Framing

This paper is based on a case study of construction labour in the central Indian steel town of Bhilai. Its immediate aim is to discuss the way in which social class on the one hand, and gender (and more especially sexual) relations on the other, are mutually constitutive. The wider agenda into which it is intended to fit, and which I explore in previous publicationsFootnote ¹ and in a forthcoming monograph, is an analysis of the hierarchy of labour in the town, of the process of class differentiation within its manual workforce, and of the eclipsing of caste by class as the dominant principle of inequality. Class has become, in most contexts, the main source of self-identification and is the most important determinant of life chances. It is the bearing that the ethnography presented here has on this issue of class differentiation that is its central preoccupation.

The gigantic public sector Bhilai Steel Plant (BSP) was constructed as a greenfield development in the late 1950s and 1960s in what is now the state of Chhattisgarh. Workers flooded onto the site from all corners of the country and many settled permanently in the town that grew up around it. By the late 1980s, BSP had 65,000 workers on its direct payroll, a total since halved—though its production and profits have steadily grown. Alongside the Plant is its spacious and orderly company township; hard by an industrial estate on which several hundred private sector factories are now located; all round a chaotic sea of residential and industrial sprawl inhabited by a vast multiplier workforce of informal sector labour. The larger urban agglomeration, dispersed over many square miles, has a population of around one million.

The regular Bhilai Steel Plant workforce is the local aristocracy of labour. In terms of consumption classes, as identified by National Council of Applied Economic Research surveys, even the households of the lowest paid workers fall comfortably into the most affluent quarter of all households in the country, and that is without taking account of the innumerable perks, subsidies and welfare entitlements that add at least 50 per cent to the value of their wages. Not only can they count on receiving this on the twelfth day of each month, it is so difficult to get fired that their posts constitute a quasi property right. They have jobs for life; and since no better jobs are locally available, hardly any ever relinquish them. With regard to income, terms of employment, life-styles and aspirations, Plant workers today merge seamlessly into the ‘middle class(es)’, which is how they think of themselves and are thought of by othersFootnote ².

Those with regular BSP jobs are not, however, the whole of its workforce. Through natural attrition and voluntary retirement, the Plant has over the past 20 years progressively replaced permanent employees with cheap contract labour, remunerated at a daily rate that is between one-seventh and one-eighth that of the lowest paid regular worker, but without any entitlement to fringe benefits, or to sick and holiday pay. Whilst the time and work discipline to which regular workers are subject is very relaxed, a greater intensity of labour is required of most contract workers. Their labour has progressively relieved the permanent workforce of the most gruelling unskilled tasks and is transforming it into a supervisory staff overseeing the labour of others. Most contract workers have no job security; all are subject to periodic lay-offs, many are made to wait weeks for their wages and some never get them. Although the regular workforce includes a disproportionate number of immigrant outsiders, the overwhelming majority of contract labourers are local Chhattisgarhis. Whilst the regular workforce is almost exclusively male, around one-third of contract labourers are women, who are susceptible to the sexual attentions of regular Plant workers (often their supervisors) and managers (to whose offices some—typically the prettiest—are assigned to clean and fetch water).Footnote ³ To the economic exploitation of contract workers, which is in significant measure what sustains the profitability of the plant and the privileges of its permanent workforce, is added an element of sexual exploitation.

In Bhilai's early days, the division between these two ‘fractions’ of manual labour was far more permeable than it has since become. An army of temporary labour was employed on construction and many were regularized as production operatives. Many others now rue the day that they spurned that opportunity because it was then possible to earn more outside the Plant as a carpenter, mason or petty contractor. As the pay and employment contracts of public sector workers steadily improved through the 1970s and 1980s, however, a gap opened up and became a gulf. In the cut-throat competition for these privileged jobs, it is the children of households that already have them who are at an overwhelming advantage on account of their cultural and social capital, and their superior educational credentials. Indeed, it is at least as likely that the son of a regular Bhilai Steel Plant worker will become an engineer, manager, chartered accountant or doctor as that the son of an informal sector worker will achieve permanent employment in the Plant. Plant workers inhabit a ‘citadel’ of state-sponsored privilege well protected against interlopers.

It is now conventional to stress that the labour market is multiple rather than dual. There are also well-guarded barriers to entry into even the most unenviable informal sector occupations. According to Holmström, ‘people at the very bottom live in little closed boxes, competing fiercely with other very poor people in other closed boxes’.Footnote ⁴ Breman similarly speaks of the ‘closed shop character’Footnote ⁵ of such employment that results from a pattern of recruitment through kinship links, that restricts sideways mobility and which inhibits the development of class consciousness. This ‘compartmentalization’ is commonly based on caste.Footnote ⁶ Though there are certainly some occupational niches that are not easily penetrated, in Bhilai by contrast many more of the labouring poor move readily and frequently between contract work in the plant, casual labour on construction sites outside it, loading and unloading jobs, and self-employment as rickshaw-valas, vegetable sellers, street vendors, waste-pickers and the like. It is only exceptionally that one of their number gets a regular job in the Plant.

If—following WeberFootnote ⁷—a ‘social class’ is the totality of positions ‘between which mobility either within the lifetime of an individual or over successive generations is a readily possible and typically observable occurrence’, those whose livelihoods depend on work of that kind constitute a discernible social class, distinct from the aristocracy of organized sector labour and with interests in some ways opposed. Bhilai Steel Plant workers often refer to such people by the English label ‘labour class’.Footnote ⁸ To suggest that this label might also apply to them would be offensively silly. As both sides agree, it self-evidently does not. Of course, there is no denying that there is also differentiation within this ‘labour class’—between those, for example, who exclusively rely on daily wages and those engaged in petty commodity production. The distinctions between them, however, are not of the same order and have nothing like the same salience as the distinctions that mark both of these out from those who have naukri (secure employment in the organized sector) and—above all—sarkari naukri (a regular job in the public sector). Typically, their children have different kinds of upbringing and very unequal life chancesFootnote ⁹; they place an unequal value on the conjugal bond and the stability of marriage,Footnote ¹⁰ have markedly different propensities to suicide,Footnote ¹¹ different ideas about the costs and benefits of industrial modernity,Footnote ¹² and different orientations to time. The lives of the labour aristocracy are characteristically geared to the future, to the education of children and their chances in the job and marriage markets, to the construction of a house for retirement. ‘Labour class’ people characteristically live for the day, in the here and now.Footnote ¹³ They have no alternative. Their penury forces them to live hand to mouth; and their job insecurity makes tomorrow uncertain.

Moving away from the wider framework into which this case study is intended to fit, this paper will firstly attempt to illustrate the radical insecurity of ‘labour class’ workers: insecurity over what, or even whether, they will be paid; insecurity against summary eviction from their homes for falling behind with the rent or for illegal encroachment; insecurity of life and limb in notoriously dangerous work, and insecurity of employment. For women in particular, this precariousness is compounded by the fragility of marriage. It is quite likely that they will be left to raise children on whatever meagre earnings they can themselves command. Not only are their wages very low but the number of days on which they will be able to find work is always uncertain. They or the children fall sick, so that even when work is available it is sometimes impossible to take. Of course, construction is a special case. Sites turn over rapidly, there is no fixed place of employment, labour requirements fluctuate day-by-day, and the world over it is an industry with a high proportion of short-term temporary workers. But though this is perhaps an extreme instance, the instability of employment and income is characteristic of most ‘labour class’ jobs in Bhilai.

Job insecurity forces workers to switch ‘lines’ of employment and I secondly wish to emphasize the fact that there is a particular range of occupations between which such workers typically and commonly move. It is this that makes them members of a distinct social class in Weber's terms (rather than being the shackled and parcelized inhabitants of Holmström's ‘little boxes’). For ‘labour class’ women the possibilities are especially constrained. Though some are engaged in petty trade or in the home-based production of country cigarettes or incense sticks, the main options are domestic service (often in the houses of Plant workers), waste-picking, contract labour in the Plant or construction work outside it. The wives of BSP husbands are seldom employed outside the home, and—unless driven to it by one who squanders his salary on drink—would never stoop to work of this kind. They are not financially constrained to do so, and nor is it respectable. For many ‘labour class’ women, however, there is no alternative. An appreciable proportion of households are female-headed, and even when there is a male breadwinner his wages are generally insufficient to feed the family. Women must work, if their children are to eat, in jobs in which they are vulnerable to sexual predation by men who are often not of their class—this being the third strand in the ethnography that I want to bring out.

Colloquially, construction work is inta-masala ka kam, ‘bricks and mortar work’. Here, masala is ‘mortar’ or ‘cement’; but in its commoner culinary context it means ‘spice’, and—as in English—conjures by metaphorical extension something ‘spicy’ or ‘racy’ (something masaledar). Inta-masala ka kam is ‘spicy’, though the after-taste is not so congenial to the liberal palate as Alpa Shah's account of ‘Tribal’ labour migration from rural Jharkhand to the brick-kilns of Bihar and Bengal would suggest.Footnote ¹⁴ As her Adivasi informants understand it, such work is less a matter of exploitative drudgery into which they are helplessly propelled by poverty than an opportunity to escape the ‘claustrophobic restrictions’ of the village and to enter a zone of freedom in which they can have fun and pursue ‘prohibited amorous relationships’.Footnote ¹⁵ Indeed, it is precisely because of the fun that they submit to the exploitation. So far, so good. What is less convincing is when she then goes on to portray the Jharkhandi social activists who rail against labour migration as puritanical middle class spoilsports who are rattled by a bit of wholesome sex and the threat it poses to the purity and proper control of women. What is striking is that all the amorous affairs she instances are, so to say, ‘isogamous’ ones between young Adivasi co-workers from the same local area; and that when the activists are allowed to speak for themselves what they turn out to be fulminating against is the oppression of tribal girls ‘by brick-kiln owners, contractors and middlemen’. In Bhilai, it is the frequency of this kind of ‘hypergamous’ sexual liaison on construction sites and the Bhilai Steel Plant shop floor that is striking, and it would be surprising if such liaisons are not also part of the situation with which Shah is concerned. To allow myself a value judgement, not all the sex is probably as wholesome as her discussion implies.

This is not to say that affairs of the sort she documents do not also occur in these contexts, or that women do not sometimes welcome and even encourage the interest of male colleagues. It may result in a respectable secondary union after their previous marriage has ended. But sexual advances are probably more frequently made by men who have power and authority over the women involved—by masons and mistris under whose immediate direction they work; or by supervisors, contractors and employers who often belong to the households of formal sector workers. It is not only that employers and their agents in the informal sector exploit the labour and sometimes the sexual services of the women they employ, though the employment relationship is certainly a crucial part of the picture. It is also that these men are commonly scions of the labour aristocracy. In Chhattisgarh, marriage is ideally isogamous, and—if such affairs develop—the wider the ‘hypergamous’ gap the less likely they are to result in stable unions, and the more likely to be seen as transgressive. The predominantly consensual relationships that occur between equals acquire an increasingly coercive edge as it widens. The ‘fun’ Shah stresses is more likely to be apprehended as ‘harassment’.Footnote ¹⁶

The asymmetrical character of many of these relationships has a bearing on my wider argument about class differentiation. It suggests something about the way in which class and gender intersect and reciprocally reinforce each other. With some help from the (pre-capitalist) ‘culture’ of their commonly work-shy men-folk, their class situation forces ‘labour class’ women onto construction sites where they are vulnerable to sexual predation by men with power over them. Some unsurprisingly prove susceptible, reinforcing the widespread belief that ‘labour class’ women in general are sexually available, which in turn provides ‘proof’ to the labour aristocracy that they themselves are a different and better breed, superior in culture and morals. Their women are pure. Class inequalities produce a particular configuration of gender relations; gender relations (and in particular sexual relations) produce a powerful ideological justification for class differentiation. It is a blindingly obvious proposition that has strong resonances with processes reported elsewhere in the world; but at least in the Indian context it has not been explicitly articulated and its significance for class formation has not been adequately appreciated.

True, DhawanFootnote ¹⁷ comes close in her discussion of the way in which highly paid women managers in Kolkata multinational corporations mark and define the class difference that separates them from their domestic servants. On her analysis, however, the ‘focal point’ in their discourse is their maidservants’ deplorable disregard for the legal bond of marriage and the legal requirements regarding registration and divorce, rather than their openness to extra-marital affairs (in which these professional women may themselves be involved). Here, by contrast, it is the supposed sexual availability of ‘labour class’ women that is the crucial marker. True, Lindberg has noted that the failure of low caste men to control the sexuality of their women impugns the purity of the whole caste, and that gender ideology thus legitimates and reinforces caste inequality.Footnote ¹⁸ My claim is that this observation applies equally well to class and that in the context described, class is the more relevant dimension. This is above all because ‘labour class’ women of all castes who work on construction sites are indiscriminately subject to male pestering.

In villages around Bhilai, the most heavily freighted distinction of caste is between the so-called ‘Hindu’ castes and the Satnamis, the largest ‘untouchable’ caste in the region. The ‘Hindu’ castes almost all rate as Shudras and have ‘traditionally’ tolerated divorce and remarriage, which were—and amongst ‘labour class’ people still are—very common. Bhilai Steel Plant workers of local origin have, however, acquired new standards of ‘respectability’, increasingly discountenance divorce, and have ideas about marriage that are increasingly divergent.Footnote ¹⁹ Whilst parents arrange the ‘primary’ marriage, in ‘secondary’ unions both men and women have considerable autonomy in choosing their own partners.Footnote ²⁰ In town, the caste panchayats have lost much of their sanctioning power, and these partners are now commonly of a different caste. Even in the past, women often left an unsatisfactory husband ‘to make a new man’ and the crucial context for this was their pivotal role in the agricultural economy. Chhattisgarhi women are renowned as hard working; the men as kamchor (literally, ‘work-thieves’). Patriarchy comes in multiple forms,Footnote ²¹ and by contrast with high caste males elsewhere, though in common with what Sen suggests is a more general pattern amongst lower castes, Chhattisgarhi men often seem more concerned with controlling the labour of women than with controlling their chastity.Footnote ²² Satnami women have a particular reputation for being unchaste, and Satnami marriages are indeed statistically somewhat less stable; but in these respects the differences between them and the ‘Hindu’ castes are greatly exaggerated by ‘Hindu’ caste discourse.Footnote ²³ That discourse is, of course, motivated. It is as Lindberg suggests: the supposed sexual laxity of Satnamis ‘explains’ their inferiority.Footnote ²⁴

What seems to have happened is that in the modern context this language of legitimization has been transposed from caste to class. The old idioms are now deployed to justify new forms of social distinction—in particular, that between the two kinds of labour. Of course this language is pre-eminently that of higher caste and class men, but it is in the nature of hegemonic ideologies that they colonize and infiltrate the consciousness of many of those they oppress. Lurid tales of sex on construction sites also suggest to many ‘labour class’ people themselves (both male and female) that it is only ‘loose’ women who work on them and that respectable ones remain at home. The consequence of doing so is a significant shift in the domestic balance of power, a polarization of gender roles, and—for women—a considerable loss in autonomy. But the economic realities are such that many households cannot afford that option, with the result that their women are revealed as ‘corrupt’ and ‘impure’, and the men as failing in their duty to provide and protect, as without ‘honour’ and—in effect—as little better than ‘pimps’. It is a discourse that denigrates both sexes. It is the ‘labour class’ as a whole that is demeaned.

The character of construction labour

The construction industry in India has attracted little sociological attention.Footnote ²⁵ After agriculture, it is the largest employer of labour in the countryFootnote ²⁶; it is estimated to contribute seven per cent of GDP,Footnote ²⁷ to employ 34 million workers and to sustain the livelihood of 16 per cent of the working population.Footnote ²⁸ Although bonded labour is illegal, even on government-funded construction projects it is common.Footnote ²⁹ By the 1980s, there were 25 separate laws that supposedly regulated the terms and conditions of workers in the industry. All are routinely ignored. Pink et al. estimate that 99 per cent of those employed in building trades in western countries like Britain are men.Footnote ³⁰ In India, up to a third or even half of those who work on construction sites are women.Footnote ³¹ Testimony recorded by the Report on the status of women workers in the construction industry suggests that sexual harassment is widespread, and that as a consequence of it jealous husbands often desert wives, resulting in an especially high incidence of female-headed households amongst the labour force.Footnote ³² But it also works the other way round: husbandless women are under greater compulsion to work on building sites where they are likely to be harassed.

In Bhilai, construction is the largest single employer of labour—especially female labour—in the informal sector; and with a booming economy, an expanding middle class and rocketing real estate values, the industry is currently buoyant. Its health depends on the steel market. In the mid-1990s, steel was in the doldrums and the housing loans hitherto available to Bhilai Steel Plant workers on extremely favourable terms were temporarily suspended. Construction slumped. It has since picked up with their restoration and the revival of steel, and with large numbers of workers coming up for retirement when they characteristically sink substantial sums from their Provident Fund pay-outs into house building. Short-term demand responds to the price of bricks, and to impending elections when unauthorized constructions are generally condoned. It is also seasonal. Labour is laid off during the monsoon; it is scarce when many return to their villages during festival periods, the marriage season and at times of peak agricultural activity. Contractors complain that workers are increasingly reluctant to leave them at all on account of high government subsidies on the price of rice and the availability of work under National Rural Employment Guarantee Schemes—though their complaints ring of employer fantasy about ‘feather-bedding’ the poor.

Much of my ethnography comes from small-scale sites, although I also observed large projects in the plant as well as the building of what I shall call, with deliberate imprecision, a sizeable municipal ‘leisure facility’. Some contractors are Plant workers running moonlighting operations or retired workers; more are Plant workers’ sons. To set up on any scale requires capital, though for those who begin at the bottom entry costs are low—just the price of some basic tools. Masons and carpenters frequently try their hand. Most fail. Employers default, forcing them back on their trades before trying again. With lucky breaks, however, some accumulate capital to invest in the equipment required to take on larger and more profitable jobs. Contracting thus provides a potential avenue for significant upward mobility within the informal sector, though it is seldom a gate-pass into the citadel of secure formal sector employment. Even the sons of a successful contractor who started life as a coolie or mason are unlikely to be as well credentialized as those who belong in it by birthright.

For large projects like the leisure facility, parts of the job will be subcontracted. Kedarnath was engaged for general building works, but the labour of other contractors assisted his and Seth Sahib's company workers with parts of the operation. The main ones are: sariya kam, fixing the iron rods over which the concrete is poured to give it strength; ‘centering’ or ‘shuttering’, setting up the wooden planks or steel plates that provide a mould for the concrete and support it until it is set; and dhalai, the casting of the concrete, a laborious and lengthy operation that cannot be suspended once started or the concrete will crack, that often lasts late into the night and may even continue throughout it, and that involves a long line of workers of both sexes passing ghamelas (shallow metal basins) filled with the mix and weighing seven or eight kilos rapidly up a human chain to the area that is being concreted—one every 10–15 seconds.

About one-third of the workforce are rejas, the female equivalent of coolie—unskilled general labour. All skilled workers are male. Though coolies and rejas share many common tasks, rejas do most of the portering, especially of bricks, which they carry six or eight at a time on their heads. Coolies mix and shovel sand and cement, and lift the bricks onto the rejas’ heads, but no self-respecting coolie will carry them on his own. Around 70 per cent of construction workers are Chhattisgarhis, but amongst women their predominance is near absolute. ‘Bihari’Footnote ³⁴ coolies and masons would rather go hungry than have their wives work on building sites and are scandalized by the seeming insouciance and indolence of Chhattisgarhi men. When I asked Lallu who carries the bricks in Bihar, I was told that ‘there we have donkeys’.Footnote ³⁵ Chhattisgarhi men put a subtly different gloss on it: ‘Here women are the slaves of men. We sit at home and they go out to work. But in U[ttar] P[radesh]-Bihar the men are the slaves of women and have to work for them. That is why they are always catching hold of our girls.’

More than three-quarters of Kedarnath's female labour in late 2003 were illiterate and few have any precise idea of their age. I frequently saw girls on construction sites that I judged to be no more than 14 or 15; only once a boy so young. Most rejas are probably between 20 and 45; a handful, older. The majority of these women are married, though a quarter or more are separated or widowed. Amongst the married, some have husbands who won’t work; more have husbands who can’t work because they are incapacitated by chronic disease (like TB) or by a work (commonly a truck) accident. Even working husbands do not reliably earn enough to support the household.

Male labour includes coolies, masons, ‘centering mistris’ (carpenters) and their ‘helpers’, and a supervisor or two. Skills are acquired on the job, and with application and minimal aptitude a coolie can rise to become a mason, a ‘centering helper’ to become a mistri. Skilled, but not unskilled, workers supply their own tools; and if female labour is included the ratio between them is about 3:7. The men are more likely to have had some schooling. Skilled masons (often Biharis) are more likely to be outsiders than coolies. Quite a few Chhattisgarhis (of both sexes) have worked on construction sites in other states.Footnote ³⁶ Most are originally from the surrounding countryside but are settled provisionally in town, though some men cycle daily up to 30 kilometres each way from their villages. There was a group of Adivasi lads from a remote village who both worked and bivouacked at the leisure facility, but it is not common in Bhilai (as it is in big cities) to see whole families camped on building sites.

Over 90 per cent of the workforce is Hindu, though some Bihari masons are Muslims. Sites are largely secularized spaces, and neither contractors nor workers bother about the astrologically auspicious time to begin construction, which is the owner-employer's headache. It is he who stages the inaugural bhumi puja to consecrate the site, though one of the workers will perfunctorily offer incense and smash a coconut by the cement mixer at the start of dhalai. Some large sites have a small shrine to Vishvakarma, the Divine Architect of the Universe, though except on the day of his annual puja nobody pays it any attention. Witches (who are usually women, tonhis) are a major cultural preoccupation in rural Chhattisgarh and a frequent topic of conversation in town. But whilst everybody concedes that some tonhis must work in construction, hardly anybody seems to have encountered one on site. One of the two accusations I heard about was not, I think, intended with much seriousness; the other related more to tensions at home than at work. It involved two rejas who were relatives and neighbours. Given the intensity of the sexual politics described below, it may seem surprising that the fear of witchcraft is relatively muted, but it seems to be of a piece with the instability of these jobs. This puts a cap on the development of intense suspicions about the secret malevolence of people one cannot escape. Construction workers seldom remain on the same site, or have the same workmates, for long.

On a big site, many are likely to have kin or mitan (ritual friends)—though married couples often avoid working together, and co-workers become fictive kin and address each other by kin terms. The caste composition of the workforce is heterogeneous and —despite the disproportionate number of Bihari masons—there are no caste monopolies over specialized tasks. Scheduled Castes and Tribes are represented roughly in proportion to their share of the population, and nearly all other workers of Chhattisgarhi origin are of ‘Hindu’ caste and officially designated as belonging to the Other Backward Classes— they are, that is, from low to middle ranking ‘clean’ castes (which form a large majority of the population in this region).Footnote ³⁷ I never encountered a Brahman construction worker—which is not surprising since they are a minuscule minority and usually have better options on account of their education. On site, caste is visible—if at all—in only the subtlest ways. Some rejas say they are ashamed to eat in front of ‘gents log’ (men), though coolies and rejas usually sit together to consume the lunch they have brought from home. Masons generally eat separately. These commensal groups are almost always of mixed caste and include Untouchables, their membership being determined more by friendship and neighbourhood ties. But although workers of clean ‘Hindu’ caste sit with Satnamis to eat,Footnote ³⁸ when it comes to the common and convivial practice of sharing the vegetables brought from home, some (not all) of the former avoid accepting them from Satnamis.

Unless previously instructed, Kedarnath's labour would gather each morning at a tea-stall on the main road near his house to be told which of his sites they would work on that day. There were often wrangles. Rejas and coolies mostly provide bricks and mortar to a particular mason who directs them. They tried to avoid Lallu because he abused them. There were also sites on which they reluctantly worked because the owner was an unreasonable taskmaster; and there were tensions over timekeeping. Puja rounded on Kedarnath's supervisor one morning when he scolded her for arriving ten minutes late. Where had his watch been on the last two evenings when they had been kept until 7.30? Whereas the contractor or his supervisor assign workers to particular sites and tasks, workers themselves decide on the precise division of labour. There is seldom much need for discussion. Yesterday, Kishora carried bricks and Jamuna mortar; today they swop over. The drudgery is lightened by joking and sexual innuendo.

When workers are taken on their wage is often unspecified. The contractor waits to judge their performance. There is no sickness or holiday pay, nor pay when weather stops work. Contractors require overtime but may not recompense it. For dhalai, however, an indeterminate overtime payment is normal. In 2004, Kedarnath was paying rejas a daily rate—well below the legal minimum—of Rs 40 (then about 50p), coolies Rs 50 and masons Rs 100—120. Over the past two decades, coolies have consistently received wages at least 20 per cent higher than rejas, and masons two-and-a-half to three times as much. These differentials go largely unquestioned. To put them in some perspective, if the household of a coolie had depended exclusively on his wage, and in the highly unlikely event that he had been employed six days a week in every week of the year, in 2004—2005 his household income would have fallen ten per cent below the official poverty line (set at the amount necessary to maintain basic nutrition). One that similarly depended on the wage of a reja would have fallen nearly 50 per cent below it. Even when both husband and wife are working, 80–90 per cent of household income may go on food. Though contractors complain of temporary labour shortages, wages do not rise in response. They compete largely on price; and—since the client buys the materials—their margins depend on the cost of labour. They are wary of inflating it.

Despite labour shortages, in Bhilai the practice of bonding workers by offering advances—well documented in other industrial settings and in construction—is absent.Footnote ³⁹ Rather than plying them with credit, as BremanFootnote ⁴⁰ reports, employers make advances only to long-serving workers and only then with reluctance. That is sometimes why workers leave; but contractors suspect that they will abscond or stop grafting if they get an advance. Rather than bond labour through advances, in Bhilai it is bonded by payment in arrears. Though contractors palm workers off with the excuse that they cannot pay them now because they themselves have yet to be paid by the owner or principal contractor, and though this is often true, they frequently retain a dabi Footnote ⁴¹—of, say, a day's pay per week. Workers are reluctant to walk out because they cannot afford to forego what the contractor owes them, and the further behind he falls the more they are bound. It is difficult to know when to bail out, and if they do so they have to take days off from their new work over the next several months to chase (often unsuccessfully) the arrears from their previous job.

Sometimes workers are not paid for weeks, or are fobbed off with a token amount and promised the balance ‘next Saturday’—though when that comes they usually get less than is owed. Often they are unsure how much it should be. They do not know what they will be given as overtime. Rational budgeting is impossible, and married couples often work for different contractors in the hope that if one defaults the other will pay. The bigger the contractor and the larger his sites, the better the chances of enduring employment. There is no such thing as a contract, and even on the leisure facility project—which lasted a couple of years—workers were often dismissed overnight on the whim of the foreman. A mistri with a lifetime's experience claimed never to have worked for the same contractor for more than two years. Some of Kedarnath's labour lasted just a few days, and of the 54 workers he employed in late October 2003, almost half were no longer with him 12 months later.

One reason that workers switch employers is that sites are dispersed, their contractor's new jobs are all far off and they cannot ride a bike. That keeps many rejas close to home. Their range is also limited by household responsibilities. Since most contractors do not allow them to bring their infants to work, breastfeeding mothers must find jobs close enough to return in the lunch break or for the baby to be brought to the site. And if they are to take jobs at all, there must be an older sibling, or a sister- or mother-in-law, to mind the younger children, though some employ a neighbourhood girl of seven or eight for a midday meal and a monthly pittance. When Jamuna went to work after her second husband ran off she was reduced to leaving their toddler alone, tethered to a steel trunk by a roped sari. Though most rejas are disdainful of waste-picking (because it is so dirty) and of domestic service (because it involves cleaning the saliva-polluted utensils of strangers), and claim that there is more ijjat (‘honour’) in inta-masala ka kam (bricks and mortar work), at some stage many are driven to such work. The money is at least as good, but more importantly the hours are more flexible and the children can be taken along.

Construction sites are dangerous; nobody wears harnesses or helmets, centering mistris work 40 feet up on six inch planks in flip-flop sandals, and during dhalai the bamboo scaffolding shifts and there is danger of collapse. During night work, much of the building—still without outer walls—remains unlit; and as it grows the rejas must carry head-loads of bricks up more flights of unguarded stairs with water dripping from the still wet concrete of the floors above. Many develop dermatological conditions from continual exposure to cement and find it painful to pick up hot food in their hands. Between 2003 and 2005, three of Kedarnath's workers sustained injuries. The most serious was when Lallu hit an overhead power line with an angle he was manhandling on a roof and was electrocuted. Fortunate to survive, his hands are now useless and work impossible. Though many would not, Kedarnath and the owner paid for his initial hospital treatment and sent rations for the next month, but now his household survives on Lallu's son's meagre earnings as a vegetable-seller. Meaningful injury compensation is non-existent, and I have never heard of a labour inspector inspecting a construction site.

Building work is completely un-unionized. The reasons are predictable: work groups are small and ephemeral, jobs and work places temporary, ‘troublemakers’ are easily fired and existence is precarious.Footnote ⁴² Workers who have been employed as contract labour in the Plant (as most have), and have experience of union leaders, have little reason to trust them.Footnote ⁴³ But though unions are absent, an ambiguous sense of solidarity is not.

Employers put word out through existing workers that they need extra labour, and workers hear about jobs through kin or more commonly through neighbours. Many go from site to site to enquire. Workers looking only for short-term employment, however, might offer themselves for hire at one of the day-labour markets (colloquially known as chauris).

The labour chauris

The two largest of these chauris are at Tehsil Chauk and at BRP Chauk.Footnote ⁴⁴ Though numbers fluctuate by season and day of the week, on average 200–300 day-labourers assemble at both gathering points each day. Workers start arriving around 7.30 a.m. and by 10.0 a.m. the market has emptied, though some who have not been hired hang on for another hour or two. Not all who come are looking for work. Those currently working together rendezvous there; some come to find an additional worker, some to gather intelligence about future jobs or to socialize, and some decide after their arrival that they do not feel like work that day. Most are employed in construction though labour is hired for many other manual tasks.Footnote ⁴⁵Rejas are often hired for catering tasks at weddings and other functions. At Tehsil Chauk, those offering different skills sit together and pass the time playing gota (a variant of jacks) and tiripasa (a dice game). Though I am unsure if it is true, a few women are reputed to be prostitutes.

These workers are not the atomized and kinless outsiders depicted in Aman Sethi's portrait of the world of Mohammad Ashraf in a Delhi day-labour marketFootnote ⁴⁶ or in Tom Gill's ethnography of a Japanese yoseba.Footnote ⁴⁷ Ashraf's existence is suspended between the positive value of azadi (freedom) and its negative corollary akelapan (loneliness). Although in a survey of 212 Tehsil Chauk workers, five were transsexual hinjras, and almost 40 per cent of the rejas were separated or widowed, most are integrated into local kinship networks. Akelapan is not their problem, but azadi is something they embrace. They choose when to work and are free from the tyranny of a long-term boss, and if they do not like the job they do not show up the next day. They generally insist on ending work at 5.00 p.m. and on being paid everyday. Many contractors regard them as recalcitrant and only recruit these workers as a last resort. A few are contract labourers in the Steel Plant waiting for their gate passes to be renewed or are between jobs with a regular contractor. For the most part, however, these are separate workforces—one strongly valuing the flexibility and freedom provided by jobs that seldom last more than a few days, the other submitting to a contractor who can provide steady work.

The asking price of chauri labour is around 50 per cent higher than a regular contractor would pay, and the gap with agricultural wages is considerably wider.Footnote ⁴⁸ Rates are, however, subject to ferocious bargaining and employers routinely renege on agreements—paying less on the pretext that the work was substandard or demanding extra tasks before paying at all. They vary by the job. For rejas, ‘party work’ as caterers’ assistants is two or three times as remunerative as working on a building site because it usually involves staying long into the night. Jobs that last one or two days pay better than those that last four or five—which is why employers regularly promise longer employment than they actually provide. Finally, rates fluctuate with the time of day. As morning wore on at Tehsil Chauk and workers became increasingly anxious about getting a job, rejas who would not accept less than Rs 40 at 8.0 a.m. would agree to Rs 25 by 10.0 a.m. .Footnote ⁴⁹ That was in 1996. At BRP Chauk ten years later, however, the rate went up as the day wore on. It was held that an employer who arrived so late must be in desperate need. By that time anyway many of those left in the market were those not really committed to working that day.

The number of days worked is extremely variable, and this is not only due to an insufficient demand for labour. Not infrequently, prospective employers fail to find workers at a price they are prepared to pay and depart empty-handed. Much depends on how committed the worker is to getting a job that day and his or her flexibility on the rate. A ‘Stakhanovite’ few are more or less fully employed; many obtain work for four or five days in one week and for only two in the next, and a significant residue resolutely stick to their asking price and more often than not go jobless. When asked if they would not be better off working five days at Rs 40 than two at Rs 70, the commonest response was a disinterested shrug, but other sources of household income seem critical. Rajiv was conspicuously relaxed about not finding work but could free-ride on his wife's earnings making country cigarettes. Sukhvaro came to the chauk with her neighbour Samrautin, but worked only half as much. Her husband earned well; Samrautin's had left her. As this implies, many (probably most) of these workers have much more in common with Max Weber's Silesian mowers than with his relentlessly profit-oriented Protestant entrepreneurs.Footnote ⁵⁰ The intensity of their labour depends on the developmental cycle of the household and its ratio of able-bodied workers to non-productive consumers.Footnote ⁵¹ Although elsewhere a woman's need to work is often a function of the lack of demand for her husband's labour,Footnote ⁵² here it is often a function of his willingness to work.

Many prospective employers first buttonhole a mason who, if interested in the job, then recruits the labour to do it. Naturally, those he takes on have been at the chauri some time and have networks. The mason does not usually demand a commission for putting jobs their way, though sometimes he is offered a lump sum for the task and profits as a contractor. Rejas need other attributes. One morning at BRP Chauk, I noted that none of the eleven women left without work could ride a bike. That is an asset when the site is far off, but is not a prerequisite since masons and coolies carry passengers on theirs. As Budhvantin, who was in her fifties, rather sourly explained: ‘Before seating you [on their bikes] those people first look at appearance. They take the young ones. . .people like us only out of compulsion (mazburi se)’. Or as Tikeshvari put it: ‘Just watch, there will not be a single young reja left in the chauri at the end. . .only the old ones and the ones who are jet black (kali-kaloti). All the beautiful ones will be called and taken’.

Chauri labour is constantly bewailing its lack of unity, and rejas the jealousies (dekhmari, jalankhor) and backbiting (chugli) that divide them. When Tikeshvari was sacked from a long job she blamed the jealousy of the one who had become the mason's mistress; and when a couple of weeks later a younger reja was taken in her place, she said it was because ‘that whore (chatkat), that prostitute (veshya) is going with him to get fucked’. Sexual competition apart, there is some implicit notion of the ‘just price’ for a day's work, and undercutting the rate is a major source of tension. It is common, but deeply and collectively resented—which is why I spoke of an ‘ambiguous solidarity’. The convention is that once a prospective employer has approached a worker, no other should negotiate with him. It is only when that worker has declined the job that others are free to bargain. One day at Tehsil Chauk, a painter grabbed Amit's bike whilst another clubbed his head and a chorus of rejas screamed abuse. Amit fled on foot, pursued by the two men; but they soon returned having failed to catch him. They had agreed a painting job for Rs 150, but when the employer came back on his scooter to conduct them to the site, Amit sneaked in with a cheaper offer. It was not the first time.

On another occasion, a mason called Santu recruited a team for a job. Before it was finished, he collected their pay from the employer and left early with the promise that he would deliver their money to the chauk in the morning. Having waited and waited next day, and having on that account missed out on a job, Budhvantin rallied the others—two rejas, two coolies and two masons—and they marched on Santu's house. Whilst the rest hung back, she hammered on his door but was told by a neighbour that nobody was in, though she would send her son to fetch Santu's daughter. When the daughter appeared and explained that her father had gone to their village, Budhvantin exploded. ‘The pimp, the bastard! He takes our money and goes wandering in the country. You tell him when he comes back that I am going to catch hold of his collar and beat him with shoes’. As this suggests, some of these women are extremely forceful, and rejas in general are not merely passive victims. If the question is why conditions in the Bhilai construction industry do not provoke more violent reactions and a more organized militancy, the answer does not lie in the large proportion of women in the labour force. When the leisure facility workers were smarting under a particular injustice, it was the rejas who were delegated to confront the bosses whilst the men kept mum. In the now all but defunct Chhattisgarh Mukti Morcha—the only genuinely radical union movement that the area has produced—many women were prominent activists.Footnote ⁵³

In response to my question about why more workers do not maximize earnings by working more days at lower rates, the most coherent responses I received were from Draupadi, who explained that when she had agreed to a week's work at below the usual asking price she had been humiliatingly abused by the other rejas; and from Til Bai who had declined such a job because of similar pressure. If undercutting the rate may be common it does not go without sanction—which in the case of the female workforce pre-eminently takes the form of character assassination, as the following fragments suggest. A reja has just agreed to a cut-price job and is trying to justify herself to the others. ‘Well’, comments one, ‘if it (your vagina) itches, go. Why tell us?’ Budhvantin cautions her companions against working for ‘that leper’ (rugaha), a supervisor who had underpaid on a previous occasion, but two of them ignore her. ‘Go, go!’, she screams. ‘They’ll come back fucked. . .people like that are banged up the anus.’

Employers can be just as crude. Jagdish arrives at the chauk in search of two rejas to whom he is ready to pay Rs 50. Kamla wants the rate, Rs 60. Jagdish mimes copulation with forefinger and fist, and observes that ‘those who do both jobs get Rs 60’. When Kedarnath's supervisor comes to recruit rejas and they wrangle over pay, he mutters that ‘those over-used cunts (bosdi) have grown arrogant’. And when my research assistant points out that the rate really is Rs 60, he impatiently says that he knows. ‘People pay Rs 60 or even Rs 70, but there they have to open their saris. That's what those bitches enjoy.’

Sex on site

Chastity and construction work are regarded as chalk and cheese. As Suraj complained, ‘people think all working women are like that’. When I reviewed with one old-timer the long list of Chhattisgarhi women from his neighbourhood now living in secondary unions with a man of different caste, commonly an outsider (often a ‘Bihari’), he sadly observed that almost all had gone off the rails as the result of thekadari, of ‘working for a contractor’. When Mohan, a carpenter on the leisure facility, explained that his wife does not work because there is no ‘honour’ (ijjat) in thekadari, I defensively admitted that my own wife worked. Naukri [regular employment], he reassured me, is something entirely different. During house-to-house surveys, women who stay at home would manage to subtly convey that, unlike those who work, they are respectable—though their gain in respectability comes of course at a price in terms of gender equality and the balance of domestic power. Stereotypically, and especially on small sites where they have more opportunity to be alone with rejas, the corruptors are masons; and it is said that rivalry for their attention sometimes brings rejas to blows. A coolie might call at the house of a reja without arousing the ‘doubt’ (shak) a mason's visit would provoke. Rejas would unselfconsciously pose for photographs with an arm on a coolie's shoulder, but the one my research assistant took of a leisure facility reja jestingly grabbing the wrist of a mason provided the foreman with a pretext for sexual blackmail.

The ties of fictive kinship that workers establish are sometimes invoked instrumentally. Poonam reckoned that Gautam gave her lighter loads because she called him jija (sister's husband). These ties are also key to the way in which affairs are initiated.

Most of this joking revolves around sex and marriage. Sita proposes that Gauri, a new acquaintance, should become the ‘ritual friend’ (mitanin) of Anjali because they have similarly obstreperous temperaments. Since Anjali is her classificatory daughter (beti), so now is Gauri and the joke develops into how she will find her a ‘mattress-vala’ husband, one physically and financially well cushioned. Often the humour lies in the word play. Damini begs for chuna (lime) to mix with the tobacco that many coolies and rejas addictively chew. Raju pretends to hear chuma (a kiss). Rajeshvari reports that she recently met her old malik (owner), the contractor who had formerly employed her; but her workmates choose to understand her to mean her first husband (bihata), provoking ribald remarks about how the bihata is more sexually exciting than subsequent men a woman might ‘make’. Some try to prevent the jokes from going too far. When Ankalaha discovered that he and one of the rejas with whom he had joked belonged to the same gotra (exogamous group), he immediately halted the banter by making her a ‘sister’. Lalita absolutely forbade male co-workers from calling her ‘elder brother's wife’ (bhabhi; bhauji), Jamuna worked alone when she could because she did not like the joking, and when 15 year-old Uma started on site she was embarrassed by the teasing and tried never to listen. Yadi larki hansi to phansi—‘if a girl laughs she's ensnared’.

Joking permits an exploration of the sexual possibilities. That provokes jealousy and sometimes violence. Lalita told of a liaison on her present site between a mason and the wife of one of the coolies which ended in a fight and them all being fired. Her own husband had been embroiled with another reja on a site on which the three of them were employed; though that did not prevent him coming to the one on which she subsequently worked alone and picking a quarrel with the mason for whom she carried bricks. In one instance, a jealous worker murdered a reja with whom he had been having a relationship by smashing her skull with a brick as she slept in the lunch break. She had gone to a fair with their supervisor. Though Bhilai has recently (and probably unfairly) become infamous for the highest incidence of reported rape in the country,Footnote ⁵⁵rejas say that the sex is not forced. Men desist if roundly abused. At the leisure facility, however, much of it seemed to amount to coercive seduction.

The leisure complex project was financed by government funds routed through the Muncipality. At the end of 2003 there were regularly 60–80 workers on site. Some were directly employed by Seth Sahib's company; some by subcontractors. Its completion was delayed when its boundary wall was blown down by a high wind, but—though a white elephant as a public facility—it is now a fine monument to bureaucratic and entrepreneurial venality. It has hardly ever been used and not one of the several dozen shops on its outer perimeter has been let. The day-to-day management of the site was under Deshmukh, its diploma-holding ‘superintending engineer’ from a middle class Chhattisgarhi family; and Kureshi, the foreman, a Bihari Muslim whose father was a Bhilai Steel Plant worker. Both were heartily detested for short-changing labour on wages and overtime, but I never heard it held against Kureshi that he is a Muslim. Seth Sahib himself was only occasionally present, but ‘Dau Sahib’—a runty and often inebriated engineer who represented the municipality—was often lurking. Since he was supposed to certify standards and the quality of the materials, the company had to keep him sweet; and when he was transferred to other duties midway through, Seth Sahib was said to have exerted himself in order to have the order rescinded. Those lower in the hierarchy were a couple of company supervisors who constantly changed but were often the sons of Steel Plant workers.

Though rejas were paid slightly less than on other sites,Footnote ⁵⁶ the job was desirable because it promised longer-term employment than smaller sites and because for most it was close to home. The work was also lighter since for everyday tasks it was over-(wo)manned. The company carried surplus female labour because it was regularly needed for dhalai when work really was intense, and usually went on into the early hours. Since much can happen in the dark, the bosses supposedly took an interest in spinning it out. On most days, however, the intensity of the rejas’ labour was not so exacting; though it varied considerably between them. A handful always had soft assignments (keeping the setting concrete wet with a hose; tidying the office) whilst the others carried bricks all day in the scorching summer sun. Breast-feeding mothers who refused to remain into the night or women who stayed home to tend a sick child would be told that from tomorrow they wouldn't have work. Others could disappear for days or storm off during dhalai and still keep their jobs. These inequities were deeply resented and led to divisions displayed in who ate together. When Santoshi, one of the privileged, returned after a fortnight's absence, two of the younger rejas decided that it was her turn to work and theirs to slack. The result was a hurricane of abuse about ‘vile prostitute bitches’ (kutti-kamini; kutti veshya). When Kureshi was reported as saying that Nitu was the only one who did any work, Gauri snorted, ‘Ha! And we people are just for fucking’. When I first dared ask why the ‘law’ is applied to some but others are assigned aram ka kam (‘restful work’), I was told that ‘that's something to be understood, not spoken of’. It soon was. The ones who enjoy ‘compassion’ (meharban) are the bibis (‘wives’) of Kureshi and Deshmukh. Sanjay supervisor would complain that his job was impossible because he could not discipline the bosses’ mistresses. Keeping those two happy was encouraged not only by the carrot of ‘restful work’ but also by the stick of being sent to another company site that was miles from home or seriously insalubrious. Worst were a couple of large public convenience facilities, one being constructed on a waste ground used for decades as an open-air toilet and densely strewn with excrement, the other adjoining a dairy where during the rains they worked up to their knees in a bog of buffalo urine and dung.

I am naturally unable to provide a complete inventory of sexual encounters on the site, nor even to be sure that all those I heard about really happened. Of those that very probably did, only one involved a reja and a coolie, an unmarried teenage girl of Kurmi casteFootnote ⁵⁷ from an unusual and extremely impoverished background, who had plainly come to the site in search of a husband and latched onto an Adivasi boy. When his family got wind of it he was whisked away to get respectably married; she went to work as a contract labourer in the Plant and got off with her Yadav (Cowherd caste) supervisor, though again his family thwarted their marriage. Shortly after she had eventually found a husband: he was murdered and she and the Yadav supervisor were arrested (though subsequently released for lack of evidence). One other affair, which culminated in the couple running off together, involved a married Oriya mason and a Chhatisgarhi Scheduled Caste reja who was separated from her husband. Three were relationships between rejas and supervisors. Kureshi—the foreman and in such matters a bull elephant—had the supervisors sacked. Two of these women had rejected his advances and—though now in eclipse—the third was supposedly still his occasional mistress. But now he was reckoned to have two younger rejas, whom he shared with Deshmukh, and was said to have propositioned at least four others. Gauri, the favourite, had been given office duties in succession to the fading Santoshi and was eventually appointed as supervisor of the other women. Though that mainly involved sitting in a chair outside the site office, she would bitterly complain about the burden of her responsibilities and the jealousy of the others who say ‘all kinds of upside-down (ulta-sidha) things about me so that my liver is torn apart’. She would have her revenge, however. ‘Beating, beating them, I will turn them into bhuts (malevolent ghosts).’ One of Gauri's duties—apparently taken over from the displaced Santoshi—was held to be that of serving as a procuress of younger rejas.

Dau Sahib, the municipal engineer, was a roué and some years after the event Suraj recounted how he had approached her:

She turned him down, but the next night was dhalai and he sent her on an errand to the office. On her return, he was lying in wait for her behind the stairs, but she managed to slip away in the other direction. Subsequently Gauri several times told her that ‘Sahib’ wished to meet her, but she never went and was exiled to punishment sites with Itvari Bai, who confided that she had had the same experience. At the instigation of Kureshi and Deshmukh, it was rumoured, Gauri subsequently managed to persuade 17 year-old Lachmi to take care of Dau Sahib's needs, and would conduct her to assignations with him on their days off. They were caught on site in flagrante delicto and it became an open scandal when Lachmi got pregnant and her neighbourhood boyfriend—who had spent time in jail—arrived on the site with a gang of drunken mates and threatened to ‘see to’ the engineer. Gauri arranged an abortion and Lachmi attempted suicide. Now some years on she seems contentedly married to the boyfriend.

Sex and class

Though I am convinced that caste is not the determining factor, in the interests of ethnographic candour I need to admit that several of these leisure facility liaisons involved a Satnami reja. The reason that I nonetheless claim that we are pre-eminently dealing with a class, rather than a caste, phenomenon is that on this particular site there were also a disproportionate number of Satnamis in the female workforce, and that most of the affairs that I heard about on other sites involved rejas of clean ‘Hindu’ caste. Though as we have seen Satnamis have a (much exaggerated) reputation for sexual promiscuity, and though that reputation may embolden men to proposition Satnami women, on construction sites it is—regardless of caste—rejas in general who are thought to be available.

Why do some acquiesce? Not a few, as I have tried to convey, are formidable personalities who are not plausibly portrayed as merely hapless victims. Indeed, as men sometimes represent it (though very likely in bad faith), it is women who ‘cast their nets’ and they who are ‘ensnared’. Others, however, are inexperienced young girls, like Lachmi, who are easier to manipulate; and Lachmi herself was, I infer, used as bait. Seth Sahib needed Dau Sahib to certify each stage of the job and relied on Kureshi and Deshmukh to keep him compliant. He did not need, and probably did not want, to know how that was done, though it is hard to believe that he had no idea. In any event, Dau Sahib wanted sex, and the foreman and site engineer could delegate arrangements to Gauri. For her part, Lachmi seems to have been ready for adventure. She came from an economically more secure household than the run of rejas and rather than being forced to go for thekadari, her father initially forbade it. She was wilful, however, complained of being bored at home, said she had always liked to watch masons laying bricks and had talked her parents round. As to others, the most likely motivations have already been alluded to—more ‘restful work’, exemption from the nastiest sites, enhanced job security and much cajoling. Given the suspicion that if not a housewife a woman is a whore, a reja might easily conclude that she may as well be hung for a sheep as a goat—and that must especially be so for those with feckless husbands who loaf all day, do none of the domestic labour and live off their wife's earnings. Nor is it hard to imagine that they may be impressed by the status and income of these men and may even compete for their attention. They would certainly expect a consideration in cash and possibly in kind.

Does that make it prostitution? Suraj was categorical. Sex is a prostitute's danda (her occupation or metier) and for none of these rejas is it anything like that. Why, Santoshi would do it ‘even without money. . .[because] her khurak [‘dose’ or ‘ration’, that is her sexual need] is greater [than that of other women]’. But if not prostitution, nor is construction-site sex morally acceptable. Rejas always refer to it as ‘wrong work, wrong relationships or wrong mischief’ (galat kam, galat sambandh, galat harkat). A galat woman who has galat relationships is opposed to one who is sahi (‘proper’ or ‘true’), which was invariably how my reja informants represented themselves. It was the affairs of others that I heard about. Their loud insistence that they themselves are sahi has the effect, of course, of underlining the possibility that there are some who are not.

The condemnation of such relationships seemed to be more severe when they involved an unmarried girl than a married woman;Footnote ⁵⁸ and to vary with the social distance they spanned—the larger the gap the greater the opprobrium. Especially for women, building work provides an opportunity to explore the field and to find a more satisfactory partner than their present one, or the one who just left them; and an ‘isogamous’ sexual liaison may well end up as a respectable conjugal union. ‘Hypergamous’ sex is far less likely to do so, and such relationships do not seem to constitute a realistic avenue for significant social mobility. Mistresses seldom, if ever, become spouses; and of the leisure facility rejas only Gauri—who always appeared the most instrumental—seems to have derived any lasting benefit. ‘Kureshi purloins our money’, said Poonam. ‘Madam’ [as they sarcastically called Gauri] purloins his.’ At the completion of the project she was able to build a house close to her mother's and somehow wangled a job doing ‘restful work’ serving water in one of the Plant offices. Already raising a child by a second husband who had left her, now she was able to bring her badly neglected daughter by the first husband to live with her. If she ‘made’ a new man (that is, took a third husband), she told me shortly after joining the site, there was no guarantee that he would accept her children and nor that he would turn out any better than the previous two. In fact, she might easily find herself supporting him too, as well as perhaps his children by a previous union. It was better, on balance, to stay single and cope as best she could. Her instrumentality, then, is easily construed as a product of circumstance, a strategy for coping with a rather weak hand; though it might also perhaps be understood as a duty imposed by the most sacred value of kinship morality, a mother's unconditional love.

How do the husbands react? None can be unaware of the temptations and some must know that their wives have lovers and be turning a blind eye, perhaps because the liaison keeps the household financially afloat. Santoshi's supervisor lover would come to the house and drink with her man. In Chhattisgarh, wrote Russel in the early twentieth century,

In contemporary Bhilai, however, things are no longer so relaxed, large numbers of apparently more savvy outsiders have settled and Chhattisgarhis have been made to understand that in matters of sexual morality their culture is shamefully deficient. What appears to be happening is that ‘labour class’ Chhattisgarhi men have been emasculated by a now more urgent sense that they ‘ought’ to be providing for their families and protecting the honour of their wives but can seldom find work that pays enough for them to do so. If the wife goes for thekadari she is suspected of infidelity, but when her husband quarrels on that account, she is likely to taunt him with his inability to provide.

Such husbands are doubtless a minority; and in the more general case—as I have repeatedly stressed---women work, not because their husbands refuse to do so, but because the household cannot survive unless both of them earn. The fundamental problem is the price of labour and the irregularity of employment—it is the product of radical class inequalities. This, however, is widely re-presented as a problem of gender relations. Women are exposed to ‘temptation’ by their husbands’ ‘inadequacy’Footnote ⁶⁰; and thus it is that ‘labour class’ women and men learn their respective destinies—to be ‘whores’ and ‘cuckolds’.

Of course, the situation described here is hardly unique and there are innumerable references in the literature to the sexual predation of higher caste and higher class men on the women of the families they employ in their fields and homes.Footnote ⁶¹ In Bhilai, elderly Satnamis tell of a time when the malguzars (landlords and revenue collectors) expected sexual access to their girls as if of right. Elsewhere, caste panchayats have tried to legislate against their women accepting employment on the land, or at least to regulate the conditions under which it is performed.Footnote ⁶² In an industrial context, BremanFootnote ⁶³ has described how in the 1920s and 1930s—and under the heavy influence of Gandhi and his disciples—the main textile workers’ union in Ahmedabad assumed that regulatory role, took up the ‘civilizing mission’ of ending the jobbers’ sexual exploitation of female labour by eliminating women from the workforce, and agitated for material conditions that would permit a single male breadwinner to keep them at home. And that is exactly what had happened during the previous century in much of western Europe, where middle class public opinion was seriously scandalized by the promiscuous mingling of the sexes in the mines and factories.Footnote ⁶⁴ But what seems to have been less clearly appreciated is that sex in the workplace may not only be an outrage to ‘decent’ middle class opinion (which is how this last literature tends to represent it), but also an important mechanism in the ideological reproduction of class.

It is a truism that in ‘traditional’ India caste purity, and hence caste status, is pre-eminently preserved through the purity of women. The corollary is that the impurity and inferiority of others is ‘proved’ by the impurity of their women. Though the ‘hypergamous’ sex I have described is less a matter of caste than class, I suggest that the values of caste leach into and give a particular shape to the way in which class is culturally constructed. Though I cannot develop the claim here, it seems plausible to suppose that the ‘naturalness’ of inequality in the world of ‘traditional’ caste tells us something about why class conflict often seems unexpectedly muted. In the present context, more importantly, the preoccupation of the ‘traditional’ caste order with the purity of women provides an idiom for talking about class relations; and on both scales of evaluation female impurity has become a key marker of inferiority. I am not of course suggesting that when a supervisor has sex with a reja he is merely performing a solemn duty of class reproduction, but what I am claiming is that this is a consequence of his actions. The supposed sexual availability of ‘labour class’ women is what most conclusively demonstrates to the labour aristocracy that they are different and better, that the ways of those from whom they are most anxious to differentiate themselves are properly regarded with ‘disgust’.Footnote ⁶⁵ It is also what demonstrates daily to labouring women that in the eyes of the world they are sluts, and to ‘labour class’ men that they are impotent drones who can neither protect nor provide for their wives and daughters. If, in Salzinger’sFootnote ⁶⁶ neat formulation, gendered identities in a Mexican factory are ‘made on the job’, on Bhilai construction sites so too are class identities, and it is moral and cultural values as much as the division of labour that makes them.