all to the shareholders in 1925 and 1926. Cotton mill shares in Mumbai had increased in value fourfold between 1914 and 1922. In the next four years, the average value halved, a much sharper fall than that in industrial securities in general.

What worried the mill-owners about the crisis was that it derived only partly from world business conditions; and largely from a

Economic & Political Weekly January 5, 2008

What is more, the direction of imports changed, in a manner that left both Mumbai and Lancashire very disturbed. In these 10 years, the share of Japan in cotton yarn import had increased from 2 to 46 per cent (in weight). And the share of Japan in cotton cloth import increased from 0.3 to 8.3 per cent.

Unfair Competition

In 1926, the Bombay Mill-owners Association made a representation before the Indian Tariff Board stating that competition from Japanese yarn and cloth was the main reason for the crisis. The mill-owners further argued that the competition from Japan was an unfair one, deriving from manipulation of exchange and deliberately repressed labour standards. On this ground, tariff protection was demanded, and was eventually granted. The case made strategic sense, since Japan was directly responsible for a fall in the share of Lancashire in the Indian market. Lancashire was less opposed to tariffs than ever before. There was also an implicit agreement between the workers and the capitalists in Mumbai on the nature of the threat from Japan. Both believed that the competition from Japan was an unfair one; only the trade unions went beyond protection and demanded wage hikes as well.

The mill-owners’ representation made three specific points. First, the depreciation of the yen since 1921 was the main reason behind a fall in the cost of Japanese textiles in India. Exchange was overvalued in India and undervalued in Japan. At any rate, the rupee-yen ratio fell by almost half between the monsoon months of 1921 and winter of 1923. The effect of the fall on business was partially neutralised, however. In the same period, Indian prices fell by 15 per cent, and Japanese prices remained the same, so that the real appreciation of the rupee was somewhat smaller. Further, Japan produced yarn and cloth mainly using Indian cotton, which would have neutralised the exchange rate anomaly somewhat. The yen, in any case, had begun to fall against the rupee from early-1924, just when the crisis became acute for Mumbai.

Comparative Costs

Second, there was a difference in comparative costs, which owed to the higher effort level of the Japanese workers. By the mid-1920s, the average Japanese weaver earned a wage about 10-15 per cent higher than that of the average Mumbai weaver. And yet, whereas direct wage costs were about 25 per cent of the cost of production of 14s yarn in Japan, in Mumbai the percentage was more than double that. The real difference was in conventions of work, and it was a persistent one. In 1931, one weaver handled six looms in Japan, an average Mumbai weaver handled two looms (Table 2 for more details on performance indicators). In 1940, 4,50,000 workers in cotton textile mills in India processed three to five million bales of cotton. In Japan, 1,90,000 workers processed the same quantity of cotton in that year.7 The mill-owners implied that this difference owed to exploitation, in particular, double shift and night employment of female and child [sic] workers in the mills in Japan, in partial violation of the Washington International La-

Table 2: Comparative Capital-Labour Ratios, 1931bour Conference, 1919.

Japan China Bombay

India had ratified this

Island

convention, Japan did

Looms/weaver 6 4 2 Wage per day (Rs) 2.25 1 2 not, and that fact fig-

Labour cost per loom (India = 100) 25 Spindles/Ringsider 600 380 181

The Labour Regime

Individuals with some knowledge of how production took place inside Mumbai cotton mills usually rejected the view that repression of labour was the fundamental cause of the discrepancy in performance. There was a current of opinion, among mill managers, economists, governments of major provinces that did not contain spinning-weaving mills, and of course, Japanese Chambers of Commerce and trade delegations themselves, that the real problem for Mumbai was inefficient labour organisation. What puzzled these observers was that, whereas the average working life of a worker in Japan was three to five years, in Mumbai it was 20 years. Why would a far more mature workforce be so far behind in basic performance indicators? Would a repressive labour regime explain all the difference? Mill-owners did not fail to acknowledge “the shrewdness and capability” of the ordinary worker, “however illiterate he may be”.10 What, then, accounted for the difference in basic indicators of efficiency?

On this question, the Japanese trade representation, the trade unions, local governments, and the engineers represented a range of viewpoints that differed from that of the mill-owners.

In his response submitted to the Tariff Board, the Bombay Agent of the Japan Cotton Spinners’ Association made a spirited attack on the Bombay Mill-owners Association view that labour standards and government protection made all the difference between Mumbai and Osaka.11 He questioned the precise calculations of exchange and average costs and rejected the numbers presented by the Association, pointed out that Japan was also an importer from India, and spent more than two-third of his first report questioning the relevance of the labour standard issue. Notwithstanding the Washington convention, was there a universally accepted international law against night-work and double-shift? “As a matter of fact, Bombay mills tried [double shift] and failed” because the workers did not want it. The mill-owners’ “objection to female employment at night does not appear to be based as much on humanitarian as on business considerations”. Female employment in Japan, indeed, was a necessity because of “the pressing population problem”, much like conditions in prewar Germany, “with this difference that whereas the doors of all countries were open to the German emigrant, they are .. shut to the Japanese. … If there is not sufficient work to go around during the day, then they must toil at night”.

Mill-owners themselves did not all speak in one voice. A significant minority among them argued that there was a difference in business organisation, which translated into better utilisation of economies of scale both in production and in marketing in Japan. ‘In Japan we found that the biggest reason for [competitive strength] was absolute vertical specialisation...three firms purchase cotton upcountry in India and in America… These organisations have huge mills which make only one type of cloth, one warp, one weft and one mixing”.12 Such integration economies were attempted also in the case of the Toyo Podar mill, formerly Diamond Mill, the only factory with Japanese participation in management.

Some of the provincial governments made a strong case against imposition of tariffs on Japanese textiles and English yarn in India. They saw themselves to be representing the interests of the consumer and the handloom weaver. Some of them introduced a new element, quality, in the discussion. “Japanese goods are prefer red mainly for better finish and lower price. Japanese mills have improved their products considerably in every direction, whereas Indian mills have not made any appreciable advance. Our Indian mill-owners with few honourable exceptions do not take as much personal interest in improving the quality of fabrics as they take in increased production and profits” [Government of Assam 1927].

Worker Unions Take Shape

In 1926 the first Trade Union Act took shape. From some time before this date, Mumbai was the site of two major trade unions that had already coordinated a few large-scale industrial actions, and whose presence was responsible for significant wage-hikes in the early 1920s. The trade unions focused on one element in the millowners’ Japan thesis, repression of labour. When invited by the Spinners’ Association in Japan to see for themselves what conditions of work were like in Japan, the mill-owners had decli ned, displaying “a little unwillingness to go deep into the matter”, alleged one union leader.13 The trade unions, on the other hand, claimed

January 5, 2008 Economic & Political Weekly

to possess a more intimate knowledge about working conditions in Japan, informed partly by reports received from their counterparts in Japan.

From the beginning, the Bombay Textile Labour Union maintained that “Japan is getting an advantage on account of the sweating of her workers”.14 The union did not deny that Japanese workers handled more machines per worker, but attributed it to repressive labour regimes. In the next four years, the sweatshop thesis acquired more colour. During the Strike Enquiry Committee proceedings at the end of the decade, the labour leader S A Dange stated, in conversation with the mill-owner Sir Victor Sassoon, “The point is to what reasons is the Japanese efficiency due. The question is whether it is the voluntary efficiency of the worker, or whether he [sic] is made to do it by an outside terrorist agency”.15 Dange went on to “point out” that the Japanese worker was not in fact taking this treatment lying down, but “he [sic] was revolting”. The other side denied that the Japanese worker was revolting, and said they “had no news about it”.16 A heated conversation between N M Joshi and T Sasakura, managing director of Toyo Podar, highlighted “the [poor] contentment of the workers, .. [repression of] communists, and the police laws”. Sasakura stood firm: “You have spoken too much about exploitation of labour. I have heard this complaint from you, and not from the Japanese women workers”.17

Sasakura agreed that the only reason “India has got behind in the matter of going ahead” was poor effort level, but disagreed on the source of this divergence. The real source was voluntary and institutionalised waste of effort inside the Indian mill. His mill had introduced practices that could reduce wastage of material and labour. A part of this attempt focused on “loitering”. What was loitering? A Managing Agent defined loitering as “sitting down, smoking and doing nothing”.18 Most Indian mills in the 1920s did not have designated periods of rest. Workers took time off when they wanted to, and the supervisors usually allowed them to go out. This practice, together with the somewhat larger-than-average numbers present in every department, had made a certain degree of idling a routine matter. In the spinning department of the mills, many men did this, so consistently indeed that surprise checks by managers invariably surprised the managers. “In Japan loitering of the kind allowed here is never allowed”. The Toyo Podar tried introducing designated breaks, but had to retreat. The mill was already caught up in disturbed industrial relations, and neither made money nor succeeded remarkably in raising effort levels.

In significant departure from the repression thesis, the trade unions also believed that Japan was not the only reason behind Mumbai’s problems. They believed that the crisis was the millowners’ own making to a large extent, and that the profits of the industry were under pressure because of faults of the managing agency system, and therefore, the solution was nationalisation.19

The Wage Factor

For the trade unions, there was a yet third problem, again created by the management, which was the low wage in Mumbai. During the Strike Enquiry Committee, it was stated many times that “workers would work better if they were contented”. More explicitly, “[w]e feel that unless the worker has sufficient stamina to

Economic & Political Weekly January 5, 2008

do his work in the factories… there is no possibility of the workers’ efficiency being increased”.20 But were wages too low to act as performance incentive? Comparison with “upcountry” Indian mill and Chinese mill wages with those in Mumbai consistently showed that wages were not particularly low in Mumbai. In 1925, the average wage of a mill worker in Mumbai was in the range Rs 25-30 per month or Rs 300-360 annually. An agricultural labourer earned an income that did not exceed Rs 100 anywhere in India. A peasant in the densely populated Gangetic plains would have earned an income, given the small size of the holdings most of them had access to, only 20-25 per cent higher than that of a labourer. When such a person came to Mumbai and gained entry into a job that would have been lucrative by any contemporary standard, the efficiency effect ought to be very strong. The allegation that the Mumbai worker reeled under low wage and poor incentive convinced few outside the union circles.

Could the Indians learn to become equally efficient? “Why not” – the mill-owners responded, but they did not offer a specific plan. The trade unions believed “nationalisation” would solve the problem. The mill-owners felt amused at the idea:

Do you mean handing over all our mills to government? I only hope they make a better job of it than they did of the Backbay scheme But you do not seem to have made a very good job of it either? If we had a national government, our position would be very different.21

For, adequate protection would then be available. The mills’ so-called solution was a tariff wall for which a national government, but not nationalisation, was necessary.

The Efficiency Issue

Within the particular context of these debates, it would seem that neither the unions nor the capitalists or the governments touched the core issue of labour efficiency head on, beyond the rhetoric discussed above. There were authoritative voices within the trade union world calling for a wide-ranging discussion on the efficiency issue.22 Yet, the fact remains that a united and convincing case on efficiency was not made from the trade union circles in any of the contemporary forums. Their representation to the Tariff Board, running into hundreds of pages, spent three pages on labour, and only a few words on efficiency. When pressed on the point of how to make Indian workers work harder, the trade unions usually raised the efficiency-wage point.

Mill-owners did not propose a clear viewpoint either. During a discussion on the point, N N Wadia, owner of one of the leading mills, offered his wisdom on why the Indian operative could not manage more looms than at present:

It is a remarkable thing that I have observed that a boy of 14 is smart, intelligent and up to the mark; but the moment he gets married, say in a year, he collapses and he becomes absolutely inefficient in another 12 months.23

Such inspired reflection did not impress everyone. Among those others were members of the technical supervisory class, or foremen, who were more knowledgeable about the shop-floor than the average mill-owner. Some of them offered a different, if narrower, understanding of the efficiency question. These people included technical men, a few economists, and a small group of capitalists who were trying to take closer control of the shop-floor.

4 An Institutional View

These individuals noticed two important differences between India and Japan. First, whereas the factory was a site of imparting instruction and welfare in the latter, Mumbai mills were notoriously indifferent to any form of instruction and welfare. Second, the Indian mill tended to hoard labour, a practice they found extremely difficult to get rid of even as competition increased.

A leading textile merchant commented in response to a question posed by a Tariff Board member:

The Indian mill-owner has not cared to educate the Indian labourer even in spheres of activities closely connected with the textile industry. He has always worked on the principle that any intellectual betterment .. of the labouring class in matters industrious would be antagonistic to his own interests as mill-owner. The result is that the Indian labourer has hardly any incentive to give of his best to his master and hardly any interest in the well-being of the industry.24

The problem was often posed in cultural terms or in terms of the quality of “patriotism”. “In Japan … the owners of the mills as well as the poor labourers are kith and kin. They talk to them and go with them, take them to excursions and please them in every possible way. Here they do not. The burra sahibs never care for the labourers and they never sympathise with what they lose or gain”.25 There is much truth in this picture, except that the problem of class distance was not wholly a cultural one.

The Jobber

The person in charge of communication with the workers, and who was by default in charge of training, was not the owner, nor the manager, but the “jobber”. The jobber was a senior worker who began as a labour contractor, evolved into a supervisor or head of work-teams, and throughout, was given over significant powers to hire casual labourers and fire them as well, on a token permission from the immediate superior. Unlike the manager, the jobber was “one of the workers”, a counterpart of the headman of worker gangs or a master artisan. They belonged in a segment of the hierarchy that was distinct from that segment represented by the Parsi or European foremen and technical supervisors. But the latter could neither do without nor replace the former. In one view, the jobber became necessary because of a class and cultural distance between the technical men and the raw labourers [Morris 1960]. However, the Mumbai jobber was only one instance of a whole variety of labour contractors who supplied workers to the modern large-scale enterprises since the mid-19th century, suggesting that search costs and supervision costs were also important factors behind the universal need for the agent.26 The much discussed and sometimes misunderstood jobber clearly had played a very positive role in looking after the interests of the ordinary workers. Pioneering figures in labour mobilisation and organisation emerged from the jobber background. Many workers had good reasons to look up to them for welfare, fair treatment from the employers, and moral support. Jobbers were also far from a homogenous class. All that said, in the 1920s, the jobber had acquired a largely negative connotation. Sections of capital felt that the jobber undermined quality of effort, and sections of labour felt that the jobber exploited the privilege to hire and fire workers. Where did these complaints come from?

42

Whatever their origins, in the 1920s the jobbers conducted much more than recruitment. They were effectively workers entrusted with the duties of a personnel manager, they were in charge of full utilisation of the machines, and in the absence of a well-developed apprenticeship, in charge of supplying and shaping skill levels as well. The managers and owners could wash their hands off training and education precisely because the jobber was by default, but rarely in practice, in charge of these functions. “People speak of the inefficiency of labour” in Mumbai mills, a Tariff Board member pointed out to G W Burley, head of the largest technical training institute in India at that time. Is it possible for the mill managers “to organise some kind of instruction for them”? No, Burley answered. “The matter is entirely in the hands of the jobber. He is a jobber because he can bring labour into the mill… They bring in people who know nothing at all about textile work, and instruction appears to be largely in the hands of the jobber”.27 Note the phrase “appears to be”, suggesting that this function was not deliberately contracted out, but was an unintended consequence of decentralising the employers’ responsibilities. Many jobbers themselves were poorly trained, but even when they were in full command of the machinery, they preferred not to teach too much, for reasons we shall see. Typically, if a loom broke down, production stopped. The weaver went to the jobber, “who will only attend to him when the jobber is free”.28

Horizontal Divide

The Royal Commission on Labour in India hinted at a horizontal divide within the working class. The workmen and the supervisors were distinguished by possession or otherwise of literacy.29 In England, the distinction did not exist, permitting workmen to graduate into supervisors. In India, literacy and unbridgeable class difference acted as a barrier to this progression. There were two apparent consequences of this problem. First, the supervisory positions, occupied by Europeans initially, were later “Indianised” with people who did not rise from the ranks, but were products of universities and polytechnics. They did not speak the language of the workers and did not belong to the same social class. Second, an intermediate layer of supervisors rose from the ranks to monitor workers. Training-on-the-job was a responsibility of the latter. But in this case, training became a part of the commercial and hierarchical transactions that tied these “jobbers” with the workers. Falling between these two stools – a largely distant and alien technical cadre and a senior worker who was more of a headman-cum-boss than a skilled technical master – formal apprenticeship never had a chance to develop fully in the Indian textile industry. At the same time, the distance between the managers and the workers deepened, rather than being bridged, by the jobber.

In the end, it turned out that not much systematic training in the more skilled tasks inside the mill happened at all. “This is the only country where the mill-owner deliberately ignores all industrial training”.30 The following exchange at the Royal Commission illustrates especially well the extreme hands-off policy with respect to training:

Q Would you tell us about the organisation of the mill with which you are concerned; have you an assistant manager?

January 5, 2008 Economic & Political Weekly

A I have no assistant manager; I have one spinning master, one weaving master and one engineer. Q Who comes under the weaving master? A The weaving head jobbers and the line jobbers. Q Has there been any attempt in your mill to train any of the weavers in the weaving shed with regard to the handling of the machinery? A You get a boy; when he comes to the mill he works with an old weaver and learns which way to do it. Q Is that the way they learn the trade? A That is the custom of the country. Q Is that the way to get efficiency? A That is the custom of the country.31

But why wouldn’t the jobber organise better quality training of the raw workers? Part of the reason was that not all jobbers were necessarily expert workers in the first place. Indeed, few were. But given their authority over the workers, the jobbers could still potentially arrange on-the-job training if they wanted to. That they did not can be partly explained by the fact that there was a contradiction between the training and the recruitment functions of the jobber.

The Casual Labour

The jobber, or equivalent senior workers notionally in charge of training, was also given the power to hire and fire workers under them. In the depressed trade situation of the 1920s, and a large pool of migrant workers available at the mill gate at the same time, the jobbers could earn a comfortable commission from those they hired. The commission was higher for less skilled workers, because the latter would find it harder to get a job. The return from creating a really skilled worker would accrue to the mill. The return from recruiting raw hands accrued to the jobber. There was incentive, therefore, to under-train or not to hire the best quality. That was not all. Mills in Mumbai always kept many more people on the rolls than were needed. The mill-owners’ standard complaint was that absenteeism reached high levels. But they did not investigate why there were high levels of absenteeism. Several contemporaries and historians observe correctly that the jobber had an incentive, namely, the commission, to hire larger numbers than were needed.32 The managers tolerated this practice as long as the jobber took care of delivering a job in time, and as long as wages were low enough.

Increasingly in the interwar period, these hands were selected as ‘badlis’ from crowds gathered at the mill gate. The jobber, who usually kept a subsidiary muster of his own, did the selection. Any senior or skilled worker could also hire substitutes, and enjoyed some of the powers that the jobber did. Once the managers agreed to take in a large number of badlis to meet a shortfall, the jobbers said: “They will create trouble...go on strike...the best men will leave the mill unless a certain proportion of the men are allowed to work, at least for half the month or something like that”.33 However, who would stay and who had to leave was a decision that depended on the price of job security in the internal labour market, which was controlled by the intermediaries.

Mill-owners, when criticised on efficiency, put the blame on absenteeism and high rates of labour turnover. They did agree that the casual labour component was a hurdle to creating good workers. But they believed that casual labour was a necessary evil because of absenteeism. “Unfortunately”, they wrote in their

Economic & Political Weekly January 5, 2008

representation “(the badli system) has to be adopted to fill up the gaps caused by wholesale absenteeism” [Bombay Mill-owners Association 1927]. This link between absenteeism and casual labour hiring, of course, left absenteeism the unexplained independent variable, implicitly attributed to cultural characteristics of the work-force. The subsequent interviews of the mill-owners conducted by the Tariff Board members revealed this attitude more clearly. As a matter of fact, absenteeism was high because of the privilege that the jobbers enjoyed in respect of hiring of casual workers. It was a reflection of the jobber power to hire and fire. Badli was not the effect of absenteeism, but its cause, and the intermediary was the cause of badli.

Propensity to Remain Absent

Absenteeism and surplus labour followed from a logic connected with the workers’ power to hire other workers. When a senior worker wished to be absent, and sent in a substitute, it was the former who received the wages of the substitute [Director of Information and Labour 1927]. Depending on the status of the substitute in the workforce, a commission was deducted before the money was paid to the actual worker on that day. Conversely, “[i]f the amount of pay that he [a senior worker] earns is sufficient he stops working...and sends in a badli”.34 And again, “I have discovered this characteristic of the Indian labourer: if he can meet his monthly obligations by working 20 days, he will certainly not work 21”.35

This hint of a backward-bending supply curve was made in other contexts of large-scale employment in the 19th century, and needs a convincing explanation. The one offered by the same speaker in the next line, “he is not interested in earning more than Rs 45 a month”, does not quite amount to a convincing explanation. I believe that the propensity to remain absent can be explained by the practice of workers having the power to hire workers. Consider the case where a worker attains seniority, which would mean attaining the power to hire casual labourers, and get the expenses reimbursed by the manager. From each casual labourer the senior worker extracts a bribe. In other words, by working, a senior worker gets his own wage, and by not working, the senior worker gets a portion of the substitute’s wage. This situation would simply make leisure more attractive than in the situation in which a manager recruited the substitutes directly. Preferences remaining the same, the worker is more likely to take time off than before, and tend to install a substitute. This state of affairs would benefit everyone, especially if the rate of pay for casual workers was lower, which was usually the case. The senior worker could in principle maintain income and yet enjoy longer leisure, the employer could get additional hours at smaller rate, and the casual worker would get a job.

Not only did the jobber find it profitable to hire more heads than the mill needed, the jobber tried to fire senior workers more often than would be the case with centralised hiring. For, the jobber lived on commissions, and any worker who believed he/she had the prospect of getting a job elsewhere refused to pay the commission demanded.36 In short, if a senior worker was given the power to fire other workers, the first people to go would be the potential competitors, namely, the good quality people. A great deal of these abuses was publicised by the textile labour union. However, the union strongly resisted any suggestion that the badli system harmed efficiency of labour [Bombay Textile Labour Union 1927:467]. The unions, indeed, were caught up in their own dilemma. The divisions within the workforce – with its jobbers, skilled workers, the rank and file, apprentices, badlis, migrants and settlers, castes and communities – made their ultimate agenda, to forge and deepen class consciousness, a huge challenge. Paradoxically, they could not succeed in this challenge without using the persuasive power of the worker-intermediaries themselves. In this situation, they could hardly afford to open another front by raising the training and quality implications of the system.

In the end, the result was a mill shop-floor with more workers than needed and a number of poorly trained hands. Such a factory site would also display a relatively greater degree of resistance to technological change, since any such changes would strengthen the skilled workers, harm the interests of the less-skilled ones as well as harm the intermediaries. This was the predicament of interwar Mumbai.

5 Conclusions

This brief sketch of emerging competition between India and Japan and a debate around the episode that discussed labour institutions has three simple lessons. The three lessons touch on, respectively, why labour institutions became a problem in the 1920s Mumbai, what was distinctive about Indian labour in the course of industrialisation that made agency so tenacious, and what message, if any, the episode carries for current debates on labour.

The first point addresses the historical question, why Mumbai retreated against Osaka or Shanghai in the 1920s. I restate here a contemporary view suggesting that modes of labour organisation that allowed senior workers the freedom to hire and fire other workers raised labour turnover, absenteeism, labour-output ratios, and reduced quality of training and apprenticeship. My interpretation of the syndrome is as follows. The labour-intermediary involved particular forms of moral hazard and selection biases. The agent raised labour turnover and labour-output ratios because the agent’s commission depended on the number of recruits, and not their quality. The agent also raised absenteeism because the attraction of commission earned from supplying others increased the value of leisure. And agency affected training because betterquality individuals were a potential threat to the power of the incumbent worker-boss. The 1920s Mumbai faced a moment of transition from a regime that had once valued numbers to a regime that needed to improve quality and productivity. While specific training had become important for Mumbai mills to meet competition, the jobbers could hardly supply skill leadership. Unlike some of their counterparts elsewhere in the world, the Mumbai jobbers were not always skill-leaders although often implicitly in charge of quality of workers. They had reason, furthermore, to resist attempts to improve skills and technology. And so did many workers under their command. It was this particular context that made the jobber a burden for capital rather than a help.

Competition from Japan exposed the problem, and the series of labour enquiries in the 1920s supplied the forums where the problem could be discussed more openly than before. In the end, protective tariffs provided respite, but only a momentary one, for in the post-depression years, Mumbai had to contend with another threat that could not be met with tariffs, competition from lowwage mills within India. Efficiency now demanded implementation of long-postponed institutional reforms. To many managers and owners, reforms meant getting rid of the jobber. Attempts to do so made worse a fresh burst of industrial disputes, a story too well-known to be repeated here [See Chandavarkar 1994].

Historical Distinctiveness

The second lesson of the story concerns the historical distinctiveness of Indian industrialisation. Why was the labour agent so tenacious in Mumbai, and indeed many other contexts of mill employment in India, even as migration increased and labour markets matured? Of course the particular crises conditions in Mumbai made the agent more useful for some sections of the business, in breaking strikes for example. But at least four other possible explanations for the tenacity of the institution can be offered. First, the contractor in India had, by a historical coincidence, been given more managerial authority than elsewhere, and this feature imparted a path-dependency effect. Second, perhaps the “headman of a community” commanded more social capital in India than did his counterpart amongst other late-industrialising working classes. Some discussions in labour history on caste and loyalties would seem to hint at such a possibility.37 A third speculation yet would be that the mill-owners in India preferred a more decentralised managerial model than their counterparts elsewhere, whether we characterise that as a “managerial failure” or not is a different question.38 The least speculative and the most straightforward would be the fourth explanation. There was an embracing macroeconomic factor that distinguished India from Japan, Britain, or New England, where the agent had been successfully marginalised. This was the weakness of institutionalised sources of supply of technicians, foremen, and professional personnel managers. More than any other factor it was perhaps a slow development in formal technical and managerial education that made the jobber continue for a long time before finally weakening.

The third and final lesson is potentially relevant for interpretations of comparative industrialisation in the present times, the theme this essay started from. Since the 1990s, India’s prospects in the world market for manufactures relative to east Asia have been discussed again with an intensity that is reminiscent of the 1920s. One difference in the terms of the debate seems striking though. In the current discourse, there is seemingly less emphasis upon, or understanding of, what role, if any, labour institutions play in deciding the outcome of international competition in labour-intensive manufactures. Much of the discussion on post-reform industrialisation focuses instead on government policy and labour laws. This is a valid emphasis in my view, but we may still be missing another dimension, the difference that quality, training, skills, and incentives make to an industrialisation drive. Incentive systems, training systems, and good communication, often matter more than wages in deciding the outcome of competition. The average worksite in Indian small-scale industry leaves much room for improvement in these respects. Having seen how important these factors were in India’s retreat against Japan in the interwar period, it seems timely to ask, how important are these factors in the competition between India and her nearest rivals today.

January 5, 2008 Economic & Political Weekly

1 I use the present name of ‘Mumbai’ for the city known in the period of this study as ‘Bombay’, except in citations and references.

2 On “inside contractors” in pre-1850 New England mills, Andrew Dawson, ‘A New Framework for Workshop Contracting: Philadelphia Machine Building, 1870-1914’, Labour History, 47(3), 2006, pp 343-59. I explore the significance of the worker-intermediary to the study of industrialisation in India more fully in ‘Sardars, Jobbers, Kanganies: The Labour Contractor in Indian Economic History’, Modern Asian Studies, forthcoming. See this article for a fuller set of citations.

3 Morris D Morris, Emergence of an Industrial Labour Force in India, University of California Press, Berkeley and Los Angeles, 1965. For an example of the thesis that the Indian worker was partially committed, see Charles A Myers, Labour Problems in the Industrialisation of India: Harvard University Press, Cambridge, Massachusetts, 1958.

4 Comment made in the ‘Open Reviews’ section of the Economic & Political Weekly web site.

5 On the effort by the employers to improve incentive systems in the larger spinning mills in Japan from about the 1910s, and the simultaneous weakening of middlemen authority, see Koji Taira, ‘Factory Labour and the Industrial Revolution in Japan’ in K Yamamura (ed), The Economic Emergence of Modern Japan, The University Press, Cambridge, 1997, pp 239-93. On the effort by the employers to supply moral and educational training, see Kaoru Sugihara, ‘Transformation of Social Values of Young Country Girls’ in Janet Hunter (ed), Aspects of the Relationship between Agriculture and Industrialisation in Japan

STICERD, London 1986. By the 1920s, the workforce in Japanese textile mills was both substantially more skilled and more assertive than in the 1890s, and than the average Indian observer of the Japanese system realised.

6 Based on data published by The Investor’s India Year Book.

7 For a discussion on the comparative data, see S K Batliwala, ‘Problems of the Indian Mill Industry’, The Indian Textile Journal: 1890-1940 Jubilee Souvenir, Bombay, 1940, pp 26-29.

8 Bombay Millowners’ Association, Report of the Indian Tariff Board (Cotton Textile Industry Enquiry) 1927, Vol III (evidence of the Bombay, Ahmedabad and Baroda Millowners’ Associations), Bombay, 1927, p 31.

9 For example, one “up-country” mill-owner said, ‘[i]t is clear that the Japanese Government… render every monetary help and give protection, on account of which the mills in India have had to suffer a severe blow”, note by Vithal Das Rathi, mill-owner, Beawar, Report of the Indian Tariff Board (Cotton Textile Industry Enquiry) 1927, Vol III (evidence of local governments, etc), p 27.

10 Evidence by R Blackwell, J Parker, J B Green, Bombay European Textile Association, before the Royal Commission on Labour in India, I(2), Bombay Presidency, Oral Evidence, 1931, p 333.

11 Report of the Indian Tariff Board (Cotton Textile Industry Enquiry) 1927, Vol III (evidence of local governments, etc), pp 177-200. See also oral evidence of T Sasakura, Manging Director, Toyo Podar Cotton Mills, Proceedings of the Bombay Strike Enquiry Committee, Bombay, 1928, Vol I, pp 274-98.

12 Stones, Proceedings of the Bombay Strike Enquiry Committee, Vol II, p 850.

13 N M Joshi, Proceedings of the Bombay Strike Enquiry Committee, Vol I, p 44.

14 Evidence by the union, Report of the Indian Tariff Board (Cotton Textile Industry Enquiry) 1927, Vol III (evidence of local governments, etc), p 514.

15 Proceedings of the Bombay Strike Enquiry Committee, Vol II, p 877.

tee, Vol III, p 1242.

Bombay Strike Enquiry Committee, Vol III, p 1071.

34 Evidence by R Blackwell, J Parker, J B Green, Bombay European Textile Association, before the Royal Commission on Labour in India, I(2), Bombay Presidency, oral evidence, 1931, pp 333-34.

35 J Addyman, Bombay Woollen Mills, Proceedings of the Bombay Strike Enquiry Committee, Vol III, p 1100.

36 The most detailed description of commissions, including these practices, is available in the representation of the trade unions. See, in particular, Bombay Textile Labour Union, Report of the Indian Tariff Board (Cotton Textile Industry Enquiry) 1927, Vol III (evidence of local governments, etc), pp 438-39.

37 Primordial ties figure in R K Newman, Workers and Unions in Bombay, 1918-29, Australian National University Press, Canberra, 1981; in Kooiman, Bombay Textile Labour, p 31; whereas in Ahmedabad, “the jobber normally represented the... authority of a caste leader”. Lakha, ‘Character of Wage Labour’.

38 Suggested in Gregory Clark, ‘Explaining Employment Institutions – A Historical Perspective’, University of California, Davis, mimeo.

References

Abraham, Katharine G and Susan K Taylor (1996): ‘Firms’ Use of Outside Contractors: Theory and Evidence’, Journal of Labour Economics, 14(3), pp 394-424.

Bombay Millowners’ Association (1927): Report of the Indian Tariff Board (Cotton Textile Industry Enquiry) 1927, Vol III (evidence of the Bombay, Ahmedabad and Baroda Millowners’ Associations), pp 138.

Chandavarkar, R (1994): The Original of Industrial Capitalism in India: Business Strategies and Working Classes in Bombay, 1900-40, Cambridge.

Clark, Gregory (1987): ‘Why Is Not the Whole World Developed? Lessons from the Cotton Mills’, Journal of Economic History, 47, pp 141-73.

Clark, Gregory and R C Feenstra (2001): ‘Technology in the Great Divergence’ in M Bordo, A M Taylor, J G Williamson (eds), Globalisation in Historical Perspective, The University of Chicago Press, Chicago and London.

Director of Information and Labour, Government of Bombay (1927): Report of the Indian Tariff Board (Cotton Textile Industry Enquiry), Vol III (evidence of local governments, etc), p 61.

Government of Assam (1927): Report of the Indian Tariff Board (Cotton Textile Industry Enquiry) Vol III (evidence of local governments, etc), p 25.

Gupta, Bishnupriya (2003): ‘Work and Efficiency in Cotton Mills: Did the Indian Entrepreneur Fail?’, University of Warwick.

Kannappan, Subbiah (1985): ‘Urban Employment and the Labour Market in Developing Nations’, Economic Development and Cultural Change, 33(4), pp 699-730.

Kiyokawa, Y (1983): ‘Technical Adaptations and Managerial Resources in India: A Study of the Experience of the Cotton Textile Industry from a Comparative Viewpoint’, Developing Economies, 21(2), pp 97-133.

Morris, Morris D (1960): ‘The Recruitment of an Industrial Labour Force in India, with British and American Comparisons’, Comparative Studies in Society and History, 2(3), pp 305-28

Otsuka, K G Ranis and G Saxonhouse (1988): Comparative Technology-Choice in Development: The Indian and Japanese Cotton Textile Industries Macmillan, London,

Wolcott, Susan (1994): ‘Perils of Lifetime Employment Systems: Productivity Advance in the Indian and Japanese Textile Industries, 1920-1938’, Journal of Economic History, 54 (2), pp 307-24.

Economic & Political Weekly January 5, 2008