In India, investigations into patterns of industrialisation and the formation of industrial labour began during the colonial period, soon after the inception of modern industry in the mid-nineteenth century. After Independence in 1947, the development of a ‘working class’ became the primary focus of enquiry into conditions of industrial labour. Scholarship on industrialization and industrial labour flourished, especially from the 1920s to the 1970s, dominated by concerns of modernisation, national politics, trade unionism and working class consciousness. In the earlier period, until the 1930s, great deal of attention was paid to various, especially negative, aspects of women's and children's work. Issues of gender relations—in the workplace and poor urban neighbourhoods—often overshadowed questions of protest and resistance. The dominance of official, philanthropic and reformist concerns influenced a preoccupation with what one might perhaps term the ‘social’ world of labour. After the 1940s, the focus shifted. The need of the hour was a greater ‘political’ understanding of the growing relationship between unionized labour and the emerging post-colonial state, political parties and local political institutions. In this stage of Indian labour historiography, issues of gender were considered entirely superfluous. The common (often stated) assumption was (and still is) that the low proportion of women workers in the modern industrial sector rendered them irrelevant to stories of ‘class’. The few women who did work in industry, mines or plantations were subsumed within the general definition of class. A more knotty question—that of the role of women in male industrial workers’ households who worked in agriculture or petty commodity production or trade or retail—was not raised at all. Were men workers members of a working class in their individual capacity or as heads of ‘working class families’? How do we characterise the work of wives, mothers, daughters and sisters of male industrial workers? These questions were not given any consideration in prolix debates about the growth and development of the industrial ‘working class’.

Women's work and their role in traditional or modern manufacturing have also had relatively little attention from practitioners of women's studies. In development economics and sociological literature, poor women are almost always of rural and peasant groups. Very recently, however, in the last two decades, there is a new interest in urban poor women and industrial work. Much of the research remains sparse and scattered and unconnected to long-term historical trends. This latter problem is likely to persist as long as gender issues continue to be neglected in historical investigations of labour, a neglect that is compounded by scholars of women's history. The historian in search of women's ‘voices’ has been limited, necessarily, to middle class literate women for whom writing and publishing had become a radical activism, a way to breach the public domain. The poor and working class women, invariably unlettered, could not leave their own authored traces in historical records. As a result, women's historians, persuaded by arguments of ‘lack of evidence’, have neglected issues of work and workers. Poor and workingwomen have fallen between the two stools of labour and women's history—further silenced and divested of historical agency.

And yet, as already mentioned, from the late nineteenth century and until the 1930s, a variety of elite discourses focused obsessively on poor urban women, their work, their visibility, their sexual and marital behaviour, their childbearing and mothering practices. These ‘subaltern’ women were a constant presence in elite discourses, which provide a rich variety of information about their lives and how they were perceived. The ‘problem’ is not ‘lack’ of evidence, but rather, as in the case of many other historical investigations, that of the nature and the often unexpected provenance of the evidence. There is sufficient evidence, certainly, to suggest that women were a critical segment of the industrial labour force at its inception. Janet Harvey Kelman, who wrote one of the earliest and most remarkable accounts of Indian labour, associated women with the ‘tragedy’ that surrounded the ‘first efforts to introduce modern mill industry into India’. In 1818, she writes, a group of ‘Lancashire girls’ were brought to Bengal to ‘introduce factory methods of work’ at the old Bowreah mill in Hooghly. ‘Heavy, white-washed tombs in the rank grass of a small cemetery near the mill compound still keep in remembrance the swift death to which many of them fell victims’.Footnote ¹ This rather surprising piece of information remains little known among India's or Bengal's historians, even though it found a prominent place in the best known and most insightful Bengali novel on the introduction of modern industry in Bengal. In Jagaddal, Samaresh Basu attributed the importation of British women weavers to MacAllister, the manager of the Fort Gloster Mills. Apparently, MacAllister persuaded the British and Indian governments to release into his custody a batch of women prisoners from an ‘island prison in the Atlantic, south of the northern Highlands’. In the novel, MacAllister recounts the story thus.

Like Kelman, Basu too refers to some of these women marrying Indian workers and settling down. While, according to Kelman, these women were victims of an epidemic, in Jagaddal, MacAllister recounts the misfortune of the more adventurous of these women who fled Fort Gloster only to end up in the flourishing brothels of Calcutta.Footnote ³

These mysterious British women were not the only women in Indian mills. From all accounts, women and children were quite a substantial presence in the cotton textile industry in western India and the jute industry of Bengal in the late nineteenth century. Both these centres of industry grew apace and by the end of the century attracted the attention of British philanthropists, reformers, the Manchester and Dundee lobbies. A variety of interests in Britain agreed on the need to regulate factory conditions in India. Mary Carpenter, celebrated educationist and prison reforms campaigner, on her return from a tour of India, argued for urgent legislative intervention on behalf of women and children workers, at that stage more than a fourth of the total textile factory workforce. The first Factories Act came in 1881, at the teeth of the mill-owners’ opposition. From this period and until the outbreak of the First World War, state and public interest in labour conditions remained focused on women and children. Long and laboured controversies among employers, reformers (both British and Indian), provincial governments, Governments of India and Britain produced a rich literature on women's working conditions and on social and gender division of labour in that period. Alongside, industrial employers, concerned about labour supply in a period of rapid expansion, generated a veritable archive on labour. Planters—of Assam tea-gardens and the overseas colonies—were heavily dependent on the state for labour recruitment and control. And they were among the few ‘modern’ sector employers who had special need for women—for both reproductive and productive purposes. Their practices and policies were the fulcrum around which questions of women's work were discussed and elaborated.

In all these areas, the concern was with the woman in her capacity as a worker. At this stage, legislation sought to regulate the condition of women actually engaged in factory work—the hours of work, periods of rest, the prohibition of night work and/or handling of machinery. These laws were for workers who were ‘special’ because they were women. And they were special in three ways: first, they also had to perform their reproductive roles as wives and mothers; second, their physical weakness limited the kind of work suitable for them; and third, they were unable to uphold their own interests and thus needed the ‘protection’ of the state. Even those who opposed protective legislation usually agreed that women workers had ‘special’ characteristics. For instance, jute mill-owners agreed that women should not work long hours but argued that due to the multiple shift system they did not do so anyway and therefore legislation was not necessary. They agreed that women workers had an important familial role to play, but pointed out that their earnings were crucial to household survival. Men workers, however, were considered not only as not having the same ‘needs’ as women but also as free agents able to negotiate mutually beneficial working conditions with employers.

In this early period, those directly concerned with controversies over labour—employers and the state—were the most prolific recorders of labour issues. A few concerned Indian and British reformers wrote a few accounts, the most remarkable being those by Dwarkanath Ganguly and Ramkumar Vidyaratna on labour in tea plantations of Assam. It is in the latter that we find glimpses of the kind of concern over women's labour that was to hold centre-stage after the First World War. The exigencies of women's sexual and reproductive roles were in conflict, these writings suggest, with their recruitment as wage labour. They dwelt on two major consequences of women's migration: first, the depredation of the family from which the women were recruited and second, the unsuitability of plantation employment for women given, especially, their vulnerability to sexual violence by European plantation bosses.

Dwarkanath Ganguly and Ramkumar Vidyaratna are recognized as pioneering labour reformers. Certainly, they were able to articulate concerns of class, gender and race in a complex inter-relationship that was to become paradigmatic of a ‘nationalist’ discourse on labour. Marxist scholars too have drawn on these two writers for a genealogy of the ‘making’ of the Indian working class.Footnote ⁴ But, some of the other key personalities engaged in discussions about women workers in this early period hint at wider inter-linkages of gender concerns, which have yet to be explored. Mary Carpenter, the pivotal figure behind the first Factories Act (1881) was also committed to women's education in India. Her chief allies were the Brahmos of Calcutta, who were in the eye of many a storm over the ‘women's question’. In 1866, Keshub Sen's Nababidhan group welcomed Mary Carpenter to Calcutta and with her and Anne Ackroyd's help, established a Normal School in 1872. Ackroyd later broke with Keshub Sen, and with the help of another group of Brahmos established the Hindu Mahila Vidyalaya. By 1878 this school merged with the older Bethune School to become Bethune College (affiliated to Calcutta University) from which the first two women graduates of the British Empire were to emerge in 1883. Mary Carpenter worked closely with Sasipada Banerjee, acknowledged as the first Bengali Bhadralok to initiate schemes for labour welfare.Footnote ⁵ Banerjee was also active in campaigns for widow remarriage and established the first shelter ‘home’ for widows called the Hindu Bidhabashrama, from which Haimavati Sen later re-established herself as one of the earliest practicing women doctors in Bengal.Footnote ⁶

Even as, in the late nineteenth century, poor women became a focus of public discussion regarding their hours of work and the suitability of factory conditions, the bhadralok debated the suitability of women's education, the possibility of the bhadramahila entering the public domain of the new professions and services and, most importantly, women's familial and marital roles in the face of sweeping economic changes. Thus, the Age of Consent debates (over fixing a minimum age of marriage for girls) raged alongside discussions about amending the Factories Act (further restricting women's work) in 1890–91. The two debates were synchronous but separate; in neither debate was there any expressed awareness of the other. Yet, some common assumptions of gender underpinned both, each bearing in greater or lesser degree the stamp of the ‘nationalist construction of womanhood’, which sought to draw all ‘Indian’ women within its ambit. These equivocations, moreover, implicated gender deeply within processes of class formation: to the establishment of a greater distance between the bhadramahila (middle-class high-status women) and the magi (low caste, poor and not respectable women).Footnote ⁷

By the turn of the twentieth century, the nationalist construction of womanhood, according to Partha Chatterjee, provided a radical ‘resolution’ to the women's question, empowering them within their traditional roles as wife and mother.Footnote ⁸ This so-called resolution, however, found it increasingly difficult to reconcile gender inequities within marriage. Unable to explain away the comprehensive vulnerability—of life and freedom—of the wife within orthodox regimes of marriage, Tanika Sarkar argues, the focus of nationalist imagination shifted from the more threatening conjugal relationship to the safer and more emotive mother-son relationship. Followed a celebration of motherhood, which found political and aesthetic fulfillment in the icon of the Motherland.Footnote ⁹ Poor and working-class women were drawn into these new concerns. Thus, in the 1920s, there emerged different strands of writings about working women. While the state and employers continued to address women's working conditions, a group of professionals—academics and doctors, especially—also focused on these women as a social ‘problem’. Kelman's book (1923) was followed by G.M. Broughton's (an Inspectress of Factories) (1924), several by Margaret Read (1927, 1931 and 1934), and one by C.M. Matheson (1930).Footnote ¹⁰ These were women writers focusing on the concerns of women workers. R.K. Das wrote a long and well-researched essay on women workers in 1931.Footnote ¹¹ Other general studies of labour contained long sections on women—P.S. Lokanathan (1929), S.G. Panandikar (1933), B. Shiva Rao (1939) and Radhakamal Mukherjee (1947).Footnote ¹² Three European women doctors—Drs. Francis Barnes, Dagmar Curjel and Margaret Balfour—did extensive research on fertility and mortality among working and working class women and published several tracts between 1923 and 1935.

These writings reflected a shift in the nature of public concern about and state regulation of factory labour. It was no longer accepted that the state should protect the ‘weaker’ sections of the workforce while adult male labour dealt with the ‘market’ without any external interference. Faith in the paramountcy of free trade over the labour market waned and the newly instituted International Labour Organisation initiated long-term efforts to create a minimum standard for labour conditions and undertake its monitoring across the globe. There were two clear differences from the earlier period. First, there was a move to equalise women's and men's working conditions by adopting similar checks on hours of work and welfare measures that were, at least ostensibly, gender neutral. Second, a set of reforms continued to address women in specific terms but in their reproductive capacity. Factory women were increasingly regarded not as workers with particular problems calling for separate remedies but as special kinds of mothers and wives—ones who also worked. Alongside debates about whether they should work or not were questions about the adverse impact of work on housewifery, child bearing and rearing. These were directed towards isolating, and if possible remedying, the specific problems of working wives and mothers. This shift was in consonance not only with the nationalist focus on motherhood but also with a growing public discourse about the ‘family’ and the alarming rates of maternal and infant mortality in India. Thus, various professional and social discussions converged in the 1920s and 30s on motherhood—both as a problem and as a solution. These concerns included elite and middle-class women, leading to debates about separate medical and hospital facilities for women and the urgent need for women doctors to attend women patients, especially during childbirth. The latter concern justified middle-class Indian women training as doctors, while it provided a major professional opportunity for British and European women doctors who faced discrimination in their own countries.Footnote ¹³ These professionals—mostly still men, of course—began to devise standards of childbirth and childcare based on their own middle class experiences. Working-class women were included in these concerns to the devaluation of their work roles. This helped industrial employers, who used working class motherhood as both a practical and an ideological instrument to reduce the (female) workforce during the depression of the 1930s. In the Bengal jute industry, for instance, the proportion of women declined by about 2 per cent between 1930 and 1940. A steeper decline was set in motion in the coal industry when women were banned from underground work in 1928. In plantations, they retained their share (about half to start with) much longer because of the perceived ‘feminine’ skill involved in low-paid plucking jobs. But elsewhere in India too (Bombay and Ahmedabad) women suffered job losses.

Similar attacks on women's jobs became more pronounced in a new round of ‘rationalisation’ in the 1950s. By 1971 women, who had been almost a fourth of the jute workforce, were decimated to a bare 2 per cent of the workforce. The decline of female labour in industries coincided with the improvement of wages and working conditions. In the jute industry, for instance, the real wages increased most rapidly from the mid-1960s to be almost doubled by the end of the 1980s.Footnote ¹⁴

From the 1940s, academic writing on labour began to concentrate on the twin aspects of modernisation-industrialisation and on progressive class consciousness as evinced through rapid unionisation of labour. The earlier prominence given to women workers as a special category eroded with the decline in their proportion within the workforce. The greatly diminished visibility of women workers bolstered a growing ideological commitment to the notion of a cohesive and solidaristic (male) working class. Radhakamal Mukherjee's pioneering book, The Indian Working Class (1947) included a chapter on ‘Women and Child Labour’, discussion of the adverse effect of exclusion of women from underground work in mines, the implication of ‘family budgets’ and the relevance of a national minimum wage for women and children.Footnote ¹⁵ Subsequently, however, the history of the ‘working class’ became increasingly equated with strikes and unionisation. The little of academic and professional writing which still chose to focus on women, for instance the studies conducted by M.N. Rao and H.C. Ganguly (1950–51), retained the earlier emphasis on sexuality and motherhood.Footnote ¹⁶

Women workers, thus, were being marginalised on multiple fronts: in actual exclusion from rapidly improving organised sector employment, from the political space unionised labour aspired to and eventually occupied and even from more general public and ‘social’ concerns of the earlier decades. Overall, we see a diminution of women's identity as workers. This paper will address a few facets of this process. First, the ‘organisation’ of industry and labour is directly related to the decline in women's employment. On the one hand, employers became less interested in women as the progress of welfare legislation reduced their cost advantage. On the other, the nature of the legislation brought women's reproductive roles into prominence and created a climate of public opinion against their employment in industry. Employers could then target women for retrenchment when ‘rationalising’ the workforce. Second, as the process of organisation created a relatively more secure enclave of employment approaching a ‘family wage’, working-class family strategy changed. In working class households where many members previously worked in industry, the trend turned towards a single male breadwinner. But this was a tiny enclave and the proportion of families with a secure foothold in this sector was on the decline. In the less advantaged agricultural and/or the ‘unorganised’ sectors, the participation of women and children remained steady and often increased during downturns in the economy when the ‘security’ of the ‘organised’ sector failed. Such adjustments within the family economy were made possible by the continuing authority of male heads of households who were able to command the deployment of women's labour in consonance with ‘family needs’. The male workers’ control over ‘their’ women's productive/reproductive activities came to be reflected in trade union policies. The unions were also participants in the emerging public and state emphasis on women's reproductive roles. The significance of such a convergence lay in the development of national-level federated trade unions fronted by political parties, which contributed to the process of ‘organisation’ of registered industries. Labour, thus ‘organised’, earned not an inconsiderable voice in labour conditions. This voice, when just emerging in the 1930s, accepted female retrenchment as a necessary strategy in the face of spiraling male employment. From the 1950s, the unions played an active role in eliminating women and hindering women's recruitment. Such policies were only in part due to the prejudices of the middle class leadership, reflecting, much more importantly, the adult male workers’ status aspirations and desire for maintaining family authority.

The paper is arranged in four sections. The first section discusses the increasing concern with women's familial roles from the 1920s, the second examines women's role in labour protests to show how their roles were represented and how their modes of protest were sidelined by the growth of unionisation. The long-term consequences of these two phenomena are the focus of the third, which draws on recent research on contemporary trends in women's employment in manufacturing. The final section explores received wisdom about women's preference for dependent domesticity and how such preference may be produced within the ideological imperatives of the family and in certain historical contexts.

Migration, Morality and Motherhood

In 1888, Ramkumar Vidyaratna's Kulikahini [Sketches from Cooly Life] was published. It was a fictional recreation of his experiences gathered while touring Assam tea plantations, and he sought to highlight their highly exploitative labour regime in the tea plantations. Questions of gender and labour were, however, inextricably linked in the delineation of the central character, Adarmani, a woman tea-garden worker. The story begins with attempts by a garden manager to recruit peasant women from north India. The manager instructed his chief agent, an arkathi, to employ female recruiters.

The author then goes on to depict how two women recruiters, Ramuna and Jhamuna, successfully recruited Adarmani, the wife of a poor peasant, and her two daughters. The recruiters found Adar in a particularly vulnerable state. Her husband had gone away with their son in search of work; they needed cash to pay the accumulating rent and interest to the moneylender. The recruiters chose this propitious moment to approach Adar. They adopted the strategy recommended by the manager. First, they held out attractions of rich clothes, ornaments and a good life. This was then contrasted with the misery and drudgery of Adar's present existence. One of the women said persuasively, ‘Your husband thinks that a little rice and a coarse cloth is enough to keep you tied to him for life. And kicks and punches are all the ornaments you need.’ She also explicitly construed Adar's decision to migrate to the tea plantation as a revocation of male authority, ‘Would your husband have let you go if he knew?’Footnote ¹⁸ The recruiters direct an ideological assault on family and marriage as a means of persuading Adar to accompany them, thereby emphasizing that she is exercising a choice between family/marriage and wage work in plantations. Their arguments were startling in their radical challenge to male authority. It could be argued, however, that the author put up these drastic arguments to help the reader anticipate their fallacy. These dialogues are crafted with great subtlety conveying, at the same time, convictions of truth and falsehood, choice and manipulation, decision and deception. There could be a third layer of argument: that for Adar to be persuaded by these arguments, they must have had some credibility, both to the author and to the anticipated community of readers. No doubt Ramkumar Vidyaratna wished to make the point that ‘a little rice and a cloth’ accompanied by the husband's kicks and punches were indeed a better option for Adar than her inevitable fate in the teagardens. In succumbing to the recruiters’ wiles, Adar was not only revoking male authority but inviting its terrible consequences: hard work, poor pay and sexual exploitation in the gardens. To highlight the terrible conditions at the plantations, he drew quite a harsh picture of women's life in the village home. These conditions explained not only what led women to repudiate their home and husbands to go to distant Assam but also underlined the grim reality of the punishing work regime at the plantations. In the village there was, of course, economic hardship, physical strain, battering and above all male control and subjugation. He repeated these arguments when Ramuna persuaded a wavering Adarmani to take the final step towards Assam. The author's imaginative reconstruction of Adar's hardships in the village and at home as expressed through Ramuna is laden with a heavy and obvious irony. His foreknowledge that conditions in Assam were in every aspect worse than Adar's present difficulties were meant to be read into Ramuna's diatribe against ‘men’ (husbands): ‘Men in general are terrible . . . . They die of envy when they hear that you will be able to earn Rs. 5 or 6, that you will able to live like a queen’.Footnote ¹⁹ There is a tension in the many long passages in this vein: on the one hand, the author quite evidently presumed that a kind of a declaration of war against the husband would, credibly, appeal to Adar; on the other, he regretted Adar's susceptibility.

Adar's story underlines and exemplifies women's transition from traditional (male) familial authority to the new sites of colonial production where women were vulnerable to heightened labour and sexual exploitation. Harilal Bandopadhyay, in a play released by the popular press in Calcutta, Arkati Natak (1901 CE; 1304 BS) comprehensively indicted the arkati system, drawing an imaginary scenario of a batch of deceived immigrants teaming up with a Naga tribal leader to resist recruiters. The victims were three women, two brought to Assam on false promises and one decoyed through a mock marriage ceremony. Referring to one of these women, the recruiter, an Englishman named Alpin, commented, “All the money gone down the drain! Jagadamba's womb would have produced enough kids to fill the garden! These females are very prolific!” In a final speech, the chief protagonist, Gangaram, under whose initiative the immigrants were rescued, dwelled on the tragic consequences of women being deceived into the gardens, ‘exploited’ both sexually and through hard labour.Footnote ²⁰

Over time, the world of capitalist wage labour, especially the poor urban neighbourhoods, appeared like a cauldron of vice, crime and disease while by contrast the rural world of family-based peasant production gained idyllic characteristics in increasingly nostalgic re-telling. The sharpening contrast drawn between rural/urban and peasant/worker had pronounced gender overtones and poor urban woman became its concentrated focus.

From the 1920s, the condition of women in the factories and the mill towns began to provoke much discussion. Kelman attributed this to the Washington Convention (1919). Public interest, she wrote, about working women's condition in India became

The Washington Convention provided for compulsory maternity leave and benefits for women workers. The Government of India pleaded that such a law was unworkable in India and was requested by ILO to furnish information about the condition of women workers. This prompted special and directed drives to generate knowledge about women workers in industries. Three women, G.M. Broughton, Dr. Francis Barnes and Dr. Dagmar Curjel were designated to enquire into ‘conditions before and after childbirth’ of women workers in the North-West, Bombay cotton textile industry and the Bengal industries. Ironically, these initiatives, prompted by the well-meaning effort to bring to Indian workingwomen maternity benefits, nevertheless provided some of the most abiding and powerful negative images of Indian women workers. Broughton's report has not been traced. Doubtless some of her findings are recorded in her book mentioned earlier. Barnes in a small report gave a detailed description of the unhygienic and overcrowded conditions in which poor urban women gave birth. This and her finding that ninety per cent working class children were fed opium became legend. Dagmar Curjel left a report of several hundred pages including her filled-in questionnaires. Her papers have become a major source of information about Bengal's workingwomen.

Her report was the first comprehensive statement, most oft-quoted, about the questionable and ‘non-family’ character of jute mill women.

Curjel's report also contained formidable evidence on appalling working conditions and a clear recommendation for a blanket application of maternity benefit legislation, but what was picked up and widely quoted were the suggestions of female promiscuity and a breakdown of the family. It was from her report that the ‘non-family’ character of urban Indian labour was discovered and the image of the women ‘who were not the wives’ of the men with whom they lived was passed on to posterity.

These were not far from Ramkumar Vidyaratna's concerns about peasant women being lured into immoral conditions of waged work. In the urban situation, however, there was a further ‘objective’ ballast to such characterisations: the low sex ratio in the mills and in the urban population. The fact that fewer women migrated to the urban areas became both the cause and the consequence of poor urban women's degradation. Janet Kelman gave perhaps the least morally laden descriptions of the phenomenon:

By the 1940s, however, such arguments, by force of repetition, had acquired more power. Radhakamal Mukherjee attributed “prostitution and spread of venereal diseases” to “a serious disparity between the proportions of sexes”.Footnote ²⁴

It was believed that social sanctions against women's mobility and visibility were effective deterrents to family migration. Men, who migrated to the city to supplement household income, would not risk their status, which influenced their foothold in village society, by bringing their wives to the city. Dagmar Engels, in her study of women in Bengal, referred to these arguments.

Abdul Hakim told the Royal Commission in 1930, ‘People of my district do not bring family to industrial areas . . . if I brought my family people would laugh at me.’ Over decades, this statement has been widely quoted to ‘prove’ that respectable women did not migrate to the cities and therefore, by extension, that those who did migrate were prone to promiscuity. To quote once again from Radhakamal Mukherjee's masterly descriptions:

Such arguments leave unexplained why better-paid workers (including those from Hakim's district), especially the sardars, often brought their wives with them.

The oral evidence collected by various labour commissions provides a different picture. It appears that women who suffered impoverishment through inadequacy or loss of male earnings by desertion or barrenness or widowhood opted for migration since economic opportunities in the village were reduced. The same applied to women who wished to physically escape oppressive fathers and husbands. Such women figured prominently in the female workforce. Except one, all the women interviewed by the commission of 1891 were widows who held that all their colleagues were widows too and that widowhood alone drove Bengali women into millwork.Footnote ²⁷ Even in the 1930s, the situation had not changed greatly.Footnote ²⁸

Many of these women came to the city alone and set up house with one of the many ‘single’ men who worked in the mills; and some came to the mill towns ‘with men who were not their husbands’. Since many of these women migrated outside the family context, they were construed as aberrant. They became objects of elite derision and came to personify the breakdown of morality in the city's overcrowded tenements. The working class neighbourhoods became associated with the collapse of caste and gender hierarchies. The Bengali women, especially, were often described as prostitutes. Doubtless, some migrants sought refuge in the impersonality of urban life after transgressing kin or caste rules in the village. In elite descriptions, these men and women acquired a particularly prominent profile. G.M. Broughton, the Lady Inspector of Factories, explained

S.G. Panandikar also elaborated the notion that the city was a hospitable refuge for those fleeing traditional social sanctions in the villages.

Contemporary official documents and independent researchers tended to confirm stereotypes based on a sharp divide between the rural and the urban. It was argued that it were the ‘single women’ and some male migrants in ‘entanglements with local women’ who settled most readily in the mill towns. These women could not return to their villages, according to Dagmar Curjel.Footnote ³¹ These women were more completely ‘proletarianised’ and their ‘rural link’ irrevocably broken. There were also workers from the bottom of the social scale—the Muchis and the Chamars—who took more readily to permanent settlements around the mills. The women of these castes were more readily de-linked from their villages: ‘[a] large proportion of their women came to stay’. R.N. Gilchrist, the Labour Officer of the Government of Bengal, observed that second generation immigrants working in the jute mills were born in the mill lines or in the neighbourhoods. They rarely returned to their native village. They were the ‘illegitimate children of jute men and women workers . . . and a large number of these women and children, who are born of the [temporary] unions, never leave the areas where they work’.Footnote ³²

Samaresh Basu stands virtually alone in challenging these sharp dichotomies between rural and urban, agricultural and industrial society. He places his fictional Jagaddal in the early period of the jute industry when entrepreneurs were moving up the right bank of the Hooghly, swallowing up rural communities to set up mills. The mills were set in the heart of rural agricultural communities and forced to maintain a delicate balance between rules and customs of village society and an emerging industrial culture. Basu depicts the changing village society in its many complexities, without idealizing a rural past or assuming an ordered caste and gender hierarchy. The most striking female character, Bimli, was a weaver who operated four looms at once. “She used a razor-sharp knife to cut the fluff out of the cloth, drank her bottles of liquor in tandem with the men, even raised her hand to the men at times.” She is drawn sympathetically, one who pursued “her own free and vagabond existence” without offence to anyone else or wishing anybody ill.Footnote ³³ Elsewhere Basu talks of the tradition of ‘Bharar Meye’, the women who were brought out in country boats and ‘sold’ into marriages and the difficulty of establishing their caste or even religion.Footnote ³⁴ Rural society, he suggests, was not frozen into caste structures and women not completely immobilized in marriage. His depictions are in conformity with recent arguments that from the nineteenth century, it was Brahmanic norms of marriage and gender relations that were being represented as generally “Indian” (or at least, “Hindu”). The customary rights and marriage practices of low castes and the poor were being eroded by colonial law and elite Indian efforts to universalize Brahminical norms and consolidate male control within the family.Footnote ³⁵ Basu's writings suggest far more continuities in the social and marital practices of the poor than officials and researchers and even social scientists appreciated.

In the 1920s and 30s, argues Radha Kumar, the concept of the woman as mother gained ascendance. And the focus was on ‘the working class woman as mother of the second generation proletarian’. She argues, in the context of the Bombay textile industry, ‘the family assumed vital importance for administrators, planners and employers’. This importance is evident in periodic family budget surveys and investigations into maternal and infant welfare. The preoccupation with the ‘non-family’ character of jute mill labour would then seem the flip side of the ‘search for a family’ in Bombay. Kumar argues that the 1920s saw parallel efforts at ‘rationalising’ the labour process as well as the process of reproduction, i.e., the family.Footnote ³⁶ R.S. Chandavarkar is sceptical of the possibility of either mill owners or the colonial state being concerned about the ‘family’ or even about ‘motherhood’.Footnote ³⁷ It is highly plausible, however, that the discourse on ‘motherhood’ came at a convenient juncture for mill-owners who were seeking ways of ‘rationalising’ (read downsizing) the workforce. The concern with motherhood not only gave them a morally justifiable argument but also allowed them to operate within an ideological space that male workers and trade union leaders shared.

In Bengal, mill owners invoked moralistic perceptions about workingwomen to postpone maternity benefit legislation for two decades. In the late 1920s, mills in Calcutta started their own voluntary schemes, partly as a means of staving off stiffer legal provision and partly in order to ensure closer control and supervision of women workers. When in the wake of the Royal Commission (1930–31) maternity benefit legislation became inevitable, Dr. Margaret Balfour confirmed the mill-owners’ argument that the main need of the hour was not statutory maternity leave with compensation but ‘welfare work’. Drawing on Curjel's earlier statement, Balfour argued that because the men did not bring their own wives with them

Neither Curjel nor Balfour was unsympathetic. They repeatedly emphasised the need for maternity benefit legislation. But their characterisation of women workers helped employers stave off such legislation. An employers’ association, the Bengal Mahajan Sabha, laid the failure of maternity benefit schemes on women workers themselves: ‘The peculiar type of female labour in the jute mills . . . does not conduce to the creation of schemes which presuppose normal family life’. Another, the Narayanganj Chamber of Commerce, argued that the ‘type of labour’ did not need maternity benefit since ‘normal family life was notoriously absent’. And also that, ‘child motherhood’, ‘indiscriminate procreation’ and ‘unmarried’ motherhood were ‘conditions that exist’ and had to be taken into account in framing legislation.Footnote ³⁹

While these arguments did not hold back maternity benefit legislation indefinitely, by the time it was passed in Bengal in 1939, the connection between working women and sexual immorality had assumed enormous proportions in public imagination. It fed, on the one hand, the working class family's desire to withdraw women from industrial employment when male wages improved; and, on the other hand, facilitated the employers’ strategy of eliminating women from the workforce. The proportions of women had embarked on its downward trend already by 1939, but from the 1950s, managers, trade unionists and policy-makers shared the conviction that it was welfare legislation that had ‘caused’ this decline. The Annual Report of the Factories Act of 1953 spelled out the argument that the proportions of women in the total factory population dropped rapidly after the introduction of Maternity Benefit (1939) and creche facilities. Women were employed, the Report pointed out, “mostly for the reason that they are cheaper than their male counterpart.” But, “[p]rogressive labour legislation . . . has made women labour costlier these days.” The Inspectors found that the general attitude of the factory management is to “do away with the women labour with a view to avoid in future expenses on special welfare benefits for women”.Footnote ⁴⁰

Such categorical statements notwithstanding, most careful enquiry fails to establish a direct relationship between passing of benefit, ‘protective’ or other ‘welfare’ legislation and the decline in women workers. When maternity benefit legislation was passed in Bengal in 1939, women comprised about 13 per cent of the jute workforce. The proportion declined slowly by a percent to stand at 12.4 in 1950.Footnote ⁴¹ The steep decline set in from the 1950s with rationalisation and formalisation: women were about 4 per cent of the workforce in 1960 and 2.5 per cent in 1971.Footnote ⁴² Thus, welfare legislation was deployed as justification by employers in periods of labour squeeze rather than it being the actual cause for retrenching women workers.Footnote ⁴³

Women and working class politics

The factories, mills and mines, which undertook the first large-scale retrenchment of women in the 1930s, were also to be the nucleus of the ‘formal’ sector covered by state regulation. These comprised the ‘organised’ sector not only because they were subjected to record and registration, but because, by the 1960s, their labour grew to be almost fully unionised with considerable bargaining strength vis-a-vis employers and the state. In historical accounts, this strongly assertive labour force is the ground for demonstrating the ‘making’ of the Indian working class. Most historians have seen in the process of organization the flowering of ‘working class consciousness’. These processes—unionisation and successful collective bargaining—have been seen, if only by omission, as gender neutral. Evidence suggests quite the contrary. The story of class formation is rendered gender neutral through a concentrated focus on successfully organised workers. The story is construed on a circular logic in which the gender identity of workers is left unmarked: (men) workers have worked in factories; factory workers have been unionised; and unionised workers have self-consciously asserted themselves as working class. It must repeatedly be emphasized, however, that it were male workers who constituted this celebrated ‘working class’ and, therefore, one primary axis along which the ‘working class’ developed was gender. It was male workers who were successfully ‘organised’, who were unionised and brought within the purview of regulated wages and working conditions. Over time, unionized men acquired a stranglehold over prized jobs in the organised sector; that is to say, (male) unions (far more than sardars in the previous dispensation) became crucial intermediaries in providing access to jobs in the organized sector. These unionized men adopted a variety of exclusionary strategies against women first and foremost, but also based on caste, region and language. The process of organisation thus went hand in hand with masculinisation. Indeed, it could be argued, women's marginalisation was imbricated in the nature and development of organised working class politics.

It is not as though women were absent altogether from the formal sector. Indian factory workers were only predominantly male. In the early 1920s, women were about 15–20 per cent of the workforce in textile mills, nearly half and more than half of the workforces of mines and plantations. Their proportions began to decline from the 1930s. Even in the 1930s, women were not invariably ‘docile’ or compliant workers. They also forged solidarities, sometimes with men workers and sometimes against them, and they undertook collective action, sometimes separately and sometimes in alliance with male-dominated trade unions. How then did women become so irrelevant to the increasingly powerful male-dominated unions within a period of about thirty years?

There is a history, though not much told, of women workers’ protests. There are also a variety of tense and uneasy representations of their militancy, both in contemporary and historical accounts. The researcher has to read them, as the cliche goes, ‘against the grain’, in order to see how women workers, as active historical agents, negotiated their gender and class identities. The problem, in this case, is that the evidence points in opposite directions. In much of contemporary writing and from oral evidence, we get two diametrically different images. On the one hand, women are described as ‘loyal’ workers, docile and manageable; on the other, in fragments and glimpses, frequent enough to attract attention, we come across the image of violent, aggressive and assertive women who took leading role in confrontations with the management and the police. Both these images have to be considered in some detail in the specific context of their representation.Footnote ⁴⁴

By and large, contemporary official and union records of working class political activities do not mention women at all. Descriptions of meetings, demonstrations, strikes, or even violent crowds, rarely contain specific reference to women or to individuals as women. Public meetings, in which women participated, were habitually described as if they were composed entirely of men, unless women were extremely prominent and therefore attracted attention. Strikes too were described without mention of women, though they must have been part of them. The term ‘men’ was presumed to include women, while other terms like ‘hands’ or ‘workers’ or even more commonly used notations like ‘mob’ or ‘crowd’ bypassed the question altogether. Addresses in leaflets, for instance, presumed an entirely male readership. Appeals to solidarity were addressed to the fraternity of workers. Trade unions imagined the working class as a homogeneous and solidaristic male population, an imagination inscribed in their very language. Women, when they were considered at all, were often expected to play a negative role in the labour movement.

The negative expectation from women found further amplification in both employers’ and unions’ stereotyping women as ‘docile’ and manageable workers. In such a discourse, women's militancy was usually displaced by victimhood. Thus they were portrayed as ‘victims’, in turns, of managerial, or state, or union violence. In all these cases, the attempt was to tap the mobilising potential of the ‘woman wronged’. Frequent references to atrocities by managers or the police were an easy and quick way for union leaders to gain public sympathy and strengthen resistance. Equally, managers talked about male workers inflicting violence on otherwise tractable and compliant women workers to delegitimise union activities. The ‘woman as victim’ was a powerful image because it personalised political resistance and often successfully harnessed public outrage and working class anger in support of strikes. But these portrayals were also double-edged in that they contributed to the passive and powerless image of women. Women's agency in moments of protest could be and often was reduced to—or even erased by—effete images of sexual violation or suffering motherhood. Thus even when women initiated strikes, newspaper and official reportage focused, not on their grievances and demands, but on the brutalities perpetrated on them. In writing about women, the image of the victim and martyr came more readily to the scribe. In such reports, the fate of victimhood loomed larger than the act of protest. Let us take an instance from Bengal: the Ludlow Mill strike of 1928 came in two phases. In the first stage, all the workers had gone on strike demanding the dismissal of two oppressive jamadars and asking for a wage raise. While the men were out in a meeting, management agents came into the mill lines and persuaded the women to join work on a verbal agreement to meet their demands. The strike was broken. When the mill management did not fulfill its promise, the women, unilaterally, went on strike. Before the men and the union had mobilised, the management, in a pre-emptive action, arrested some of the ring-leaders including six Telegu-speaking women of the preparing department. A large contingent of women marched to the local thana to demand their release. The police attacked the women and a fracas ensued. In reporting this incident, all the leading newspapers highlighted a single incident of a young child being separated from the mother. Ananda Bazar Patrika screamed the headline, ‘The Infant torn from Mother's Arms’ and went on to say, ‘after this, the other women workers were beaten up ruthlessly. They were pulled by their hair and dragged up . . . . One 60-year-old woman was beaten to unconsciousness . . . a one-year-old baby was torn from his mother's breast and flung to the ground.’ (11 and 13 June 1928) The case became so emotive that the District Magistrate, Gurusaday Dutta, went to investigate the case.Footnote ⁴⁵

The imaging of women as victims of managerial violence became more effective when linked with the metaphor of sexual violation. This is not to argue that women workers were not subjected to sexual molestation, as indeed the earlier description did not intend to underplay managerial and police brutalities. There is a great deal of evidence to suggest that violence was frequently used against women. But, incidents of violence could acquire heightened political significance in different contexts. Marking out men in authority like managers, supervisors, the police or even durwans and sardars as sexual predators afforded a potent and personalised symbol of exploitation. Thus, allegations of sexual abuse against European managers or Indian supervisors formed a basis for collective action, sometimes by women themselves, and often with the help and support of male colleagues.

In protesting against sexual abuse, women often invoked the notion of honour (izzat), which had enormous appeal for the working community. Invoking potential anger against sexual abuse to muster working class solidarity, however, meant that women workers were acting in their capacity as workers to reinscribe their gender oppression. The notion of honour, embedded in a cultural discourse that privileged above all women's chastity and ‘purity’, was a particularly male conception. Therein lay its power to tap men's emotions and attract their support. Thus, women workers sought to negotiate the contradiction within which they were situated vis-a-vis managers and working-class men. They thus gained the support of fellow workers in their protests against employers. However, they were drawing on a notion of honour in which they were a mere vehicle for assertion of male status and power. Women were incapable of agency in this discourse, which depicted them as victims. Such a notion of honour invoked, inevitably, the ‘protection’ of male workers against threatening managers and their agents.

But that such ‘protection’ was a fragile refuge and that honour like many other instruments was double-edged were proved repeatedly. The male notion of honour was constructed, typically, in the context of the family. But it was susceptible to extension from the corporate family to a wider community. And in the latter context, women became the site on which communities traded insults, ‘honour’ assuming a highly symbolic and codified meaning in battles between workingmen. Thus, women did often invoke ‘honour’ as a mode of resistance, but this same notion could be deployed by men workers with women as the passive sites on which imageries of ‘protection’ and ‘violation’ were played out. Notably, the ‘community’ on which women's ‘honour’ was inscribed was never static. When women successfully invoked honour to rally support to their cause, this community was co-terminus with class. But it was not always so. The construction of community and the encapsulation of its honour could be defined by language, religion or habitat. The boundaries of community, shifting continually across the interstices of these definitions were drawn and redrawn in moments of crisis and protest. At times, ‘honour’ of women was subsumed within the community and deployed during communal conflicts and rivalries. Such a societal construction of women's honour was essentially male and when deployed by male workers for asserting their identity, during class or community conflicts, such political symbolism reinforced the passive and powerless image of the working-class woman. The role of the victim was extended, from subjection to state and managerial brutality in moments of class conflict to a comprehensive vulnerability within and outside shifting categories of community.Footnote ⁴⁶

The extreme vulnerability and oppressive situation of the working class woman cannot be overstated. Nevertheless, they acquired the reputation being excessively violent, aggressive and capable of leadership, collective action as well as individual initiative. We find mention of such women from the beginning of the history of industrialisation, but they acquired a special edge and prominence in the 1920s. Kelman said of the Bombay cotton mill women,

She cites a case where women workers collected the money and hung a clock in their room. They refused to let the management pay for it or have anything to do with it. They may have feared that the management would tamper with the time. The attempt to reduce wages and workers in winding and reeling, which were women's departments, led, in the twenties and thirties, to a series of strikes spearheaded by women.

In March 1926, women colour winders in the Rachel Sassoon Mill went on strike against the introduction of cheese-winding machines and remained out for five days. On 2 January 1928, women winders of the Jacob Sassoon Mill refused to work when threatened with a wage cut. Initially 250 women stayed away from work. Soon after, the spinners joined them. Next morning, all the workers went on strike. This sparked off the general strike of 1928. In reporting the strike, the Labour Gazette sought to distinguish between the women's grievance and the causes of the general strike. The women's original dispute was about the reduction of wages in the winding department but its extension to other mills, said the Gazette, was due to opposition to the Indian Textile Tariff Board's recommended system of working. On the one hand, women's specific initiative is recorded, but subtly undermined as narrower interests vis-a-vis more generalised ‘class’ issues. But these strikes confirm the especially adverse impact of rationalisation on women's departments as well as women's ability to begin and sustain collective action on ‘economic’ issues. In fact, as the crisis deepened, many mills had to close down and others rapidly reduced workers in winding and reeling. Between August 1932 and 1933 there were twelve strikes by women against wage-reduction and retrenchment. Women workers went on strike on 9 February 1935 in Jacob Sassoon Mill demanding an increase in wages and the dismissal of the winding master, who, they said, was harassing them. It lasted two days and they resumed work ‘unconditionally’. A week later they went on strike against the introduction of a nine hours’ day starting at 7.30 a.m. While this accelerated frequency of strikes stemmed from a heightened sense of grievance, their greater effectiveness and sustainability was due to their ability to unionise. Between late 1930 and 1932, unions intervened in all strikes by women. The Girni Kamgar Union and the Bombay Textile Labour Union realised the potential of women's collective action. The Sassoon Alliance silk mill strike by women in 1932 was sustained over 5 months because the union played an active part in setting up a strike fund and distributing rations.Footnote ⁴⁸

Women in cotton textile industry in Bombay thus took collective initiative when the ‘women's departments’ were threatened and they often won the support of the major textile workers’ union. But women workers in other industrial centres did not share this experience. The jute mill women suffered more than most male workers in the limited mechanisation planned and executed in the mid-thirties. In 1941, in the Anglo-India (Middle) Jute Mill, Jagaddal, 118 women were replaced with 18 men in a mechanization drive. The women did put up some resistance but it proved ineffective.Footnote ⁴⁹

In a marked contrast to Bombay union practices, in Bengal trade unions were indifferent to (and often complicit with) women's retrenchment. In both these industrial centres, however, women workers were associated with ‘spontaneous’ strikes at a time when the labour movement was moving towards ‘organisation’. This was perhaps more marked in the case of jute mills where women resisted unionisation even in the sixties and even when formally unionised preferred to act outside it and on the basis of personalised relationships forged within the mills or in the neighbourhood. Prior to the growth of a strong trade union movement, which was delayed in the case of Bengal compared to Bombay, all workers acted in manners more or less ‘spontaneous’. They protested over immediate or relatively short-term issues like ad hoc demands over wage hikes, holidays, dismissals or restoration of allowances. If workers wanted an oppressive guard, supervisor or clerical staff sacked, they would organize sudden strikes. Some strikes protested wage cuts or perceived interference in customs like the conduct of religious festivals. These strikes were usually disorganised, sectional and prone to individual and group violence targeted at the state, the police, the management, and in some cases, against other workers. The Bengalis targeted up-country men in the late 1890s when migrants were crowding into mills depressing wages. Several riots followed from confrontations between Hindu and Muslim workers in the late 1890s and again in the 1920s. Strikers and ‘loyal’ men were prone to attack each other at the picket lines, at the mill gates and in working class quarters and settlements. These ‘protests’ followed no distinctive pattern and when women protested on or against much the same issues—harassment, wage cuts, gate-passes or acts of managerial, supervisory or clerical staff—they too adopted similar tactics. Like other workers they were apt to reach for the broom or the stick, chase the supervisor or walk out of the mill.

The balance, however, changed as unionisation made headway. The male workers were progressively organised, in Bengal especially after the general strike of 1937. Strikes were less violent, better prepared, better led and often reached the negotiation table where educated middle class leaders dealt directly with Government and mill authorities. The nature of demands changed. Trade unions bargained for industry-wide demands, long-term welfare measures, social insurance and labour legislation. The leaders believed, sincerely, that these changes were in the interests of workers and that higher gains were achieved from collective action. But women tended to remain outside the ambit of such trade union organisation.

Even in the 1950s and 60s women tended to protest over immediate issues and without support of unions. Their ‘irrationality’, which the unions failed to curb while they remained of some numerical importance in the industry, was in part a response to the union's indifference to their specific interests. But their activities have left an abiding impression in the collective memory of the jute industry. Workers, union activists, managers and supervisors talk about strong and aggressive women who were apt to make quick use of the broom. Such women rejected the discipline not only of the factory but its counterpart, the union, too. There is much ambiguity in the way these women are represented. At one level, there is considerable admiration for the courage women displayed and often a great deal of gratitude. This was evident in Bejoy Hazra's account of how women surrounded him and escorted him to and from the mill for days to protect him against arrest.Footnote ⁵⁰ Reports of strikes and ‘disturbances’, however, also dwelt on adversarial women in order to amplify female deviancy. Women workers were regarded as sexually deviant; and rough and disorderly women in the public space discomfited managers, the police and even union officials. This was in strong and marked contrast to public acceptance of middle class urban women's participation in picketing and demonstrations. These women drew on the moral force of nationalism, especially with Gandhi's encouragement in the twenties and the thirties. Factory women, however, were seen as ‘public women’ not only because they worked in the public world of production but also in their association with prostitution. When women strikers violated gender norms by taking to the streets—a symbolically male space but also frequented by the prostitute—derogatory associations undermined women's labour militancy.Footnote ⁵¹

Over three decades, from the 1930s to the 1960s, women became increasingly marginal in working class politics. With the growth of trade unionism, their ‘spontaneous’ modes of protest came under renewed attack, not only from determined employers but also from well-meaning union leaders. The union leaders unfortunately neither extended their good intentions to espousing women's special grievances nor did they make any effort to include women within the ambit of union activity. Women's inability to find some space within the (increasingly) powerful trade unions reduced their ability to resist directed and gendered retrenchment policies of employers. Their share in formal sector employment was sharply reduced, increasing women's dependence on male income and reinforcing male authority within the household. This long-term trend became increasingly entrenched; and values of female dependence proved remarkably resilient in the face of pressures from the labour market.

Recent trends in women's work: Identity and Autonomy

From the 1920s and 30s and up to the 90s, the formal manufacturing sector has continued to eliminate women from their workforce.Footnote ⁵² From a study I undertook in July 1997 comparing a rayon mill and a jute mill, it appears that while an older ‘crisis-ridden’ industry could undertake this elimination relatively gradually, a more dynamic rayon mill, which benefitted significantly from liberalization, was more hostile to employing women.Footnote ⁵³

By the 1990s, two decades-long uncertainties and periodic crises had rendered jute mill jobs highly insecure. Nevertheless, in comparison to the ‘informal’ sector, mill employment was still lucrative. As before, the most coveted were the small number of ‘permanent’ jobs but there were also a whole range of not-so-permanent jobs graded according to security of employment, number of days of work available and social security entitlements. The permanent workers were the best represented by the unions and tended to follow narrowly exclusionist strategies. For instance, unions and management agreed to reserve job ‘nominations’ to male heirs. Only when such male heirs are not forthcoming, would female heirs qualify. In new recruitment, as in case of the ‘zero record’ or ‘voucher’ workers (the most casualised section of the workforce), women were not entertained at all. Women tended to cluster in the middle band of ‘badli’ workers. And their numbers (as well as proportions) had declined further since 1970s.Footnote ⁵⁴

Women, it is often argued, are more adversely affected by economic liberalisation because they bear the brunt of the household's poverty. In India, economic restructuring and liberalization has been introduced piecemeal from 1991. The consequent assault on the organised sector, undercutting of organised labour as well as the political weakening of trade unions will, together, it is feared, lead to a general squeeze on labour. As more male workers are shunted into low paid jobs, household incomes will decline, more women will be forced into poorly paid work in the informal sector and total family welfare will decline. Women will suffer more because when there is less to go round, their share of household resources declines more steeply.Footnote ⁵⁵

These consequences follow, however, from existing gender inequities in the household and the labour market. Women will suffer more; they already do so; besides, not only will they suffer more than they do now, but also, presumably, more than men will suffer. Existing, persistent and long-term patterns of gender disparity and discrimination are inscribed into economic change. Economic actors like the state and its agents, employers and their agents, workers (male and female) and trade unions operate in this playing field, which is far from level.

It is thus important to examine the gender characteristics of the market in the long term. From a survey of garment industries in Delhi, garment and electronic industries in the Thane-Pune region of Maharashtra, the Coimbatore textile industry and five small industries (garment, leather, prawn processing, silk spinning and reeling) in Bengal, Banerjee offers a profile of the woman worker in the 1970s. Most of the women were casual workers working, typically, with low capital investment in tools and technology. There was lower familial investment in their education, health and nutrition. Most women workers were married, burdened with the entire responsibility of housework and childcare.Footnote ⁵⁶

The labour market shows two clear gender characteristics. First, women rarely take over men's jobs by undercutting them. Men, too, despite high unemployment, do not usually undercut women. The terms of competition in this labour market are somewhat differently determined. The gender composition of a workforce is primarily subject to male decision-making. Male workers’ familial controls over women's labour allow them to acquire the flexibility (usually associated with women) in jobs that are lucrative and where such flexibility is desired.

Employers’ preference for women workers, especially when male labour becomes overly demanding, cannot be expressed by a direct replacement of men by women (or vice versa). Men resist ‘women's jobs’ unless they are significantly redefined or relocated. The most common situation in which this happens is when machinery is introduced for a previously manual job. In periods of crisis, however, or if the terms of the jobs improve, men do lay claim to women's jobs. Both these tendencies were evident in the jute industry in different periods. To undercut male workers with female or child labour, however, employers have to undertake job-differentiation (usually wage cuts and casualisation) or even major reorganisation of production (usually shift from factory/workshop to home-based). This latter was the case in the biri-rolling industry. Thus, women do undercut men, but usually not at their own instance and not without considerable entrepreneurial initiative. There has to be an evident ‘preference’ on part of employers for them to move into men's jobs.

Over the years, there has been a marked tendency for men to move out of unremunerative traditional household activities like weaving. In West Bengal in the 1990s, for example, the family continues to be involved in handloom weaving by deploying women's ‘spare time’. The earlier taboo against women's working at the loom has broken down and women now weave at home while the men join the new power loom factories as wage workers. The men moreover continue to handle the ‘business’ end of the household industry—procuring orders and marketing finished goods. As a result, women have no control over the earnings from this productive labour they undertake in addition to housework and their traditional task of yarn-processing. The goods they produce are usually of poor quality since they have been given no opportunity to acquire the necessary skill or knowledge.Footnote ⁵⁷

Given these patterns, it is difficult to try and predict the impact of liberalisation on women's employment. Casualisation is not new in India and was never specific to women. The standard conventions of large-scale industry, except for a brief fifty-years period, have never been ‘rational’ in the Indian context and are breaking down elsewhere as well. Employers’ preference for casualised male labour often coincides with male claims on ‘women's jobs’, especially in periods of high unemployment. These tendencies have contributed to a long term, often drastic, exclusion of women from more rewarding occupations like factory jobs. Periodic exceptions are more due to supply bottlenecks than to changing entrepreneurial preference. It may in fact be argued that neither national nor transnational corporation strategies are likely to make any significant dent in the labour market in this regard. For instance, the spurt in export industries in the 1990s did not lead (as in the 1970s) to expansion in women's employment. In the case of Glaxo, a pharmaceutical company, desire for ‘flexibility’ in the 1990s led not to augmentation of the female workforce they had employed in the 1970s but to fragmentation of production and putting out to smaller enterprises. The new employers have preferred men workers who are more amenable to flexible working.Footnote ⁵⁸

In India, women usually join the labour force after marriage when they already have children. These women are constrained by their family roles as wives and mothers. They are less mobile and unable to commit longer hours to paid work. This is the largest and most vulnerable section of the female workforce, hampered by lack of education, mobility, opportunity to acquire or upgrade skills or find alternate livelihood. They are most liable to be pushed into working more intensively as unpaid family labour or as cheap labour. They do not offer the advantages that the factory girl in Thailand or Taiwan does to employers. In fact, married Indian women workers are less flexible than men; the men acquire greater flexibility by abjuring housework and childcare responsibilities, but also by controlling women's and children’ labour. So women and children often enter and exit the labour market in response to the male breadwinner's employment situation. They enter the market when it is shrinking and terms for labour are unfavourable—male earnings become more inconstant, inadequate or suspended. They withdraw when the employment situation begins to improve and the men are able to ‘win the bread’. Thus women are flexible workers—but more for the family than for the employer. The family's particular requirements place limits on their flexibility in the labour market.

First, the non-negotiability of housework as women's responsibility limit the time they can spend at work. An important part of their work begins after they return from their workplace. A 1997 survey reveals that there has been almost no change in the pattern of housework distribution. When women get help with housework, it is from other women members of the household or at best a teenage son. Adult men, even if unemployed, will not help with housework. This means that women cannot do too much overtime; they prefer to limit the amount of physical strain at the workplace;Footnote ⁵⁹ and they have to take more leave and holidays because they tend to allocate themselves lower quantity and quality of food and are, therefore, more prone to illness themselvesFootnote ⁶⁰ and/or because they are responsible for the sick and old in the family. It is the most vulnerable women—unmarried, deserted, divorced or widowed—who offer flexibility from sheer compulsion. In one study it has been found that married women put in long hour of overtime only when it is compulsory.Footnote ⁶¹

Second, training and gender link the workplace and the labour market with household hierarchy. On the one hand, young girls and women are denied scope and access to education and training by the family, long before they enter the labour market; on the other, machines and workplaces are gendered too. A sexual division of labour is embedded in each social and technical division of labour. Thus men and women tend to participate in different spaces, shops or sections of the factory and they usually operate or set up different ‘physical technologies’ that require skills or knowledge defined as male or female. This is justified as the ‘natural’ complementareity of their roles but is usually accompanied by a ‘vertical sexual division of labour’—accounting for discriminatory wage gaps, patterns of promotion and supervisory functions.Footnote ⁶²

Third, the dominance of personalised and informal channels of recruitment allows male workers to influence the gender composition of the workforce. Since inception, large-scale industry has depended on recruitment through myriad social networks based on kin, caste and regional affiliations. Such a ‘system’ became staple in large-scale industries by the end of the nineteenth century. Over time the ‘informal’ channels have been institutionalised.Footnote ⁶³ Given the slow growth rate of organised sector and the tight labour market, recruitment itself is fairly sluggish. There is almost no direct recruitment now. Recruitment takes place either through union channels or by nomination. The latter indicates the right of retiring workers to ‘nominate’ a successor. If a worker dies while in service, their ‘heirs’ replace them. Retiring workers sometimes auction this ‘nomination’ to non-kin candidates on the basis of the highest bid. Attempts to prevent such practices have led to the re-entrenchment of the ‘family’ in recruitment considerations. In some units, union and employers have agreed to restrict nomination to a ‘son’ or a ‘son-in-law’, closing off women's access to jobs by nomination, which was the channel through which many ‘widows’ had earlier entered factory jobs. Thus factory jobs become ‘property’ and women's inheritance disabilities are reflected in their access to factory employment.Footnote ⁶⁴

There are many fall-outs of these gender characteristics inscribed in the workplace and the labour market. Male supervisors’ notions of what kind of work is appropriate for women influence the pattern of their employment. Male values of segregation and male workers’ interest in retaining control and use of women's sexuality and labour are written into hiring practices. Moreover, the supervisors and other clerical staff can extract sexual favours from women in return for access to jobs which increases the opprobrium of factory work for women and ensures their withdrawal from such jobs when higher male earnings allow them to do so. Often, women move from highly paid factory work to most ill-paid home-based work. On the basis of a study of beedi-workers’ attempts at organisation in Hyderabad, Rohini Hensman argues that husbands and employers have a common interest in keeping women confined in the house while at the same time earning an income. And husbands have the social sanction to use violence to enforce their wishes.Footnote ⁶⁵

The better-paid male workers use the recruitment system to effectively exclude women from the more prized jobs (within the factory or from the factory altogether). Possibly they fear that women's induction will threaten their ability to maintain high wage levels. But why do some of them not use their personal and family channels to recruit their own mothers, wives, daughters or sisters in high paying jobs thereby augmenting their household resources? Obviously, the few who might desire better incomes for their women relatives are unable to contravene established rules of gender segregation. So, the more lucrative jobs go to male protegees while wives and daughters may be given the lower paid jobs where women tend to cluster. In part this also reflects a preference for male rather than female earners in the family. The better-paid workers are better able to maintain non-earning adult female relatives, and often do so. To have wives living in various modified forms of seclusion and domesticity in the city signals higher social status. Such practices further contribute to ghettoising women at the lower rungs of job ladders. Since their ‘own’ women do not work, the better paid and more influential workers (often also active union activists) have no obvious interest in promoting women for better jobs.

Thus, male decision-making in the family and the workplace impact on women's job decisions. In the situation of a tight labour market, competition among workers may run along gender lines too. A striking example of this is quoted by Rohini Hensman Banaji from an interview of workers at Hoechst:

Domesticity and Dependence

Discussions on women's work address two levels of concern: first, the obvious and immediate questions of livelihood and survival in the case of the poor; and second, more broadly, an engagement with the Engelsian question of the autonomy women may (or may not) derive from waged work. The question of women's employment and autonomy must, however, also address the family, which is the critical site for the construction of female dependence. Hilary Standing concludes from her study of women workers in Calcutta, that the meaning of employment for women lies in the ways in which ‘women's lives are bounded by the family’. The family, she argues, ‘is the critical site in the construction of female dependency’, which has implications for women's entry into and exit from the labour market and their capacity to secure their material situation. These structures of dependency are, moreover, related to, but not determined by the changing demands for female labour, or their class position. Between women and the labour market, the family remains an arena of intense ideological mediation.Footnote ⁶⁷

In her analysis, there appear four structures of familial dependency: economic, implying financial dependence on male kin; legal, through gender assymetrical marriage, domicile and inheritance; ideologies of domesticity, which constitute women as dependents and men as ‘breadwinners’; and psychic and social construction of feminine identity, which subjects women to male (sometimes older female) authority figures.Footnote ⁶⁸ Access to employment may challenge some of these, rarely all of these. More importantly, the terms on which the access is gained, and the ability to continue to maintain this access is very often determined by these structures themselves. Thus, the kind of women who work and the kind of work they do are very often contingent on male decision-making in the family, euphemised as ‘family decision-making’ or ‘household strategy’.

Importantly, patriarchal construction of female dependence is not uniform. Rather, ‘class’ (previously ‘caste’) differences are based on differential constructions of femininity. Multiple patriarchies constitute women of different caste/class status in different ways, which bear erratic and quasi-functional relations to each other. The visibility of lower caste women—in workplaces and streets—denigrate their men. The economic and sexual exploitation of lower caste/class women marks out the upper caste/class men's ability to ensure the security of their own women. Men of subordinate castes and classes struggle, in turn, to reverse the conspicuous denial to them of exclusive access to and control over their women's sexuality and labour power.Footnote ⁶⁹ Thus upward caste/class mobility gets easily fixated on visible and ensurable symbols of familial control over women: withdrawal from paid work; dowry and arranged early marriages; deployment in status-raising housework.

So far as the women themselves are concerned, there are varied and sometimes contradictory responses. Poor women are less able to stretch the given norms of feminine identity than urban middle class women. And they find themselves trapped between their understand-ing of discrimination and their aspirations for higher socio-economic status. In female workers’ own telling, they realise the actual similarity in the content of male and female work and point out that these are obscured by constructions of skill. Equally, they themselves often echo managerial and union distinctions of male and female tasks, especially on the grounds of physical strength and strain. These distinctions are sometimes integral to the expression of feminine identity. Given the hard manual labour that is routine to their lives, a public statement of physical weakness is a means of claiming feminine frailty. They also accept the management's arguments about protective legislation being the main reason for denying women jobs in the organised sector—these constitute a public recognition of, say, motherhood and sexual protection. But this is usually accompanied by an understanding that ‘protection’ applies to better-paid jobs. They also tacitly accept that for most men, their job is crucial to their identity. In their own cases, familial roles are more defining, publicly and privately. Women in insecure and heavy work in the informal sector also share the primacy of familial identities, but they often have no choice but to submerge these in the face of the demands of employers at great cost to their family lives, their health and their leisure time. The double shift looms very large in women workers’ day-to-day experience of work. They recognise and accept the non-negotiability of domestic work and its responsibilities (the primary defining feature of their feminine identity) and are, therefore, more ambiguous about their role as workers and commitment to the workplace.Footnote ⁷⁰

The ideology of domestic femininity has a special attraction for poor women. Paid work is usually monotonous, always physically demanding and carries the danger of sexual harassment. They are thus more susceptible to being pushed further into the family but simultaneously always vulnerable to being pulled out of it and into paid work during household exigencies. Paradoxically, this trend is reinforced by the difference in the earning potential between husband and wife, but also results in furthering this difference by directing women towards even lower paid home-based work. The 1970s survey of a jute mill found that an overwhelming majority of women (all from lower castes) believed factory work to be unsuitable for women. They would prefer ‘home-based self-employment’ drawing on traditional skills like sewing despite its low earnings. This response did not vary by level of income.Footnote ⁷¹

For all these women, aspirations for a daughter or daughter-in-law specifically excluded jute mill employment. Rather, marriage to a man who could afford to keep them out of the mills was more desirable. In the organised sector the withdrawal of women from ‘outside’ work became evident from the 1930s and 1940s. With yet more improvement in wages and working conditions since the 1970s, factory workers aspired to upward social mobility—education for their sons and dowry marriages for (non-productive) daughters. The withdrawal of the wife from paid labour outside the house is usually the first step in such a progression. These confirmations of the domestic ideal create more difficulties for women who do work in factories. In many jute mills, patterns of women's employment have held over a century. In 1891, Taroni told the Commission that only widows in unfortunate circumstances worked in factories.Footnote ⁷² Throughout the 1920s and 30s observers commented on the predominance of ‘single’ women in jute mills.Footnote ⁷³ In 1970 in the National Jute Mill, more than half the women workers were the sole earners in their family. The other half had been ‘forced’ into millwork by the husband's loss of employment.Footnote ⁷⁴ In Wellington Jute Mill in 1997, women workers were also almost entirely composed of widows and deserted women.Footnote ⁷⁵

It is not surprising that women workers do not consider mill work ‘suitable’ for women and endorse the role of the ‘housewife’ as most desirable for their daughters, even though the price of such desired marriages often cripple poor families. K.P. Chattopadhyay in a survey in the 1940s noted spiralling dowry demands among the urban poor.Footnote ⁷⁶ It is now widely admitted and bemoaned.

These attitudes were never universal.Footnote ⁷⁷ By the 1990s, more women appeared appreciative of organized sector jobs. Older women who had worked in the mills many years were not happy to relinquish their jobs and were not so particular about daughters and daughters-in-law abjuring factory work. However, they acquiesced with ‘family’ decisions to maintain non-working young adult women ascribing such decisions to the men in the family. Some women were more assertive: questioning managerial and employment policy regarding the exclusion of women. In the rayon mill mentioned earlier, women had a more uniformly positive approach to factory work, even though they readily accepted the limits of tasks and areas imposed by management and union negotiations.Footnote ⁷⁸

In the South Asian context, women garment workers of Bangladesh represent a remarkable exception. An active preference for docile workers has resulted in a predominantly female workforce in this industry. But the women workers did not comply with managerial expectancy of docility; indeed, their active involvement with unions led to a decrease in women's employment in the industry in the 1990s. Predictably, employers prefer unmarried or single women who constitute more than 60 per cent of the workforce (whereas they are about forty per cent in the total female workforce). Despite lack of security, considerable gender discrimination in wages, long hours and deplorable work conditions, women have a positive attitude to wage employment in the industry. They have benefitted, not only in intangibles like autonomy, self-confidence, improved conjugal lives and matrimonial relationships but also in decreased fertility, increased age of marriage and reduction in dowry demands. While wider societal attitudes remain negative towards these women's ‘public’ activity, they themselves retain a positive self-image. One of the surprising findings of a recent survey is that 35 per cent of the sample comprised women who had left their husbands in the village to search for jobs in garment factories.Footnote ⁷⁹

Amrita Chhachhi's survey of retrenched workers in the electronics industry in Delhi focuses attention on how much the women value their status and identity as skilled workers. These women resisted entering the unorganised sector and felt the loss of physical mobility and economic independence acutely. Many of the younger women, married and unmarried, were forced to undertake low paid or home-based jobs, which they saw as demeaning.Footnote ⁸⁰ In some cases thus wage employment has led women towards some forms of ‘autonomy’: self-confidence, assertiveness and new values of self-worth. In the Indian scenario, however, most workingwomen are already wife and mother and familial dependency continues to mediate their access to and experience of wage work. As a result, in certain historical circumstances, the improvement in male working conditions (as in organised industry) has promoted and fostered poor women's withdrawal into domestic roles, which have been the primary means through which they seek opportunities of self-improvement and status raising.

Conclusion

This selective sweep over a century of women's participation in manufacturing industry in India highlights some long-term processes of gender in the constitution of industrial labour. Chronologically, there have been at least three discernible shifts in this period. In the late nineteenth century, there was considerable concern over women's wage labour employment, which took the form of state regulation of women's working conditions in mines and factories, and repeated attempts at restricting women's recruitment for plantations. In the 1920s, the focus shifted to the working class mother and to welfare legislation aimed at protecting the working-class family. This shift could be characterized as a discursive one, having some impact, positive and adverse, on women's working conditions but not reducing their employment to any great extent. The social marginalisation of women workers set in trend in these decades was reinforced in the 1950s, by their political marginalisation. The process of formalisation and organization included both increasing state regulation and the development of federated unions attached to political parties. The effective unionization of male workers, started in the 1920s, led in the 1950s to considerable improvement of working conditions within the ‘organised’ sector. The interests of management, faced with spiraling cost of labour, and unions, operating in a labour surplus economy, converged towards a masculinisation of organized labour.

On the part of employers, abjuring women was partly a rationalization strategy in the face of increasing cost of labour, but the unions played an active role in both retrenchment of women and in creating effective barriers to their entry. These strategies adopted by organized male labour can only be understood in context of the role played by the family in mediating women's access to wage labour. In the Indian case, capitalist development has certainly not eroded the control of the family over women. If there has been change it is in the opposite direction; increasing violence and the absence of social welfare pushes women more deeply into the folds of the family, and this is truer of poor women. Moreover, the ideological mediation of the family has given men several advantages over women in the labour market: greater flexibility, a buffer against trade cycles and the relegation of the entire reproductive burden on women.

One of the distinctive disabilities of Indian women vis a vis the labour market is early marriage and child bearing. Women entered the labour market as wives and mothers, overburdened with double and triple shifts. This pattern began to change in the 1970s, when a gradual rise in age of marriage created an admittedly small but significant group of unmarried women who were mobile and able to access new opportunities of wage labour in export processing zones or with multinational companies. This was not wide enough a trend to lead to ‘feminisation’ of labour and, some economists have argued, was reversed in the 1990s. How far the liberalization policies of the Indian state since the 1990s have impacted on the broad gender characteristics of the labour market is not, however, very clear as yet.

Recent economic policies may have undercut the ‘domestic’ aspirations of the women of hitherto more secure and upwardly mobile working classes. Women's own access to such employment has always been very restricted and is likely to be more elusive. At the same time, other groups of women are being able to access employment before marriage—though poorer women's access to such employment, which emerged in the 1970s, may now be under some threat. For a vast majority of poor women, their work remains in the most vulnerable and insecure sectors. A majority of women continue to respond to ‘family’ imperatives as to when and where they work. It is likely that the recent changes in the labour market will reinforce rather than erode the pervasive value of dependence, which continues to inhere in definitions of Indian femininity.