about archives

Vvgnli

Ailh

Publications

Research Projects

Research Partners

Archival Sources

Acknowledgements

Events

Contact Us

Home Web Survey Search  Collection Map    

Outline of a History of Labour in Traditional Small-scale Industry in India

Tirthankar Roy

At 1800, India had a significant presence in the world as a manufacturing country. Possibly about 15-20 per cent of its working population, or 15-20 million persons were employed in industry at that time. All of it was small-scale artisanal industry, and the great majority of the production units were family-labour oriented or 'households'.       

The nineteenth century brought  about different conditions. Trade with industrializing Europe destroyed a great deal of artisan livelihood in India. At the same  time, the 'globalization' of the Indian economy through trade, investment and colonial rule initiated a limited growth of mechanized factory-based industries in India. These were usually industries that used abundant natural resources intensively. Examples  are cotton textiles, jute textiles, sugar, paper, iron and steel, etc.. The latter type of enterprise has been called 'large-scale industry'. All other forms of enterprise can be termed 'small-scale industry'.       

How   did the nature of industrial work change between  1850  and  1950? Mainstream  labour history in India has been concerned almost exclusively with labour in large-scale industry and more  or less ignored labour in small-scale industry. Implicitly, there has been  a  belief that large-scale industry is what defines 'industrialization', whereas small-scale industry ceased to matter in industrialization. The  former was expanding  in income  and employment, whereas  the latter was in decline in competition with large-scale industry. As a result of such beliefs, anyone reading mainstream labour history is likely to form an exaggerated view of the change in industrial labour in colonial India.

Recent works on small-scale industry have called for a more balanced reading. These works have pointed out that small-scale industry survived on a very large scale, and that segments within it modernized in the colonial period. Labour in modernizing small-scale industry must have changed more gradually and in different ways than did mill labour. This perspective raises the following question. How do we conceptualize 'work'  and 'workers' in modernizing small-scale industry? The present essay is mainly an attempt to answer this question.    

The essay has three parts. Part 1 (The Context) sets out the background, with a review of the state of labour history scholarship, the macroeconomic context, and with an attempt to develop a realistic typology of industry from labour history point of view. The  typology makes a distinction between 'traditional' and 'modern' small-scale industry. This essay will be exclusively concerned with the former category, because only on this category is there sufficient research to build upon.¹ Part 2 reports what we know  and can plausibly speculate about employment conditions in traditional small-scale industry in the long run. Part 3 offers some  broad hypotheses  on traditional small-scale industry in the post-independence period by using the survey of the preceding section. The 'conclusion' suggest some areas of research and issues in comparative history.

 Part 1. The Context

 1.1  Typology: Large-scale and Small-scale Industry  

Standard histories of Indian industrialization deal mainly with a type of firms described as 'modern industry' or 'large-scale industry'. Large-scale industry can be defined by three basic characteristics, relating to technology, organization, and government  regulation. First, large-scale industry used machinery and steam-powered technology. It was the relatively more capital-intensive sector in manufacturing. Second, it was organized in large factories sometimes employing several thousand persons, rather than in small factories or in 'households'. Households are defined as units where members of the owner's family were the main workers. Third, these large factories satisfied the official definition of a 'factory'. This definition has changed over time. Today  it applies to any unit employing 10 or more workers and using electricity or 20 or more workers and  not using electricity. Once such a unit is registered officially as a 'factory', it becomes subject to government regulations under the Factories Act concerning the wages and welfare of the workers. A unit officially registered as factory can sometimes  escape implementing  these regulations. But overall, the Factories Act has quite strongly influenced employer-employee contracts inside large factories in India.

By contrast with large-scale industry, in numerous industrial firms in India neither machinery, nor large factory, nor government regulation played significant roles. These formed  the relatively more labour-intensive component in manufacturing. We call this sector 'small-scale industry'.

Industry types can be distinguished further based on a fourth characteristics, vintage. Large-scale industry was a  new  type of enterprise. It was a product of the Industrial Revolution. Machinery, even the technicians, needed to run the mills were for a long time imported from Britain. The technological-organizational paradigm followed that existing in mid-nineteenth century Britain. The biggest business partner was Britain. Except in the cotton mills, the owners were mainly British.

For the majority of the small-scale industry firms, on the other hand, the products and technologies dated before the colonial period. An example is handloom weaving of  cloth. This set can be  called 'traditional' small-scale industry. There were, however, a few small-scale firms that were modern in origin. Compared to large-scale industry, these were units of smaller scale and usually unregulated. And compared to traditional small-scale industry, these were usually of recent vintage, used machinery to a greater extent, and had higher average scale. We can call this type 'modern small-scale industry'. In the interwar period, a significant growth of modern small-scale industry took place. Examples are foundries, cotton gins, jute presses, edible oil extraction, rice mills, flour mills, etc. Table 1 summarizes the main characteristics of the three types of industry.

The term 'modern'  can be misread and needs to be qualified. The dividing line between  traditional and modern small-scale industry was not a very sharp one. Most types of modern  small-scale industry supplied old products. Thus, grain milled by machinery and that milled by hand both supplied the same consumers, but by different technologies. Sometimes, however, new technology led to a new product altogether. For example, machine-milled rice was often seen as a distinct product from hand-pounded  rice, and in some cases these involved different grains. In many cases, money  made  in a traditional small-scale industry tended to be invested in starting firms that were bigger in scale, technologically more developed, and more capitalized. These would  satisfy the definition of 'modern small-scale industry', even though the modern  and the traditional in this case were two stages of a continuous process of evolution or adaptation. The example of handloom weavers who set up small factories equipped with power-driven looms is perhaps the most appropriate.

The  dividing line between small-scale and large-scale was sharper. But they too had close relationships. Large-scale industry supplied raw materials to small-scale. Workers  often  moved   between  them. And   small-scale industry workers and entrepreneurs sometimes learnt their skills and acquired new ideas by working in large-scale industry. The former could even buy secondhand machinery  from  the latter. Textiles supply examples of all three situations.

Table 1. Different Types of Industry Defined

In terms of conditions of labour, there was a great deal of continuity over time within traditional small-scale industry. Change did come. I shall argue later that a labour market came into being. But the change was a slow one and mediated for a long time by traditional institutions. By comparison, there was a sharper break when agricultural or artisanal populations entered large-scale industry. In modern small-scale industry too there was a sharper break, but we can say very little about it.

1.2  The intellectual context

History of industrial labour in India has dealt overwhelmingly with large-scale industry or mill labour. It has seen work and workers in the mills as representative of a transition from traditional to modem occupations and culture. It was an uneasy and incomplete transition in most views, but the core problem for labour historians nevertheless. Within this tradition, concerns have shifted over time, from the mechanisms by which an industrial labour force was formed, to the constitution of 'working class' culture.²

Mainstream labour history in India has more or less completely ignored small-scale industry. The oversight is curious. For, the great majority of manufacturing workers remained, and continues to remain, employed in forms of enterprise that fall outside 'large-scale industry'. At 1930, as little as 5 per cent of industrial employment were  located in mechanized and formal factory units, a small subset of which has engaged  almost all of labour history scholarship. The rest were employed in such occupations as handloom  textiles, leather, pottery, metal-working, jewellery, wood­work, etc.. Most of these were traditional and small-scale. In 1991, the percentage of formal factory workers was less than 30 and in steady decline since 1971. The rest of industry was probably not dominated by  traditional small-scale industry any more. But a significant part of it had originated from latter roots. The labour historians' preoccupation  with mill  labour has clearly no  relation with  the quantitative importance of such labour in India. The sources are in fact ideological.

One source of this preoccupation is the analytical categories and framework used in conceptualizing industrial labour. The language of Indian labour history is rooted in the Marxist concept of the 'working class'. By definition, the 'working class' is cut off from traditional industry and employment practices, or the debris of 'precapitalist relationships'. Indian labour history of recent vintage is primarily engaged with measuring  the 'classness' of the working class. Labour in traditional small-scale industry cannot be analytically handled by this scholarship and, therefore, is forgotten by it.

A second likely source of this oversight is a widely held belief that large-scale industry effectively supplanted traditional small-scale industry in the colonial period, that these two were  basically competing oppositions, and the former was clearly winning the competition. In labour history this view has remained implicit. In the old nationalist interpretation of colonial India the view was more explicitly stated. In recent industrial history, which has been strongly influenced both by nationalist thought and by Karl Marx's writings on India, the idea revived. Marx was generally dismissive of traditional small-scale industry, even more so in the context of colonial countries.

In this view, economic contact between India and Europe produced three changes. First, it destroyed traditional small-scale industry. Second, it created large-scale industry in its place. And third, the new enterprise did not compensate for the loss of the  old so that  a net retardation or 'de-industrialization' resulted from trade, investment, and colonial rule. Such a view of Indian industrialization has been so influential in the 1970s and the 1980s that it merits being called the 'orthodoxy'. The main evidence in support of the orthodoxy came from two sets of facts. The first is an old  and misinterpreted example  of decline in the traditional textile industry in competition with mills in Lancashire and later those of Bombay. And the second is a decline in total industrial employment during the censuses of the colonial period (1871-1931). We  shall re-examine this evidence later on.

Elements of this orthodoxy have been in dispute from time to time. But in recent works in the 1990s on industrialization in South Asia, it has not only been questioned again, but also questioned in a manner that an alternative paradigm can take shape.³

In this alternative view, wit h some exceptions (some branches of textile industry for example), traditional small-scale industry did not compete with large-scale industry. As a result, vast numbers of such workers and occupations could continue. But they did not continue in a changeless fashion. The main source of change was commercialization. Most  types  of traditional small-scale industry experienced increased long-distance trade from the middle of the nineteenth century. There was decline in production for subsistence, in production under various types of non-market distribution arrangements, and in production for local, rural or periodic markets. Markets  integrated due to the railways and to  standardization in consumption. Commercialization  increased competition within traditional small-scale industries. That, in turn, led to increased specialization and division of labour. Less efficient and non-specialized forms of enterprise, such as the households, declined in favour of specialized wage-labourers. The decline in census employment reflected this internal competition.

This alternative paradigm can be called a Smithian-Hicksian view as opposed to the Marxist view that has ruled Indian labour history for a long time. The analytical tradition associated with  Adam   Smith  and   John   Hicks  tends  to  stress commercialization as the key to economic growth. The tradition associated with Marx tends to see machinery as the centrepiece of economic growth. The former view is capable of explaining a decline in manufacturing employment, and yet it does not see traditional small-scale industry as such as declining.

How does a paradigm shift in industrialization change the writing of labour history? How do we redefine 'work' and 'workers' in manufacturing in the new context? Needless to say, the new view encourages seeing in industrial labour more a continuity than a violent break. But what remained the same? And what changed? In Part 2, these questions will be addressed. Before we come to that it is necessary to form an idea of the quantitative extent of large-scale and small-scale industry.

 1.3 Employment  statistics

The relevant employment statistics are presented in Table 2. By studying these figures, we can come to a few reasonably firm conclusions on industrial labour.

In the colonial period, industrial employment was stagnant in absolute terms. It may  even have declined by comparison with the nineteenth century. The exact size of the labour force can be disputed, but this conclusion is unlikely to change. There was stagnation also in industrial employment in relative terms. That is, industry's share of total employment hardly changed. This inertia or stagnation derived entirely from trends in  employment  within  small-scale industry. If we measure  'large-scale industry' by employment in officially registered factories, and 'small-scale industry' by employment  located outside such factories, then the scale of the latter fell quite substantially between 1881, 1911 and 1931.

In 45 years after the end of colonial rule in 1947, industrial employment increased in absolute terms. But the implicit growth rates were not very high. Between 1961 and 1991, the growth rate was a little over 1 per cent per year, substantially below the rate at which supply of labour (population growth multiplied by participation rates) expanded. Not surprisingly, the stagnation in industrial employment in relative terms continued in independent India.

For independent India, however, it is quite clear from census data that the low rate of growth in employment did not signify an overall decline, but a decline in certain segments. This pattern reflected an organizational change that improved labour productivity. Family-labour oriented units were declining rapidly, and small wage-based  units expanding even more rapidly. What appears as an overall inertia, was actually a process of growth in wage labour and specialization.

We do not have the means to test directly whether or not such an organizational change had already been going on in the interwar period. But we do have several sets of data that suggest that indeed it had started long before 1947. First, women's participation in industry declined steadily. Women were and are overwhelmingly concentrated in household industry. Therefore, this decline in participation reflects the decline of the households. Second, real income in 'small-scale industry' increased significantly between 1901 and 1947⁴ even as employment declined, which suggests that the decline affected specifically low-productivity workers. These were likely to be households. Third, there are reliable data from major types of traditional and small-scale industry - chiefly leather, handloom textiles in cotton and silk, carpets, and a few others - which suggest that the scale of production increased in the first half of the twentieth century with stagnant or even declining capacity indices. Once again, we can postulate a shift from low-productivity units to higher-productivity ones.

To sum up, during the colonial period, employment was shrinking in some segments  of manufacturing, whereas employment was rising in other segments. In the most  widely held interpretation, 'small-scale industry' as a whole was in decline in competition with large-scale industry. This is disputable. It is more likely that small-scale industry as a whole was a dynamic  sector. But certain segments and certain forms of industrial organization within it were in decline. The trend derived to a lesser extent from competition with large-scale industry, and to a larger extent from growth of markets and competition within small-scale industry itself. Part 2 on labour begins by elaborating this point about internal commercialization.

 Table 2. Employment in Industry

Notes: The extent of large-scale industry is measured by employment in officially registered under the Factories Act. Any unit employing 10 or more workers and using power  should be registered, but in fact, very few units with around 10 workers actually register. Units usually registered have a much higher average work-force. 'Small-scale industry' means all units not under the Factories Act. 'Family-labour oriented' units are called 'Household Industry' in the census. 'Wage-labour oriented' units are the residual, consisting of small factories and workshops.

 Source: India, Census of India, various years; India, Statistical Abstracts for British India, various years; India, Statistical Abstract, various years.

Part 2. Work and Workers in Traditional small-scale industry in Colonial India  

2.1 Commercialization

The sixty years between the opening of the Suez Canal (1869) and the Great Depression (1929) saw an almost continuous growth of external and internal trade. A national market emerged  in a  number  of basic goods and  services that were imperfectly if at all traded before. Labour became steadily more mobile. The basic infrastructure needed to hold a market-economy together was more or less completed in these sixty years. Large-scale industry was a product of these changes. Traditional small-scale industry was also transformed by these changes. There was increasing integration of the market for its products. There was a shift away from production for own  use, or use as gifts and tributes, to production for market, especially non-local market; shift from local to long-distance trade; and changes in consumer and producer behaviour induced by the possibility of long-distance trade.

As markets integrated, competition within the crafts intensified and eventually weaker  manufacturing  traditions decayed. The weaker traditions were represented universally by the rural artisan. This was the context in which household labour was extensively employed. This was also the context in which the male artisan was very often a part-time agricultural labour himself. Broadly speaking, rural industry had the following core set of attributes: (a) production for subsistence (of for the producing household, community,  or village), (b) primarily barter, grain-based exchange, at best production for a  small local market, (c) peasantry-dominated social relations of production, and (d) narrow or non-diversified markets and skills. Examples were leather processing, coarse weaving, utility metals, etc. Commercialization could upset this network by altering the structure of costs as well as tastes. For example, competition drove out the non-specialized rural artisan who almost always produced a poorer quality good  even if at a lower price than in the cities. Expanded cash transactions drove many artisans out of barter and forced them to migrate. Persistence of restrictive social relations in the villages had a similar effect, when much improved communications and a less hierarchical society in the towns increased the costs of staying on in the village.

The cities and rural clusters that had artisan settlements and that emerged as main  points of trade, were well-placed to receive craftsmen, diversify occupations, utilize old skills in new ways, and also compete more fiercely between themselves. The advantages of these places were manifold. They were often production centres for major inputs such as cotton yam. They were invariably the centres of trade in both raw material and finished goods. Thus, they had merchants who knew these trades better than the artisans. Being less hierarchical places they often provided artisans themselves with the opportunity to become traders.

This agglomeration tendency further increased specialization via a restructuring of the production unit. Women  migrated  less frequently than men. Therefore, the household as a production unit was much less common  with migrant workers. For a number  of reasons, factories or closely supervised putting out systems expanded. Finally, as specialization and division of labour progressed, new quality-improving or cost-reducing innovations could be tried. Almost always, these innovations did not replace labour outright, but made labour more productive. Before coming to the impact that markets had on labour, some  specific industry experiences are cited below to illustrate how commercialization could change institutions.

The most  important example of outright market-creation is leather, especially tanned hides and skins. This product became a major tradeable in the late nineteenth century. The capital and expertise accumulated in this trade was the foundation for a diversified leather industry to build up later, now one of South Asia's leading exports. This was originally a rural craft. It was performed by quasi-serfs under conditions of extreme inequality. In most places hides were subject to barter. Even where a market formally existed, servitude arose both from caste-hierarchy and the interlocking of markets. That is, it arose from the fact that the main customer of the leather artisan was also the peasant-employer. An export market from the 1870s, and the need for different kind of processing weakened this serfdom. The market enabled massive migrations into the city, the rise of the merchant-owned factories and rise of wage labour.

The case  of handloom  weaving is more  well-known. Here, the vulnerable segment consisted of artisans who had the following features. They were engaged in cotton cloth as opposed to silk, in coarser and plainer type of weaving as opposed to skilled decorative weaving, were mainly rural and included a large number of artisan-cum-labourers or members of agrarian labour castes who wove on the side. Two types of weaver were  forced to look for alternatives. The first included those engaged in coarse-medium  cloths and facing  competition from the powerloom   in increasing intensity. The second included those engaged in rural coarse weaving who stayed poor and vulnerable because of their dependence on local markets. There was exit from both spheres. But the former class tended to look for alternative options within weaving. They  migrated to new  urban textile production and trade centres and restarted as weavers. The latter tended to leave weaving when alternatives, usually casual labour of various kinds, became available. In contrast with both, the silk weaver never faced competition from machinery in a serious way, but was affected by other factors shaping the textile market, clubbed under 'commercialization'.

In handloom  cloth, especially silk, long-distance trade was far from a new invention. But trade almost certainly increased in extent in the colonial period because of and through four processes. First, mass-produced and mass-consumed cloth from England  and the Indian mills created large-scale marketing networks of recent origin. Second, introduction of imported inputs had a similar effect on trade network. Third, there was concentration of weaving at points of internal trade, the railways and spinning mills. And fourth, there was universal decline of spot sale in favour of contractual sale, a  process contemporaries called increasing 'dependence' of the weaver  on  the merchant. How  greatly the work-site changed as a result depended on the extent or nature of relocation of the industry. In a group of towns in western India that became both points of trade and major destinations of handloom weavers from depressed and less competitive regions, the industry was reorganized in factories. In semi-rural weaving, and in other regions, the household proved much more stable.

A  range of industries intensive in craftsmanship - carpets, shawls, engraved metals or silks - were always urban and always commercial. In the colonial period the extent of urban concentration increased and there was a qualitative change in markets. In many   examples of precolonial markets, these products went to consumers with relatively little intervention from merchants. These consumers represented local power or custom. For example, carpets or engraved metals made in the towns of the Western gangetic plains were sold to the local rulers; or silks in Peshwa towns were governed by royal patents. This pattern of consumption changed in favour of exports or long­distance trade within India. In these crafts, the greater part of production and exchange in the mid-twentieth century happened in large agglomerations of artisan community and trade, and were subject to many layers of merchants. In other words, the twentieth century craft town was a place where bargaining, freedom of entry and exit, anonymity of transactors, were greater that before. The markets were both more complex and more competitive. In the highly skilled crafts, the old production units survived to a great extent because  they served the essential functions of preservation of skills and restriction of skills. The institutional changes experienced here were more subtle and slow-acting.

2.2 Work  and Workers

How  were work  and workers affected by commercialization? Words that have a known   meaning  in the context of large-scale industry - 'work', 'wage-labour', 'factory', or 'labour market' - reappear in the context of traditional small-scale industry, but in a changed meaning. This is because commercialization of work lagged behind  commercialization of products. In other words,  in employment conditions tradition persisted longer, even if in an adapted form.

The most strikingly 'new' type of work-site was the factory. In the interwar period, factories were usual form of work-site in the industries describe above. In major handloom  towns and destination of migrant labour - Sholapur being the most important example  - factories predominated. In carpet-weaving in northern India, factories were the usual system. In leather, again, wage-labour was performed inside large tanyards. But such factories could be different from a large mill in essential aspects. Generally, there were two broad types of factory in traditional small-scale industry. The first employed wage-labour, and only wage-labour. The second was a site owned  by  the merchants of the product, where putting-out contracts were executed by several groups of intermediaries who brought their own workers. The second was quite a different thing from a modern factory.

Except where factory in the former sense prevailed, and where large-scale migration had created a pool of workers available for hire, there was no explicit 'labour market' in existence. The former sphere had grown in the colonial period. But in many industries, it was a relatively small segment.

The general case was that of recruitment of labour into a small firm without full-fledged hiring. This happened in broadly two ways. Firms using mainly family labour employed workers from within the family, or sometimes hired surplus family labour of neighbourhood firms. And masters hired apprentices. The family firm and the master-apprenticeship system were the two general pre-factory types of unit that survived  the  colonial period  and  participated in  long-distance trade and industrialization.

In both cases, children were employed. But as long as they were employed within the family, within the master-apprenticeship system, or within localized limited hiring of non-family labour by families, children worked without the existence of an explicit market for the services of child labour. Increasingly in the post-1947 period, a market  for the services of children grew up. Today, the traditional contexts of employment  of children has disappeared, and children are recruited on wages from an available pool of hirable, often migrant, child labour. The same broad trend can be seen in the case of the adult male labourer as well, and probably for adult females.

 2.2.1 Factory

The forces that led to the significant growth of factories in certain industries were the following. In some cases - such as Sholapur cotton handloom or tanning generally - large-scale rural-urban and interregional migration broke up household or communal   systems of production, and created concentrated pools of hired labour. In general, wherever migration was subdued, wage-labour was slow to grow, and grew in a more subdued way in the form of families hiring assistants. The examples are, Tamil Nadu handloom  weaving, carpet weaving in Eastern UP⁵, etc. A second reason was the employers' or merchants' need to supervise production especially where the product was intensive in craftsmanship. This was one reason why carpet weaving came into the factories in Western UP. Thirdly, there was a more conventional way the factory begun, when a craft partially mechanized. Textile processing is one example. In Surat town, processes connected with gold thread manufacture represent another example.

 2.2.la The Handloom factory: Symbol of artisan accumulation: 

In both tanning and carpets, the factories were established by the merchants. In handloom weaving, they were mainly established by persons of weaving community.

Sholapur, the premier textile town of the Bombay-Deccan, represented the highest all-round development of handloom production in the twentieth century. It became  a mill town and a major cotton trading town in the decades after the railways connected it with Bombay. From the 1890s, Sholapur began to receive a large number of migrant weavers from Telengana. Some of them entered the mills. But many settled down  as handloom weavers. In the interwar period, the town probably had 7-8000 looms located in small factories owned mainly by the Telugu-speaking Padmasali weavers. These weavers dominated capital as well as labour.⁶ By 1930, there was a mixture of small family firms with 3-8 looms and a few hired workers and large factories with 20-40  looms worked mainly  by hired workers. The second type dominated. By then, loomless migrant weavers staying in dormitories and available for hire had become a large and visible group in the town. Increasingly in the interwar period, weavers came with their families. The women and children often found work in the winding-warping-sizing mill attached to a factory or catering to several of them. These factories employed relatively faster techniques of processing than those used inside households. Many of these families retained their ties with the places of origin through frequent homeward journeys. But the frequency of these trips came down over time. The existence of a large wage-labour group had also invited attempts to organize them into unions. At the end of the interwar period, the Handloom Workers' Union had been formed and managed to enforce two relatively minor provisions of the Factories Act, regarding hours of work and  weekly  holiday. The 1940s saw  the first strikes in the industry. But after independence the impetus to form unions weakened on its own.

 2.2.1b Tanning: The role of migration and hierarchy: 

The rise of the factory and wage-labour happened  in tanning the most extensively, in the backdrop of large-scale migration of tanners from the countryside to the hides and skins trading towns. The migration was most visible in the case of North or Central Indian Chamars who entered labour in Kanpur, Agra and Calcutta. In South India, on the other hand, there was extensive entry of local agrarian labour castes into tanning factory labour.

The number of large tanneries in India rose from 13 in 1901 to 66 in 1939, and labourers working in them doubled. Of 1921 census employment in leather, factories employed  5 per cent in United Provinces, and 4 per cent of tanning workers in Madras. But  these figures are an underestimate because the 'factories' here refer only to concerns officially registered as such. A provincial survey of unregistered factories was carried out during the Royal Commission on Labour, 1931.⁷ The survey showed that, in Madras, 776 small tanneries employed close to 10,000 workers, whereas large tanneries employed  4,000. According to the 1931 census, the small and large factories engaged about a quarter of the workers in tanning in this province. It is fair to assume that the proportions were similar in India as a whole, and rising steadily. While tanning was usually a male occupation, in some factories of Bombay, women and children were employed  sorting wool.

The word  'factory' could include a wide range in sizes and types. A 1940 report on Punjab described a hierarchy of tanning units in the province.⁸ The description seems to be applicable to most parts of India.

At the smallest scale was the 'village tannery'. This was still mainly an artisan-owned  unit. It worked with family and neighbourhood labour, used local cattle, but was contract-bound to the hide merchant from the small town. At the middle was the 'town tannery', factories with 5-10 hired workers each. The owners took part in both contractual and spot markets. A fair number of them were of artisan background. But the majority was hide traders, usually Muslim. At the top of the scale were the 'big factories' employing 100 or more workers each. There was some European capital in these ventures. But the main source was the larger of the Muslim merchant firms. The first type could be found scattered in rural North India. The second occurred in small towns, but usually those with slaughter-houses and large spot markets or auctions (mandi) in hides. The third was a feature of major ports and industrial towns such as Calcutta and Bombay. The  labour recruitment practices in the largest tanneries were probably similar to that in the mills. But this history is yet to be written.

In general wages of the least skilled worker inside a tannery were not very different from the wages of an agricultural labourer. In that sense, the tanning workers' low opportunity costs helped the owners of factories. On the other hand, the tanning worker benefited from a freer social environment and steadier work. There is evidence to suggest that economic mobility was faster and more likely inside a tannery than in agriculture. About 1900, the wages of the ordinary 'cooly' was about Rs. 4 a month. The unskilled labourers could easily command a wage of Rs. 5 to Rs. 8, and the skilled Rs. 10 to Rs. 30. Either of them could live comfortably on Rs. 5 a month.⁹

 Custom  did not completely vanish, but often persisted in the tannery in the form of direct or indirect hierarchy. Hierarchy could permit not only low wages but also poor working conditions inside the factory. Further, the distance between the workers and the managers was apparently quite sharp inside a tannery, and remained so for a long time. Inside factories in Madras, for example, the main body of manual workers was drawn from the agrarian labour castes, chiefly the Paraiyans. The white collar jobs, on the other hand, were sometimes performed by people who owned land. In such cases, the interactions between the 'tannery men' and the 'factory men' reproduced, in a much  milder way, the hierarchy in the villages that both had left behind.¹⁰

 2.2.1c Carpet: Interplay of export-merchants, master-artisans, and their apprentices: 

Major  urban centres of carpet weaving in India in the interwar period - Amritsar, Srinagar, and Agra  most notably - saw the establishment of large factories with sometimes  as many as 300 looms each. Most of these were owned by foreign carpet trading firms. Some  of them had  operations in Persia and Turkey, which  were destroyed during the First World War, possibly with the effect that their India trade grew  and the Indian factories became more visible. There were several reasons why the factory suited the carpet industry. Certain operations, chiefly dyeing, needed new investment and systems of quality control. For delicate or expensive work, weavers were  not always trusted enough to be left unsupervised. But most importantly, carpet weaving  in these towns belonged in an Indo-Islamic craft tradition that used master-apprentice  systems  most  extensively  and  rarely used  female   workers.  A predominantly  male and team-work-based  production unit was much  more  easily assembled in a factory than families.

The carpet example is rich in details about contractual terms in the urban skilled crafts of northern India, and for this reason it merits a detailed look.

The employment contract in the factory was not a wage-contract. It involved three parties and not two: the owner, a master, and workers. Typically, the owner contracted on piece-rate with the intermediary, also called ustad or karigar. The latter created his team. The team consisted of both grown up and trained men, and young apprentices called shagirds or disciples. The first group sometimes appears as partners of the ustad, sometimes as his employees. The second  group was  almost always engaged on time-rates. For the tasks they did were wide-ranging and ill-defined. In the large factory, the ustad was an intermediary. In smaller factories, the ustad and the owner (called karkhanadar) often became indistinct. The karkhanadar was technically one who owned  the factory. But in practice he was an owner with artisanal background. The large factory was no factory in the modern sense. It was the shed where a putting-out contract was implemented. The 'employer' received the benefit of putting out, terminable contracts as a strategy towards risk. The employer avoided one disadvantage of a factory, having to communicate with and monitor the direct producer. The shed ensured that the work progressed under  their eyes. All these features suited the European factor very well.       

The  artisans, on the other hand, found the factory consistent with an old tradition. Until the mid-nineteenth century, carpets were made in these towns in factories or karkhanas owned by courtiers and rich men of the town. These factories employed a class of skilled masters known as ustads, literally maestros, and 'karkahnadars'. In the twentieth century, the weaver-intermediary symbolized two elements, of which one was old and one new. The old element was his function as a repository of skills, one of the main senses of the term 'ustad'. In this sense, the work-contract explicitly accommodated learning. Workers were supervised by a  super-worker who  taught them the craft and had their trust and respect. When the master moved  from one  factory to another, he took a whole team with him. On the other hand, the ustad was a contractor under the merchant or karkhanadar. When he worked under a European factor, the ustad was a product of foreign trade.

Three types of customary relationship, cooperation, credit and apprenticeship, gave strength to north Indian karkhanadari. The first occurred among ustads or among karkhanadars. The second occurred between the ustad and his merchant employer, or between the skilled workmen and their ustad/karkhanadar. The third bound unskilled workers and new  entrants to the ustads. These relationships enabled trust, supply of labour, formation and retention of skills, and helped producers respond to expanded market with relative ease.

Cooperation, or what a contemporary writer misleadingly called 'guilds', is a state where masters of apprentices combined to regulate wages, quality, 'limit[ed] supply to demand', and 'settl[ed] trade disputes'.¹¹ In the prewar decade, such controls could still be seen in places where the industry was in decline. But, in towns that were growing very rapidly, there is very little trace of informal guilds of this kind, not to mention formal ones. in the long run, such customary forms of cooperation came under pressure from sustained new entry in carpets, the ownership of factories by merchants, the decline of the old type of apprenticeship system (see below), and the entry of rural families in carpet production as in the Mirzapur area. By and large, even within the period of  the book, cooperation seems  to have  given way   to market-mediated coordination systems in which the European firms played a major role.

'Indebtedness' evokes powerful negative images of the crafts in present-day India. Post-independence official statements about small-scale industry unfailingly equate debts with exploitation of the generic artisan in the hands of the generic capitalist. Existence of debts tends to be seen as sufficient ground to replace private trade networks by the government-aided cooperatives. In colonial India, debts were pervasive in the urban crafts. But they appeared in a manner that cannot be boiled down to the term 'exploitation'. Usually, debts represented an exchange of privileges. It offered the employers an instrument to secure steady labour supplies. That instrument came  at a cost. Its effectiveness depended on the condition of the industry and the worker in question.

The  trader who wanted a stable contract had to offer a banking service to his ustads. In Amritsar, the former could be compelled to take over weaver's debts to the moneylender.  For otherwise, a hard-up weaver often went into hiding, looked for alternative jobs, and became irregular at the factory.¹² Indebtedness, in this case, reflected competition for the master-worker's services, and could rise so high in good times as to put the employer at risk in bad ones.¹³

The  smaller factory-owner or karkhanadar, likewise, never actually paid out a sum  to his employees-cum-partners or employees-cum-apprentices at the end of a task. Regularity in payments was a feature of trader-karkhanadar contract, but almost never of employer-worker contract. The worker withdrew money as he needed. The employer wrote his dues as the tasks were completed. The employer kept an account that was, all descriptions concur, transparent to both parties. Disputes, and the charge of breach, were  unheard of. The financial status of a karkhanadar, and of the best workers he employed, were  similar enough to ensure roughly equal bargaining power. The account was  checked about four times a year, on four festive days, and finally settled on one of these. They were interest-free. The Muslim artisan had an aversion to interest. But nor did the Hindus, forced by competition or by custom, indulge in usury.

The baqi, as such loans were  called, thus, performed an essential banking service to the workers. It smoothened transactions and enabled them to provide security to their savings. This service was freely available. In carpets, the system existed in Amritsar and Agra. It was rarely seen in small places. The urban worker, unlike the peasant, did not possess collateral, and thus needed a banker willing to lend without tangible security. The only pledge he could offer was a commitment to work, and the only person who could cash that pledge was, not a moneylender, but the karkhanadar. The cities that looked for a stable labour force invariably offered this arrangement.

The skilled artisan, however, was always a valuable asset in industries like carpet. Thus, baqi was also the means to control labour. Supply of loans was a means to enforce quality, where for example deductions were made for bad work. It was a means to retain the best workers.¹⁴ The towns displayed not only reputation for honesty in accounts, but respect for the rule that no employer would engage a worker who was indebted to another karkhanadar, except by repaying that loan himself. In practice, the average worker  was perpetually indebted. But the best ones were clearly so out of choice. For, it was never a problem for a good worker to clear off debts from current incomes.

This ambiguity was  an essential feature of the system. Employers could be obliging, or use debts to bargain wages. The combination of loan account and piece-rate payments was  potentially a complication the workers might fail to understand. It could also lead to a condition where the workers remained tied even when there was neither work  nor wages. Whether the system favoured the workers or was used against their interest depended on demand  for labour. In a rising industry, employers needed to secure labour supply, but debts as an instrument for labour-tying became loosely enforceable. A good  worker invariably had employers competing  for him. For the interwar period in general, baqi in the United Provinces towns appears to have been driven by a persistent shortage of skilled labour. During decline, on the other hand, workers might not be able either to leave or to earn. And, as the decline itself might make  credit supply more monopolistic, they faced harder bargaining on wages and interest rates. Echoes of these adversities came from small towns in decline.¹⁵

 Apprenticeship was   also an  institution inseparable from such three-way contracts. The apprentice was a young  boy  in training, a student, and never 'the household drudge  of the master'.¹⁶ The training aspect was reflected in the generally accepted duration specified for an apprentice to be recognized as a master in the making. A  credit account started as soon as an apprentice was taken. It began with an exchange. Gifts to the master were returned in the form of a loan to the worker.¹⁷ In most cases, the talented worker could repay the loan. But more often, he persuaded the next employer to take it over.       

In no other craft were children more extensively employed outside the family. The  possibility of employing children derived partly from the work itself, partly from tradition. There are three levels in a decorative craft, conception of design, execution or the pure physical labour, and coordination of the two. In most complex crafts, the first level is separable from the others, whereas the latter two are variously tied in the labours of the main artisan, the handloom weaver for instance. In carpets, all three levels were in principle separable. The task of knotting piles was simple and not too exhausting. And by a few ingenious devices, it did not need the reference of the whole carpet. One such  device was the talim, a system that came from Kashmir, but has distant relations in central Asia. Repeatable sections of a design were transferred onto graph paper (called 'cartoons' in west Asia), and works as a reference for the weaver, where he or she is an adult. Where he was too young, the matrix was written out, in a coded script to avoid piracy. With the help of this graph, each knot could be a coordinated by oral instruction.¹⁸ These tasks needed separate and literate groups of workers. Once these tasks separated from weaving, the weaving itself could be done by persons skilled only in tying knots. These were boys between nine and fourteen. They followed a 'reader' behind them, who dictated the oral instructions in a rhythmic chant.       

Usually, a normal-sized loom in an Amritsar factory involved four persons. Two of them were boys under twelve years, the third a boy in late teenage, or a senior apprentice, and the fourth the master himself.¹⁹ Apprentices were divided and hierarchized according to whether they were domestic, hired, or quasi-domestic, and according to the age of entry and duration of stay. It is possible that some of this difference derived from the condition of the industry. Good times induced quicker entry. The usual age of entry into apprenticeship was nine, though six year olds were not unknown.  The  boys earned an income  about half that of an adult weaver. On becoming  an expert weaver, a distinction attained in late teens, they earned a wage somewhat  above  the average in textile crafts. As they reached this stage, chances increased of their leaving for a new job, 'just as they are beginning to be really useful'. There were abuses too. In Amritsar, for example, the wages of the boys remained rigid as they rapidly improved as workers.²⁰ This, however, was not the general case. We do come  across examples of productivity-adjusted apprentice earnings.  

The  Amritsar data shed some light on the process of recruiting a child into a carpet factory. It involved an intriguing contract, in which apprenticeship, tied labour, and debts are so closely intermeshed that it is impossible to say which was the dominant compulsion. Typically, a man, not infrequently a weaver himself, brought his son or grandson to the master, and took a loan from him, to be repaid from the wages of the boy. The boy  worked  as an apprentice for a few years and, if a good worker, was considered an adult entrant into the labour force like any other. Merely the existence of an advance does not signify the motive for the contract, all apprenticeship start with a ritual gift-exchange, In less commercial and more tradition-bound crafts (such as music), the apprentice had to make substantial gifts. In a commercial craft like carpets, the ustad seemingly paid a higher amount. In either case, there was no trace of usury in these transactions.

In the normal course, apprenticeship was a joint learning-hiring system. Good times induced quicker graduation of apprentices into independent contractors, if not into ustadhood.²¹ On the other hand, in bad times, or in a state of decline of custom, apprenticeship could degenerate into exploitation of new entrants and weak workers, without necessarily the promise of a better future.

In the craft monographs of the early twentieth century, children workers tended to be referred as a curiosity, at best in a tone of reassurance:

        I watched the boys [at the Otto Weylandt factory at Agra] for some time.

        They looked quite happy, and the work did not seem to be arduous²².

But by the 1930s, children in the carpet factories had become extremely visible, and begun to pose a delicate legal dilemma. Perhaps in no other industry would one see pre-teen and early-teen boys constituting an almost entire factory's work-force. And where the owner of the factory was a European, and not one of the artisan caste with whom one might take an apparently offensive practice for granted, a great deal of explaining took place.

An  occasion arose during the Royal Commission   on Labour  which, in the course of evidence in Amritsar, met George Stevens and Gerald Davies of East India Carpets, and two master weavers as well. The central theme in the probe that followed was children. The managers' defence of the practice of employing children rested on two grounds. First, the contract with the master left actual employment and labour management  the latter's prerogative. And second, there was an impression that children in the Indian carpet factories were probably better-off both in wages and working conditions than in Persia. The conditions in Persia were a legend, and had apparently invited some sort of censure from the League of Nations. It is more likely that, given the ignorance of educated  Indians about conditions anywhere outside the British Empire,  the  Persian  case  could  be  conveniently exaggerated  by   European manufacturers. From 1930, labour legislation was more than a credible threat. Who would  bear the cost of substitution of children by adults, if labour laws came to that? The managers of East India carpets pointed at the ustad, who was solely responsible for employment   contracts.²³ That solution was consistent from the merchant's point of view, but inconsistent from the ustad's point of view. For the child apprentice was the counterpart of the ustad, one could not be sacrificed without the other.

The problem  did not need immediate action in the 1930s, especially since the market for Indian carpets abroad revived greatly. But after 1947, two circumstances affected these customary transactions vitally. First, numerous artisans, especially the ustads, emigrated from India to Lahore, Multan and Lyallpur. The effect in Amritsar in particular was devastating. And secondly, the formal sanction on child labour now became  too strong to ignore. From the 1950s, the north Indian factories began to decline. The karkhanas and karkhanadari contract became obscure, and with it went many  of the ustads and aspiring ustads. And yet, interestingly enough, child labour did not disappear. It merely reincarnated inside the family firms in the semi-rural Eastern UP  industry. More on this change in Section 3.4 below.

 2.2.2 Family firms and family labour

Domestic work  was widespread, but except in weaving, not enough is known about it to enable us to generalize on divisions of labour and the nature of work and leisure. There were two divisions of labour inside a family - one based on sex and the other based on age. Inside a weaving household, one would generally see adult men working as weavers, adult women on winding  and sizing operations, and children as assistants in both weaving and winding. In silk, one specific task was always reserved for children, manipulation of dobby. More generally, children were employed in handling any system of manipulating the warp-threads to institute designs, other than the jacquard. Both boys and girls in weaver families operated the harness. But if these were  boys from the family, they would be put to weaving by the time they reached teenage. If girls, they stopped working at about that age. If they were boys from other families, this work was considered apprenticeship, when, 'through observation and practice they gradually pick up the skill of weaving..'.²⁴ A jacquard can easily replace this labour. But it cost money. In this way, scarcity of capital might have encouraged relatively larger family size in artisan families, for which there is some independent evidence.

Were  the child worker happy inside his home? The alternative, of course, was rarely a school. In most decorative crafts, the alternative was working in the workshop of a family friend, or in a city, under a master. We do not know enough to compare the two situations, that of being a helper in own home against being a helper in someone else's home. But between factory and home, we need not hastily assume that the home was necessarily a haven for the child. Boys themselves might well have preferred the company  of other boys, which was  only possible in a factory. And as N.G. Ranga harshly reminds us,

 .. parents are indeed  veritable exploiters .. and this exploitation, however,  merciless, inhuman  and  unjust, is shielded by so-called parental authority and filial devotion. The evil decreases in intensity and barbarity as the earnings of the families increase.²⁵

 2.2.3 Neighbourhood  hiring of children

In industries and regions where the main system was family labour, children were available for hire in large family firms. Such hiring necessarily involved boys, and never girls. We can call this system 'neighbourhood hiring'. This was extensive and  possibly  long-standing such  that it seemed   in 1930   as a  near-formal apprenticeship such as that in carpets. N.G. Ranga wrote about such a system in South Indian silk weaving. His respondents explained apprenticeship system in the same terms as a school, a learning institution, but also a disciplining one. The loom was meant to discipline unruly children, who would otherwise grow into 'disorderly youth and men, predisposed to drunkenness and brawls'²⁶.

In Dharmavaram silks, boys were employed in family or outside family to learn the craft. Weavers who had more than one son let the younger sons work for another weaver who needed assistance. In the first six months of training, the boys got nothing. Thereafter they took a token wage, about a tenth of adult family earnings a month. On attaining adulthood, around the sixteenth year, the father and prospective employers discussed whether the boy was reliable and skilled enough to deserve adult wage. If he was, the boy then started saving up to buy looms. In a year or two the apprentice could begin to contract on his own.²⁷ An important point added in connection with the silks of Peddapuram and Uppada is that, the more complex the work the more pains did the master take on the novice, and lower were apprentice wages. This implicit payment is not a general feature. On the question of payment, two motives, labour-shortage and training costs, often pulled in different directions. In the fine cotton weaving of Salem town we find reference to the contrasting situation, so great is the demand [for labour] that employers will give an advance of Rs. 100 to Rs. 150 to a boy's father in order to secure the boy's services. No interest is charged but the advance is regarded as a debt due from the boy.. By this system it becomes possible for a thrifty weaver to make money  out of a large family.²⁸

 2.2.4 Master-apprentice system

Carpet is the best example of formal master-apprenticeship. The system was, in fact, prevalent in most skilled crafts performed in the cities and towns of the UP by Muslim  artisan groups. This formal and more or less conventional contract had four features: (a) wages were historically decided, (b) there was collective control on terms of  employment,  (c) apprenticeship led to entrepreneurship, and (d) wages  were inversely related to complexity, that is, wages contained an implicit training fee. In these senses, the north Indian Indo-Islamic apprenticeship was distinct from any employment  of children outside the family.

Both the family firm and the master-apprenticeship had an association with skill. Exactly how production was organized is a relatively immaterial question in the unskilled crafts. It is an important one with the skilled crafts. Because here the producer did  not just produce, but imparted training, created skills, conserved skills, and restricted access to skills. The institutions doing these were of two types. The family-firm  internalized training and access to knowledge within  the family.  The apprenticeship internalized training and knowledge to  small, typically all-male, collectives that crystallized around master-disciple lineage. The family-firm and domestic labour were usual among Hindu artisans, the apprenticeship system among the Muslim artisans, most clearly visible in the towns of UP. There were exceptions to this Hindu-Muslim contrast, but they do not alter the rule. In both cases, there was a clear economic  motive involved, access to cheap labour, investment, insurance, or bequest, etc, and an element of authority or affection which softened and cemented the economic ties. In family, that authority was clearly parental or gendered. In apprenticeship, where did it come from? On  this point there is an important contrast between the standard historical paradigm of apprenticeship, medieval Europe, where the authority was exercised via formal corporate institutions, and India where no vestige of formal authority can be seen. Guild in the sense of a corporate distinct from public authority was  almost non-existent in India. What seems to have existed in its place was an ideology of competence, which I call the notion of 'ustadhood'.

How  were these systems affected by commercialization? Unlike in the case of the rural artisans, in the skilled crafts, the production unit did not dramatically weaken, obviously because there were  no alternative system to organize low-cost informal training. However,  distant markets, unknown  consumers,  risks of world market fluctuations, all demanded greater control by merchants upon direct producers. What we  see, therefore, is a superimposition of new exchange contracts on the old production unit. With family firms this took the shape of standard putting out, as in cotton handloom  weaving. With ustad-shagird system, it took the shape of a specific kind of putting-out which tied two dissimilar contracts - merchant-ustad, and ustad-worker – the whole system being known in north Indian towns as karkhanadari. The carpet factory was in effect such a system, though on a bigger scale.

2.2.5 Women

Women   rarely worked as full-time labourers in traditional small-scale industry. The incidence of full-time work may have increased over time. On the other hand, part-time work of various kinds declined. The scale of the second process was clearly larger than that of the first, leading to a net contraction in women's participation in industry (see the discussion in Section 1.3 above). Important examples of decline include cotton spinning, processing of yam and cloth allied to weaving, general helpers in rural tanning, etc.        The extent of women's participation in industrial work varied by craft, by type of  production unit, and by culture. Thus, in handloom weaving, women were extensively employed inside a family or, more rarely, outside it as labourers in the processing department of a factory. But, in metals of all kinds, women were rarely employed. Women were always employed on industry when the main system of work was a family firm. Women were  never employed in a master-apprenticeship system under  male ustads. But there are many instances of women of several families or women   of a village collectively working together. I shall call this type 'domestic-collective labour'. In some such cases, especially where delicate skills were involved, a parallel and practically invisible apprenticeship may have been at work. But we can only infer its presence. Finally, in north Indian urban Muslim craft tradition, women worked  only in specific crafts, and almost always in domestic-collective labour.

Pure domestic  labour has been mentioned already (Section 2.2.2). There is some  evidence on domestic-collective labour. One common  example of this came from weaving villages all over India, where women and girls set up warps and sized warp-threads in a common field, usually under the shed of a tree. This system is now rare, but not extinct yet. In rural tanning, women fetched water, and very rarely performed some leather finishing processes such as currying.²⁹ In both these examples women's  work declined. In one case, it declined due to the urbanization of tanning. In the other case it declined due to the use of new tools in winding, warping, and sizing, often as adjuncts to a handloom factory.

In the decorative crafts of UP, there were several significant examples of women   working  in domestic-collectives. The most important was, and remains, embroidery. The three elementary conditions that induced hiring of women household workers in a craft like embroidery were: (a) women knew the craft already in its noncommercial  uses, (b) the industry was depressed and needed to contract sales and reduce price, and (c) demand was seasonal. Seasonality of demand, and the ability to adjust price to demand, are frequently preconditions for the use of surplus domestic labour. The embroidery skills of the urban Muslim women in nearly all parts of India did not evolve out of an established manufacturing tradition. It evolved from the fact that, 'needlework [was] a somewhat rare accomplishment among Hindu women, [but] much  more a part of the Mussalman girl's education than the "three R's'"³⁰. But despite an  accumulation of skills, the range of their application was narrow and tradition-bound, such that the commercial use of these skills tended to suffer from poor quality.

A  decline of elite consumption probably induced increasing domestication of specific kinds of craftsmanship, embroidery being one example. Another example was weaving of the Paithani turban at a time when the traditional turban was beginning to become  old-fashioned. In the 1930s it shifted from the Momin male weaver's domain to that of his family.³¹ In mid twentieth century, several north Indian towns employed women  extensively in crafts that once seem to have engaged specialized male workers. The  former worked  in their homes, but necessarily on contract. Inevitably they compromised  on quality of work. The scale and sequence of this transition remains obscure.

If women lost livelihoods as part-time workers, did a new opportunity of full-time work open up for them in the factories? It did in fact, but on a very limited scale. The  handloom   factory, as mentioned  before, did employ  women   sometimes. Employment  of women  and children in the tanning factory appears extremely rare. But there do exist some exceptions. In Punjab factories, women ground sajji. or saline earth, and babul bark.³² Modern small-scale industries like the gins and presses were also a new  field for full-time work. On the other hand, within the large mills like cotton and jute factories, there was a retreat of women in favour of migrant male workers in the interwar period.

Part 3. Work and Workers after 1947

Broadly speaking, traditional small-scale industry in post-independence India saw a transition from social contracts of various kinds to market-determined contractual terms  in respect of employment  of labour. This basic process was  mediated by government  policy and structural conditions.

 3.1 Government  policy

After 1947, government  policy on industrial regulation influenced traditional small-scale industry significantly. The exact nature of this influence is a subject that cannot be  handled in this paper. But some broad  hypotheses can be advanced. Independent  India began with a strong sympathetic sentiment towards traditional small-scale industry - a legacy of Gandhian politics and ideology - but a very poor and very unreal comprehension of traditional small-scale industry - again a legacy of Gandhi's mythical ideas about the crafts. This had contradictory impact on traditional small-scale industry.

Some aspects of government policy favoured traditional small-scale industry. For example, textile policy reserved items for the handloom industry. In general a number of  manufactured products - traditional and non-traditional - were reserved for production in small-scale units.

But numerous  other aspects affected traditional small-scale industry adversely. For example, the main  instrument of government intervention was the cooperative society, on a mistaken notion that the merchant was an evil exploiter of craftsmen. This  policy strongly discouraged private trade and  enterprise and encouraged corruption and wastage of public money. The reservation policy discouraged the use of economies of scale in marketing where such economies existed. Small firms did not want to grow  into large firms, and did not want to become visible for fear of losing patronage. Being small, they could not establish brands. Being small, they could  cheat consumers  with impunity. As a  result, the quality of products and engineering skills in traditional small-scale industry have remained primitive by the standards of many developing countries that successfully produce and export products of traditional small-scale industry.

Another  way government  policy discriminated against traditional small-scale industry until the mid-1980s was  its highly insular macroeconomic  policy that discouraged exports and foreign investments. Traditional small-scale industry, being labour-intensive, had excellent export prospects. Some of this prospect is being realized only now, after economic policy reforms. Contact with foreign markets and technical skills were further limited by restrictive policy on foreign investments. Exports having been  shut off, the limited scale of the domestic market imposed constraints on traditional small-scale industry's employment growth.

Finally, labour laws, and the nature of their implementation, created widespread fear among small entrepreneurs about these laws and encouraged evasion by underreporting scale of firms. This factor has induced a bias against officially registering as a factory, and against increasing the scale of firms. There are signs that a very large number of factories in traditional small-scale industry went underground or were  dismantled in favour of smaller more  discrete type of firms after 1950 because of the pervasive fear of the bureaucrats and the trade unionists.

3.2 Structural conditions

If the effect of policy was mixed - on the whole negative - structural factors strongly favoured  traditional small-scale industry. Factor-endowments favoured labour-intensive manufactures. The supply of labour to industry has been sustained above  all by one of the highest rates of population growth in the post-war world. Together  with low levels of investment by  governments  in education and skill-formation, population growth  has ensured that the supply of relatively unskilled labour to industry remains  very high. Traditional and small-scale industry has benefited by the availability of this bottomless labour pool. Low purchasing power and in some cases traditional preferences favoured these industries too.

As traditional small-scale industries grew, they agglomerated further. In some cases, the growth of major craft towns of colonial India has been truly staggering in the last 50 years. Surat at the turn of the century probably employed about 5-6,000 weavers in silk and lace. Today, the direct descendant of weaving, the powerloom, provides employment  to about half a million. Moradabad brassware engaged 7-8,000 full-time workers in 1924. In the 1990s, an estimate places the town's metal workers at 150,000. Not more than a few thousands were found in the carpets in Mirzapur-Bhadohi area in the interwar period. 300,000 is the approximate figure in the 1990s. By the most recent report (a television documentary of 1997) Firozabad glass employs 400,000, many times the number it did 60 years ago. These cases capture a steadily increasing share of the informal sector in industrial wage-labour.

Along with staggering orders of employment growth, there  has  been modernization in infrastructure - marketing, credit, input-supplies. There has also been some  modernization in capability. A major product of metalware centres is generalized brass-casting; that of pottery centres industrial ceramics; that of glass-ware centres bottles rather than bangles; and of textile centres, rayon saris, polyester shirting, cotton cloth for exportable casual-wear, rather than old-style bordered cotton saris.

And  lastly, there has been limited mechanization in these industries. The most well-known  case of mechanization is the conversion of handlooms into powerloom weaving  factories, a transition that has involved several million industrial workers. Through these means - mechanization and modernization of infrastructure - the distance between traditional and modern small-scale industry has been bridged in many cases. Limited modernization and mechanization have happened without significant change in the average scale of firms. Technical economies of scale remain small and marketing economies of scale have not been utilized adequately either because of capital-shortage or perverse incentives created by the government.

How  have these changes affected work and workers?

3.3 The growth of wage labour and a labour market

The decline of the households has been continuing for a long time. It speeded up from the 1971 census. Why has the family declined as a firm? Five hypotheses can be offered, some of them can be seen at work in the colonial period evidence.

First, as supply of labour increases relative to capital, investment funds become less easily available. Loss of access to capital must be a reason why the average family finds it difficult to survive. Second, as trade expands and markets integrate, production concentrates in towns that have strong comparative advantage over its competitors. This agglomeration encourages migration, migrants tend to be males rather than families, further encouraging a break-up of the family. Third, the simple needs for division of labour and specialization - means to compete better in an integrated market - are served less efficiently inside a family. Fourth, surplus labour available for industrial employment now originates from a variety of backgrounds, many  did not  have prior experience in industry and thus had no prior ties with traditional employment  institutions. Fifth, with increasing division of labour, craftsmanship has become less important than before, and with it the family firm or the apprentice system has lost one important reason for their existence, the training function.

By  a simple extension, each one of these factors can be applied to infer a decline of master-apprenticeship systems of northern India. A further reason for its decline was migration of Muslim ustads to Pakistan at the time of the Partition.

Thus, wage-labour and the labour market expanded in post-1947 India. This is the lesson of employment statistics and the experience of major sectors. In this extension of wage-labour, there is continuity between the colonial and post-colonial periods, and similarity between agriculture and industry in post-colonial India. Field-work based studies suggest that labour hiring now happens almost exclusively in markets for casual labour.³³ Social ties matter to some extent, but are not critical variables.

Recent  descriptions of the work-site lead one to speculate the following changes. On  the one  hand, tiny factories, typically with an employment of 8-10 workers, seem to be the dominant organization in most traditional small-scale industries whereas  the family firm has steadily receded. Only in handloom weaving, the artisan family still has a considerable presence. But this is one of several anomalies created by the protection the industry has received from the government. Elsewhere, as in carpets, factories sometimes involve family labour, but are rarely pure family-labour units.

In average scale of unit, there is apparently a break. In the 1940s it was usual for a craft observer to talk about steady growth of the factory. But large craft factory, officially recognized as a factory, are today rather rare. There has been expansion of small back-yard workshops rather than 'factories' registered as such. Why did the trend stop or slow down? Why is wage labour being employed increasingly in tiny workshops? At least one reason why the larger factories disappeared from view is political. There is systematic under-reporting of factories in post-1947 India. Fear of the regulatory regime and  of the political party-backed trade unions is pervasive in small-scale industry. The petty government  officer can be both destructive and exploitative for the weaker capitalist.       

Coming  now  to the shop-floor, there has been proliferation of new tasks where training is not a critical resource. Old-style apprenticeship, and in turn, the sway of ustadhood  have clearly decayed. The intermediary in employment is typically a 'labour-contractor', and not a master in the old  sense of the term. The  term 'ustad' is remarkably  rooted in northern India. But it does not necessarily carry the 1920s meanings.  One likely reason why it does not is further division of labour that reduced the importance of the skilled supervisor. Another possible reason is de-skilling and casualisation across the new tasks. A yet third factor, and related to the other trends, has been great increase in employment of vulnerable labour like children.³⁴ Finally, labour is far more mobile now than before. There has been very large increase in migration of labourers from the poorest, agrarian, and over-populated states in India.³⁵ Before 1947, artisan migration involved mainly the artisan castes and communities. But the typical migrant labourer in northern India today is likely to be of agrarian origin, a labourer working for subsistence wage, a migrant from Bihar, and possibly a teenager. All this is a world away  from ustads and shagirds of the interwar period. In Western India, the picture would differ only in minor detail. In south India, such vast pools of surplus labour are scarcer. But short-distance migrations between semi-arid countryside and industrial towns have been  a typical channel of labour supply. Whereas  in many instances, the artisan-migrant in the interwar period ended up as an entrepreneur or a factor, the average migrant today is likely to end his career as a casual labourer.

3.4 Child labour after 1947: Out of family and apprenticeship into the market

Section 2.2.3 suggested that children were  employed  in primarily three contexts in pre-colonial India: family, apprenticeship, and some neighbourhood hiring between  families and probably between masters. The households, as we have seen, declined in the long run. The apprenticeship contract too does not exist quite in the same  form or same force any more. The figure of the ustad in the 1920s sense has disappeared from  most  carpet-weaving areas. Work  has standardized everywhere sufficiently to dispense with a privileged teacher of the craft. Employers are too well informed and in control to need a privileged intermediary 

Notes

___________________                                

1 The essay draws on my  Traditional Industry in the Economy of Colonial India. Cambridge, 1999. However, it also uses raw material on labour not used in the book.

 2  This shift in interest is illustrated by two major works on industrial labour set a quarter century apart, Morris D. Morris, The Emergence of an Industrial Labor Force in India, Berkeley and Los Angeles,  1965, and Dipesh Chakrabarty, Rethinking Working-Class History. Bengal 1890-1940, Princeton, 1989.

 3 Roy, Traditional small-scale industry, and Tirthankar Roy, 'De-industrialisation', Economic  and Political Weekly, forthcoming.

 4  S. Sivasubramonian, 'Revised Estimates of the National Income of India, 1900-1901 to 1946-47', IESHR. 34(2), April-June 1997.

 5 United Provinces or Uttar Pradesh according to context.

 6 On Padmasali migration into Sholapur and the history of handloom factories in this town, see Douglas Haynes  and Tirthankar Roy, 'Conceiving Mobility: Weavers' Migration in Precolonial and Colonial India', Indian Economic and Social History Review, 33 (1), January-March 1999, and Douglas Haynes, 'The Logic of the Artisan Firm in a Capitalist Economy: Handloom  Weavers and Technological Change  in Western  India, 1880-1947', in Burton Stein and Sanjay Subrahmanyam   (eds.), Institutions and Economic Change in South Asia. Delhi, 1996.

  7 Royal  Commission  on  Labour in India, XI (Supplementary), London,  1931. Although the survey was claimed to be a census, for many provinces it does not appear to be so.

 8 K. Quddus Pal, 'The Punjab Tanning Industry', unpublished report prepared for the Punjab Government, Lahore, 1940, pp. 70-72.

 9 H.G. Walton, A Monograph  on Tanning  and Working in Leather in the United Provinces of Agra and Oudh. Allahabad, 1903, p. 28.

 10 Indian Industrial Commission, Minutes of Evidence, 5 vols. Calcutta, 1916-8, Vol. III, evidence of G.A. Chambers of The Chrome Tanning Co., Pallavaram, p. 323.

 11 Kunwar  Jagdish Prasad, Monograph on Carpet Making in the United Provinces. Allahabad, 1907, p. 6. Jhansi is an example.

 12 Royal Commission on Labour, p. 90.

 13 Example, a speculatory episode in the 1890s in Amritsar, C. Latimer, Monograph on Carpet Making in the Punjab. 1905-06. Lahore, 1907, pp. 16-7.

14  Ahmad  Mukhtar, Report on Labour  Conditions in Carpet Weaving. Labour Investigation Committee. Delhi, 1947, p. 16.

 15  For one among  several descricpitive accounts on these nuances, see Prasad, Monograph  on Carpet Making, pp. 12-4.

 16 Census.  1961, Uttar Pradesh, Woollen Carpet  and Blanket Industry of Uttar Pradesh. XV(VII A). Allahabad, 1964, reference to Shahjahanpur, p. 19.

 17 Mukhtar, Report on Labour Conditions, p. 16.

 18 For a sample text, see Indian Carpets. Bombay, undated. section titled 'Kashmir'.

 19 Report of the Royal Commission on Labour in India, London, 1931, p. 97.

 20 Royal Commission on Labour, evidence of masters Muhammad Ramzan and Rajbai, pp. 104-5. Some aspects of the arrangement we will be dealing with below, are also described in a recent book on Pakistan, where the typical interwar Indian contracts seems to have remained more unaltered than in India itself. See the excellent report, S.M. Shah, Hand-knotted Carpet Industry of Pakistan. Peshawar, 1980, pp. 8-40.

 21 For example, Prasad writes of the prewar decades in northern India: '..the great demand  for labour makes the instruction less thorough and, very often, raw workmen are turned out', who obviously competed with their masters, Monograph on Carpet Making, p. 9.

 22. A.C. Chatterjee, Notes on the Industries of the United Provinces, Allahabad, 1908, p. 59.

 23 Report of the Royal Commission, pp. 97-8, and Royal Commission on Labour, p. 90.

 24. The  quotation from India, Draft Report on the Himroo  Weaving  Craft of Aurangabad (Maharashtra), All India Handloom Board, New Delhi, p. 12.

 25. N.G. Ranga, The Economics of Handloom Industry. Taraporevala, Bombay, 1930, p.113.

 26. N.G. Ranga, Economics of Handloom. p. 110.

 27, Ranga, Economics of Handloom, pp. 81-2.

 28. Ranga, Economics of Handloom, p. 136.

 29 Report of the Madras Exhibition, 1855, cited in A.C. Chatterton, A Monograph on Tanning and Working on Leather in the Madras Presidency. Madras, 1904, p. 2.

 30. W.S. Hadaway, Monograph  on Tinsel and Wire in the Madras Presidency, Madras, 1909, p. 10. Gift-exchanges between families during marriages involved, with the southern Muslims, home-crafted cloth bags filled with nuts, spices, or money.

 31. Indian Tariff Board, Written Evidence recorded during enquiry on the Grant of Protection to the Sericultural Industry. Vol. I, Calcutta, 1935, p.304.

 32 R.L. Anand and Ram  Lal, Tanning Industry in the Punjab. Board of Economic Inquiry Publication No. 61. Lahore, 1931.p. 26.

 33 Jan Breman, Footloose Labour, Cambridge, 1996.

 34 Studies on child labour in these clusters suggest some patterns of change in the labour market. The child worker  in northern India today is usually employed in household workshops and  not factories, for example, in carpets they are employed in rural households to perform agricultural and industrial labour; he/she does not originate in the city but is an immigrant from distant and impoverished village; is nowhere mentioned  as a 'trainee', 'student', or 'apprentice', but universally as a cheap labour. See, for example, B.N. Juyal, Child Labour and Exploitation in the Carpet Industry: A Mirzapur-Bhadohi  case study (New Delhi, 1987). In a survey of Khurja pottery, Burra was  told that the child-turned-adult earns less than an adult entrant into labour, Neera Burra, Born to Work, (Delhi, 1995), p. 131. If generalizable, this is the opposite of what any  apprenticeship employment of children must imply, and shows that the motive for employing  children is the vulnerability of the segment where they come from, rather than potential incomes from training.

 35 For an insightful and detailed descriptive survey of supply of labour and conditions of work in some major artisan clusters in north India, see Burra, Born to Work. On the labour market and employment practices in the informal industry of Surat town, see Breman, Footloose Labour.

 36 See on Pakistan, Shah, Hand-knotted Carpets, pp. 33-4, 41. In the rare instances in southern India where children were formerly employed, and which could not adapt, the industry declined. At Walajahpet in Arcot, carpet factories were apparently run under family cooperations, and children employed as domestic workers. The Factory Act ended  employment  of children and by breaking up the family work-unit, 'played an important role in hastening the decline of the industry'. It needs to be noted that the children moved  into more unskilled occupations, and did not necessarily leave the labour market to acquire skills, as a result Census of India. 1961, Druggets and Carpets of Walajahpet. DC (VII-A-VII). Madras, 1965, p. 20.

37 India, Indian Carpets in World Market, All India Handicrafts Board. New Delhi, undated,