SPECIAL ARTICLEEconomic & Political Weekly EPW JUly 26, 2008109Dispensable Daughters and Bachelor Sons: Sex Discrimination in North IndiaRavinder KaurWhile pursuing the puzzle of the unwantedness of daughters, I was struck by certain parallels among contemporary northern jat and southern gounder peasant households.1 Both these social groups desire a family that consists of only one son. Why has the preference for children been reduced to such a minimum in these communities? To answer this question one has to return to the question of how families consciously and unconsciously decide on the size and composition of their families. Indeed, this is the pressing question facing us as we try and understand why large parts of the country evince a strong preference for sons and have rapidly taken to new technologies which allow couples to control family size and achieve a desired family composition of mostly boys with one or no girl. This paper, instead of going into the details of the already existing rich research on the causes of son-preference, female infanticide and female foeticide, argues that by concentrating entirely on the dyad of son-daughter, in which sons are preferred while daughters are not, we are missing out on significant gains yielded by taking other cross-generation and same generation members of the family into consideration. To develop our understanding, we need to take into account the varied roles men and women perform in the agrar-ian economy. Perhaps the key to the north Indian conundrum is not that sons are wanted and daughters are not but that daughters are not wanted while daughters-in-law are absolutely essential to the family. This will help explain the current “rush” to acquire brides from far-flung parts of the country by Haryanvis who have reduced their own female population to a pitiful 861 women for every 1,000 men. The paper also argues that we are helped in solving the puzzle of long-standing female adverse sex ratios in the north and north-west by including in our analysis family strategies vis-a-vis the sons in the family. Historical and contemporary data shows that the preference for sons is not as all-encompassing and undiluted as we have been led to think. Evidence is that among large agrarian castes more than a certain number of sons was often frowned upon and the family’s treatment of sons was often severely differentiated, privileging one or two sons while others were left to a lesser fate. There is no doubt, however, that unwanted sons did not meet the fate of an early death, which unwanted daughters did. Even today, a male foetus is rarely eliminated wilfully while innumerable female foetuses have been condemned to never being born as is made transparent in the recent censuses. The paper argues that for peasants in the northern countryside, especially those for whom fragmentation was a constant fear, the solution to the optimum family was not simply female infanticide but a combination of female infanticide and non-marriage of some sons. Both these strategies helped to reduce claimants on family Daughters may not be wanted but daughters-in-law are necessary for family well-being and perpetuation. Similarly, not all sons in the family receive equal treatment and those who are left bachelors suffer a lesser fate. This paper attempts to move beyond currently available explanations of low sex ratios and daughter elimination. While supporting the hypothesisthat large peasant castes in the north and north-west practised infanticide, non-marriage of men and polyandry as strategies to control family numbers in relation to available resources, this paper makes three arguments: one, that these strategies occurred together, two, that one needs to go beyond this explanation to understand why daughters were the dispensable ones and, three, that the number of sons wanted was by no means unlimited.I am extremely grateful to T N Madan and A M Shah for valuable comments on a draft of the paper. And, as usual, to Surjit S Bhalla for discussions.Ravinder Kaur (ravinder.iitd@gmail.com) is with the Department of Humanities and Social Sciences, Indian Institute of Technology, New Delhi.

SPECIAL ARTICLEJUly 26, 2008 EPW Economic & Political Weekly110resources. For men who were deliberately left unmarried, commu-nities sanctioned fraternal polyandry in which the wife of the married brother was shared by the others. A certain number of men remained bachelors or entered into sexual liaisons with lower caste women but were not accorded a marriage of their own.1 Existing Explanations for Daughter EliminationEarlier research has suggested various reasons why families wanted fewer daughters than sons – (1) hypergamy – the need/desire to marry daughters into higher status families placed a burden on parents; (2) dowry – similarly needed to acquire desired grooms or to conform to “honourable” norms of marriage; these two “cultural” culprits have to be understood in relation to marriages as alliance building and ultimately as part of political strategies which may be accompanied by significant economic gains at the local and regional level; (3) women’s lower value in wheat farming systems which require less of their labour or where their labour is less visible than in paddy farming systems; (4) seclusion or lack of participation in productive activities as an extension of the above (prevalent among certain northern high castes); (5) kinship, marriage and descent systems in which men inherit property and women move to unrelated and distant families to live with husbands; (6) patrilocality – the daughter moves to the husband’s home and does not inherit immovable family property, and hence, is unable to either contribute to her natal family or offer old-age support to her parents which her brother can; (7) furthermore, scholars such as Oldenburg (2002) provide a historical reason pointing out that the concentration of proprietary rights in the hands of individual males during the colonial period diluted unstated but customarily recognised female rights in property contributing to their unwantedness. Other reasons, ritual and practical for needing more sons (includ-ing physical defence of the family and its assets) then become icing on the male privilege cake leading to daughters being considered the unwanted or burdensome children. The historical record of female infanticide and excess female mortality in the north and north-west of India has been well-docu-mented in the censuses conducted by the British and in independ-ent India. Many scholars have analysed the female deficit [Agnihotri 2000; Bardhan 1974; Miller 1981; Mitra 2001; Vishwanath 2000, 2004, etc] and noted regional and castewise patterns. In addition, several ethnographic works have documented and analysed the persistence of low sex ratios among dominant peasant castes or among some urban propertied groups. It is among the same groups today that infanticide has been largely replaced by elimination of female foetuses through sex-selective abortions, even as son-prefer-ence and declining sex ratios spread to other social groups earlier unaffected by it. Jeffery and Jeffery (1997) report a history of infan-ticide and low sex ratios for the jats of Bijnor in Uttar Pradesh. Pet-tigrew (1978) and Hershman (1981) note female infanticide and adelphic polyandry among the small landholder jats in Punjab. Vishwanath (2004) discusses female infanticide in north and north-west, where not only elite rajputs, but also populous landown-ing groups like the lewa patidars, thakurs, jats, ahirs, khatris and gujjars practised female infanticide and had low sex ratios. More recent research confirms low sex ratios among the same groups – a survey in 10 villages in Morena in Madhya Pradesh showed child sex ratios of 392 among the gujjars, 400 among the yadavs, 417 among the rathors, 583 among the jats and 714 among the brahmins [Premi and Raju 1998]. Linking female infanticide in the past with the present desire for an only son family among the southern peasant group of the kongu vellalar gounders, Stephanie Vella says “Moreover the ideal progeny in this community appears to be today that of a single child, preferably a male, and traditionally they had only two sons so as to avoid the division of land. Therefore, infanti-cide and abortion were perhaps above all a method of family planning in the absence of more modern means of contraception” [Vella: 2005] (or sex determination, one might add). My concern is to explain the occurrence and persistence of daughter deficit among large peasant groups who are also gener-ally the dominant castes in the northern part of the country. Some historians and demographers have contested sociological explana-tions which centre on hypergamy and dowry as the cause of female infanticide and neglect of daughters. They argue that this explana-tion – the pride (hypergamy) and purse (dowry) – as Miller (ibid) puts it may fit upper castes such as rajputs who practised hyper-gamy and extreme female seclusion and hence exclusion from “productive” work but does not explain rampant infanticide, neglect and contemporarily female foeticide among peasants with smaller holdings and little practice of hypergamy. In this context, Oldenburg (2002) argues that female infanticide was practised by isogamously marrying peasant castes for “family balancing” reasons and that dowry became a significant cause of female infan-ticide only with economic changes brought about by British colonial policies leading to a further tightening of the noose around the girl child. She rejects explanations forwarded by the British for female infanticide – hypergamy linked to dowry and expensive marriage celebrations. Caldwell et al (1983) and Bhat and Halli (1999) also rebut the argument that dowry led to a desire for fewer daughters arguing instead that it was the shortage of marriageable men in the early part of the 20th century that led to the rise of dowry or groom price. Bhatt and Halli, however, attribute low sex ratios in the north/north-west to hypergamy and agree with most anthro-pologists that it led to infanticide of women at the top of the hierar-chy (as it was difficult to find higher status grooms for women in the highest clans and keeping them unmarried was considered dishonourable) and breakdown of caste endogamy at the bottom (as women moved up, a shortage was created at the bottom rungs). Men of lower status or from poorer households had to pay bride-price or buy wives and make socially less desirable marriages. Parry (1979) also reiterates the hypergamous marriage hypoth-esis for infanticide and neglect of girl children among the rathi rajputs of Kangra but is unable to explain why the brahmins who also practised hypergamy did not kill their daughters, merely stating that the top group obviated this need by marrying them to equals. It is possible that the political gains yielded by hyperga-mous marriages were not of importance to brahmins who were rarely politically significant in the north and north-west. But for most agrarian castes, especially dominant ones, hypergamous marriages were important for furthering status ambitions and were aspired to by many; a point Oldenburg entirely misses in her desire to lay all blame at the door of the British. As Parry’s

SPECIAL ARTICLEEconomic & Political Weekly EPW JUly 26, 2008111discussion (ibid) of biradari reform and similar efforts in Gujarat discussed by Shah (1982) and documented for Punjab during the colonial period show, attempts at restricting hypergamy by making isogamous marriage mandatory within endogamous units and reducing marriage expenditure generally failed due to ambitions of deviating families. The same process is still visible among Haryanvi jat clans today. 2 Producing the ‘Needed Family’Many scholars seem to converge on the reasoning that the elimina-tion of daughters had more to do with producing the kind of family agriculturists needed in the politico-economic terrain of the north and north-west. Such a family was definitely one with more boys and few girls. Miller (1981) calls it an indigenous form of family planning. DasGupta in her comparison of fertility decline in the Punjab and in Europe in the middle ages, states that household strategies to maintain a balance between population and resources in the first half of the 20th century, in the face of fluctuating numbers, included male celibacy (bachelorhood), reduction of female children through infanticide and neglect (since female non-marriage was not acceptable), informal polyandry, or buying of wives when needed. According to her, marriage was also used as a strategy to control population. She argues that in recent times the successful control of women’s fertility has made it possible to begin to reduce levels of permanent bachelorhood of men in this society (1995:491). Jeffery and Jeffery (ibid) make a similar argument for the jats of Bijnor in Uttar Pradesh and argue that infanticide, neglect of girl children and non-marriage of men were strategies “to regulate population growth in the interests of conserving the resources of the lineage” (p 239). There evidence is trenchant because it reflects a situation that has not changed from when it was first documented by the British in 1872. In 1990, they found that the demographic picture in the village they studied, Nangal, was still characterised by excess deaths of girl children and a relatively large number of unmarried adult jat men (ibid: 232). Jats and rajputs in villages of Etah district of Uttar Pradesh show similar patterns.The ethnographies of peasant landowners in what was the undivided Punjab (Punjab, Himachal Pradesh and Haryana) appear to support the hypothesis that infanticide, non-marriage of men and polyandry were strategies to control family numbers and that more importantly, these occurred together. Female infanticide or neglect of girl children were the tools adopted to keep the number of girls and women down. Equally, non-marriage of some sons was also a technique to limit the numbers of heirs and prevent fragmentation of land during the lifetime of brothers. Marriage of only one son also prevented infighting between brothers/couples which could lead to untimely subdivision of the land.2 The key ethnographies available to us – Hershman (ibid) and Pettigrew (ibid) for Punjabi Jats, testify to the extensive practice of surrepti-tious polyandry among them. Thus Hershman says “Polyandry is a recognised state of affairs in Punjabi society and it is basically condoned” (p 187). Pettigrew, who conducted fieldwork not just in one village but in a large area states that she came across numer-ous instances of a woman being married to one brother (p 53). Even though society disapproved of the practice, it accepted its widespread occurrence among agricultural castes. The patterns among the jats were often emulated in the caste of agricultural labourers, the chamars, across the north Indian plains. 3 The Dispensability of DaughtersWhile providing a plausible explanation, these arguments do not succeed in laying out why it was daughters who needed to be got rid of if it was a matter of keeping a balance between resources and humans. Yet, it is clear that between sons and daughters, daughters were more dispensable and more burdensome and less valued and hence the ones to be culled while men were simply left unmarried. It has been mentioned that while men could be left unmarried so that they reproduced only through a single wife held in common, the marriage of all daughters was a must. Among some high status families within clans it was also important that women be married into families of higher standing than their own. The necessity of marriage of daughters and in the case of higher status women, of an appropriate marriage, provide some of the reasons for why it was thought better to get rid of daughters at birth. But it still does not answer the question as to why middle peasants who were not beholden to marrying their daughters into higher status families or who could even receive a bride price for their daughters would kill them. In such families, the labour of women was also a necessity unlike among the elite rajputs where the extreme seclusion of women rendered them economically “useless”. While bringing the so-called “cultural” explanation of hypergamy and the necessity of marriage of women back into the picture, it is important to take a clue from Parry (ibid) and a recent paper by Chakraborty and Kim (2008) which points to the close links between kinship and marriage and the stability of northern political systems. Hyper-gamy was a major political strategy of clans and political power more often than not also brought economic gains with it.4 Daughters vs Daughters-in-law In the following sections, let me put forward my hypothesis which I feel adds something to the existing explanations of daughter elimi-nation. I am suggesting that to gain an explanation of daughter dis-preference, we need to look at a broader matrix of roles and relationships than simply the sons vs daughters axis. We need to look at the whole gamut of relationships in the family – the husband-wife dyad, the father-son dyad (or by extension brother-brother), and, interestingly, the mother-in-law-daughter-in-law dyad. This is obviously a simplified model/sketch since families will consist of multiple sons, daughters-in-law, etc. The point being made is that to produce a peasant family necessary to run the peasant farm, the basic unit necessary is not merely the agnatic unit of landholders but that it is important to consider individuals in their various roles. Men are equally important in the role of fathers/husbands/sons and we have to recognise that women are crucial to this perpetuation in their role as wives/daughters-in-law. Obviously, the glaringly missing figure here is that of the daughter/sister. Why is she missing, is she not in anyway necessary to the family? First, let us see why the others are necessary in the landscape of north/north-west India. The labour of males is necessary although there might be differentiation as discussed below as to the value of the labour of different brothers. The male as husband, father and son is necessary in a patriarchal, patrilineal system to run the farm,

SPECIAL ARTICLEJUly 26, 2008 EPW Economic & Political Weekly112to keep it viable and to inherit land and reproduce the family to pass on the land to. Additionally, it is important to recognise that peasant agriculture is dependent on both female and male labour. That female labour in the north may not be given the recognition it deserves is another matter [cf Miller 1981]. But in the landscape of the small peasant family farm both men and women are equally needed to make the land viable. The woman is responsible for several supporting activities and even main activities without which the family farm would collapse. Indeed, this is what makes marriage an “economic necessity” and leads to recognition of the worth of the wife if not of the daughter. Especially, in agricultural systems, in which crop growing and animal husbandry are combined and symbiotically linked, an enormous amount of labour falls upon the woman’s shoulders. Folk wisdom recognises it and so did Malcolm Darling when he said that the farmer would soon lose his land if he did not have a wife. He attributes a man’s inability to pay his revenue dues to his single status, since an “akela adami” (bachelor) would not be able to perform well agriculturally; and a widower was considered to be “half-paralysed” [Darling 1925, 1977: p 53]. Indeed, the British encouraged widowers and widowsto remarry. Leviratic marriage of the widow ensured that her labour would remain in the family and her claims to manage land on behalf of a young son and thus gain control over it would be blunted. Buying of brides and bride-price, practices which appeared appalling to the British initially were ultimately tolerated since they kept revenues flowing [Chowdhry 2007]. Among women’s roles, the role of the wife is equally important as a producer and a reproducer. She is the biological reproducer of sons and hence of the family, even if the family is “socially” repro-duced by the man. When a woman grows old, these roles are taken over in the family, not by her daughter but by the daughter-in-law, the son’s wife [Chowdhry on the importance of the daughter-in-law: 1994:50]. Hence, wives have always been needed even if not too many. The number of wives brought in could easily be controlled by the family but the birth of excess daughters could only be control-led through infanticide or neglect. The number of wives and repro-ducers could be kept down by allowing brothers to share one.3 But if the family needed another wife and one was not easily available in the community (due to daughters being killed or married into higher status groups) or was too expensive, she could be recruited from another caste or even from a long distance. A wife’s duties were made clear to her and the wife who did not produce children, especially sons was in danger of being sent back or supplemented with a co-wife. A woman who was widowed and did not have male heirs would either be married to the brother of the deceased man or sent back to her natal home [Hershman 1981, Chowdhry 1994].Thus, if one goes over the “needed” members and the “needed” roles, the one that is conspicuously missing from the scene is the daughter/sister. It is eminently clear that while a son can take over the father’s mantle – his position and his role in the natal family (and this is ritualised in north India in the “pagri-rasam” through which the transfer is effected after the death of the father, making the son the de facto head of the family (rather than the surviving mother), the daughter cannot take over the mother’s. The son can reproduce the father’s family but the daughter cannot reproduce her parental family. Her reproductive labour belongs to another family, i e, consanguine women cannot reproduce the lineage. Her productive labour may be of some use to her family – in some communities she is allowed to contribute her labour until she marries or during frequent visits to the natal home in the early years of marriage [Palriwala 2001]. But often these visits are inter-preted by the daughter as well-deserved periods of rest from the hard labour endured in the marital home. In the past, marriage at an early age ensured that she did not provide help for long although there are instances in which the marriage of a daughter is delayed either because parents need her labour or need time to accumulate an adequate dowry. In her in-laws family, the new bride immediately takes on, first the domestic tasks, then farm tasks, and the sooner she proves her fecundity by bearing a child, preferably a son, the better it is for her. So it is the daughter-in-law who steps into the mother’s shoes as she takes on the productive and reproductive roles of her mother-in-law. Thus wives and daughters-in-law, though they may be disliked and often barely tolerated and though their status may remain uncertain until they become mothers and mothers-in-law themselves, are needed and acquired somehow or the other. A reading of the history of jats and other agricultural groups shows that not all families among them could afford to be punctilious about where they got their wives from or how. Hence the easy violation of caste and other social injunctions among these groups as far as marriage was concerned. Jats and rajputs in Punjab were known to marry into castes below them and even intermarry with women from the untouchable castes [Hershman: ibid 180, Parry ibid, Chowdhry 2007]. In fact, most ethnographic evidence points to the tolerance of a variety of marriage forms – with or without dowry, bride-price marriages, hypergamous and isogamous marriages, those conforming to the norms of endogamy and those violating them, existing in the same space. What kind of a marriage a family or a man made depended on his socio-economic circum-stances and comparable status goals. The desperate need for brides and daughters-in-law is under-lined in contemporary Haryana and Punjab, states suffering from acute bride-shortage due to their extremely skewed sex ratios. In these states, men are bringing as wives women from states as far-flung as Assam, West Bengal, Kerala and Maharashtra, women with whom they share neither language nor culture. Even their caste and religion does not matter [Kaur 2004, 2007]. In Morena, a low sex ratio district in Madhya Pradesh, when families were asked who they would marry their sons to if there were no girls, the answer was that girls would come from “other families” [Field notes from Action Aid Project 2007]. To put it crudely and instrumentally, while families may love their daughters, daughters are the only “dispensable” members of the family. Not only are they structurally dispensable (looked at from the point of view of a single family or household) but they represent “outgoings” as far as household resources are concerned in contrast to sons.4 And if one is balancing members against resources, then fewer daughters are better and if there are none, then there is generally no regret. People go to extreme lengths to procure a son – in the old days through frequent and prolonged childbearing, in contemporary times by undergoing repeated abortions of female foetuses – but there is no pressing desire to

SPECIAL ARTICLEEconomic & Political Weekly EPW JUly 26, 2008113have a girl. More recently, even older couples in Punjab who are well past their child-bearing years are trying to have sons with the help of expensive assisted reproductive technologies. As one such man, Makhan Singh of Gurdaspur district, said “Only a son can wipe the paap (sin) of childlessness” Makhan Singh is 57 and his wife is 51 and they were visiting a fertility clinic to try and have a child (son) (India Today, December 22, 2003). The felt ritual value of a daughter earning merit through kanyadaan is fast disappear-ing. The son’s importance is continually reiterated including in his role as a brother. Most north Indian families express the view that a brother is necessary to facilitate the marriages of sisters. 5 Are All Sons Wanted? The demonstrated unwantedness of girls should not blind us to the fact that historically and contemporarily, there has not been unadulterated love for more than the necessary number of sons among farming communities. Proverbs from Punjab and Haryana warn against too many sons, even as the ideal family continues to be one consisting of more sons than daughters.5 In Bhagwatipur village, in Rohtak district of Haryana, a pandit family was not thrilled when the wife gave birth to twin boys at her third pregnancy recently. One of the sons was subsequently sent to her parent’s home [AA Project Rohtak 2007]. While sons may not ever be elimi-nated at birth or their birth actively prevented, not all were/are granted equal status in the family. These less-favoured sons are often second class citizens within the household. The family delib-erately chooses to let them remain bachelors or tries to disinherit them or makes them share one wife. Basically, they are reduced to what the Chinese call “bare branches”– men who will not have families of their own. In the Indian countryside they are referred to as “bechara” (one without food or resources). “Bachelors are much pitied in Pandit society” wrote Madan in his study of Kashmiri pandits (1965:101). Most ethnographies of the Punjab region document the phenomenon of the bachelor brother who was a marginal member of the household; his well-being dependent on how much land he stood to inherit and who among the other broth-ers coveted his share of the land and thus consented to take care of him. In Haryana, I was told by an informant that the status of a bachelor is that of a family servant. According to another inform-ant, bachelors are often not allowed into the house – they are sent food in the ‘gher’ (cattle shed). In Punjab, they similarly spent their lives in the cattle shed, called the haveli. A jat Sikh woman respond-ent while describing her husband’s father’s brother who remained unmarried said “he was not given any share in the land. So far as his living was concerned, he used to shift from one place to the other; sometimes at the Gurdwara, at times he would live with his sister or with his brother” [AA Project, Fatehgarh Sahib 2007]. In the Haryana field site, murders and suicides of bachelors were reported (village Bhagwatipur, Rohtak district, AA Project 2007].The incidence of bachelorhood (referred to as celibacy by some demographers) in north India– 7 per cent in the age group 45-54 in 1911 [Bhat and Halli 1999] in contrast to 3 per cent in the south pointed to a substantial number of unmarried men. The presence of unmarried men has been documented by a large number of ethnographic studies in the north [Darling 1925, Parry ibid, Jeffery and Jeffery: ibid, Minturn 1993 and others] with the result that a stratification of men emerged within families and endogamous groups – those who got to marry and those who did not.6 The kind of marriage various men merited could also vary. Pierre Bourdieu, commenting on the differential fates of older and younger brothers in Algerian peasant families says “Moreover, although there is no official recognition of any privilege for the eldest (of the boys, of course), everything conspires to favour him to the detriment of his younger brothers, to marry him first and as well as possible, that is, outside rather than inside the lineage, the younger brothers being destined for production rather than the exchanges of the marriage market or assembly, for work on land rather than the house’s exter-nal politics” (1977:69).7Demographers and sociologists have remarked how the short-age of women impinges negatively on men who are on the bottom rungs – whether of the family or of the clan. Men who are less desirable – who lack land, education, are jobless or have been tinged by scandal, are the ones who cannot find brides. My recent studies of long-distance marriage in Haryana and Uttar Pradesh (2004, 2007) reveal that it is sons who are unable to get local wives due to the acute shortage of women and whose families make little attempt to find them spouses, who arrange their own marriages with women from Assam, West Bengal, Maharashtra and Kerala. In doing so, they are apparently circumventing the family’s strategies to balance resources and family size. With individuals being less governed by the family, this shows the active agency of men in not leaving their fate in the hands of the family – parents or married brothers. An equally important factor is the contemporary margin-alisation of the bachelor in the rural household. As Amar Kaur, an elderly woman of the balmiki caste in Sirhind district of Punjab remarked, “no doubt, in the past too there were bachelors but not as many as today. But what is important to note is that, “previously, neither there used to be any fight over the distribution of land nor the women used to hesitate in giving food to the unmarried men. Instead the unmarried men had a status in the house; even after they were dead they were respected. The women of today will swear a hundred times before giving food to their younger brother- in-law” [Action Aid Adverse Sex Ratio Project 2007]. Amar Kaur’s views point to the transformation occurring in rural areas where effective nuclearsation of family relationships is taking place even within the joint family structure. It is the conju-gal relationship that is gaining in importance in contrast to the earlier emphasis on the agnatic bond. As couples focus more on the well-being and aspirations of their biological children and on the nuclear unit they tend to ignore obligatory relationships with unmarried siblings and even old parents. The demographically small-sized modern family has less time, space and resources for those outside its immediate circle.While marriage of all sons upsets familial calculations of suste-nance from the land, it explains the acute urgency and need for men to find off-farm sources of livelihood. The desire for govern-ment jobs, the investing of large sums of money to purchase public sector jobs such as those of havaldars or in the army, points to the necessity for small holders to diversify livelihood strategies. It is interesting that the discourse of non-marriage of males and the difficulties of finding brides in Haryana today is tied not to its abysmal sex ratios but to unemployment. It is often claimed in

SPECIAL ARTICLEJUly 26, 2008 EPW Economic & Political Weekly114Haryana today that as soon as a man finds a job he will get several offers of marriage. In the family, male livelihood insecurities indirectly feed into a sharpening of son-preference and instigate higher investments in male children as the family continues to repose hope for support in the male. Even among seemingly well-to-do agriculturists such as jats in the north and gounders in the south, it is the real and perceived precari-ousness of agricultural livelihoods that explains the desire for a fam-ily consisting of a single son. As the benchmark for living standards continues to rise and with it the costs of living, and agricultural yields plateau, the desire is to put all eggs in the basket of a single son. As older generations informed me in Fatehgarh district of Punjab, the younger child-bearing generations have no fear of the mortality of this one son. Parents of single sons amplified on the costs of educa-tion and the unviability of holdings if more than one son was pro-duced. Data from our study in Fatehgarh Sahib district of Punjab (the district with the lowest sex ratio in the country in the 2001 Cen-sus) reveals that if the first child is a son, the parents believe that the family is complete. In such a scenario the question of wanting a daughter does not even arise. This is the phenomenon of the “self-inflicted” one-son family. With the desire for zero daughters, the task of redressing the sex ratio imbalance becomes even more daunting and consequently our strategies for preventing daughter elimination need to be much more radical than those presently in operation.Notes 1 Stephanie Vella (2005) discusses the case of the gounders of Tamil Nadu. The information on Jat households is from fieldwork in Fatehgarh Sahib conducted as part of an Action Aid-funded project on adverse sex ratio (2007). 2 North India largely followed the Mitakshara system of inheritance in which all sons became heirs to ancestral property at birth. Hence, the peasant was always fearful of land fragmenting into economi-cally unviable holdings if the number of sons was too large. Keeping some sons unmarried, as elabo-rated later, also served the purpose of reducing the number of heirs. According to informants, the new inheritance laws which favour the equal division of property between sons and daughters are causing further upheaval in the north-Indian countryside. The share of land inherited by daughters is being reverted back to sons/brothers with payment of prohibitive transfer costs. This is possibly another reason for the tightening of the noose around the female foetus (Field notes from Action Aid Project in Fatehgarh Sahib, Punjab).3 Berreman’s work on Himalyan polyandry (1975) clearly outlines the context in which the system flourishes. Wives are recruited to the group as and when needed and polyandry serves a useful purpose if men have to practise multiple livelihood strategies of which male migration is one. 4 Daughters are popularly referred to as ‘paraya dhan’ or as guests in their own natal homes. In China, another country with a low sex ratio, bringing up a daughter is referred to as “watering the neighbour’s garden”. Even though society needs daughters who will become daughters-in-law, each family sees itself as not needing daughters and needing only daugh-ters-in-law. In Gujarat (Mehsana district), another state suffering from low sex ratios, exchange marriage (known as ‘ata-sata’, in Punjab as ‘vata-sata’) in which a family gives a daughter in exchange for a daughter-in-law is becoming common. Families with no daughters have to pay bride-price, or marry women from lower castes or tribal communities or their sons have to remain bachelors. 5 Prem Chowdhry (1994:50) quotes a proverb from Haryana which translates as “A daughter after two sons brings prosperity. Three sons in a row bring beggary.” Interestingly, Miller reports a study by Smith for 18th century Japan which showed that the Nakaharans practised a sort of family planning which was integrally related to the needs of agricultural production such that children not needed were killed. The author found evidence of male infanti-cide, generally of later-born males even though female infanticide was preponderant [Miller: 1981: 43]. She also quotes a study by Beals in south India where medical care given to girls and to later-born boys showed that both were less desired (ibid:101). 6 In the south, among the famous case of nambudiri brahmins, only the eldest son was allowed to marry a nambudiri woman. The other brothers had liaisons with nayar women [Gough: 1959]. A M Shah mentions a caste in Gujarat in which the eldest son remains, by custom, a bachelor, but succeeds the father as the head of the joint family with legal and moral authority. At death his corpse is first ritually married to a sacred plant in the crematorium and then cremated (personal communication, March 23, 2008). Praveena Kodoth (2006) mentions the opposite for nambudiri women who were not married even though the eldest marrying male could practise polygyny. For women who were left spinsters, a marriage ceremony had to be performed when they died and before they were cremated. These cultural requirements point to the importance of marriage as a lifecycle phase and to the reality of family marriage strategies which left certain men or women married.7 Recent research in Rohtak district of Haryana reveals that after the marriage of the first son, the marriages of subsequent sons may be simple affairs. The method followed is that of “chunni charana” – a group of five individuals goes to the bride’s house, a simple ceremony consisting of an exchange of garlands takes place and the bride is brought home. A full ceremonial marriage is not performed because, “social obligations” have been taken care of by the first son’s marriage.ReferencesAction Aid/IDRC (2007): ‘Addressing Adverse Sex Ratio in Selected Districts of Himachal Pradesh, Madhya Pradesh, Rajasthan, Punjab and Haryana’, Unpublished Country Reports from the Study.Agnihotri, Satish (2000): Sex Ratio Patterns in the Indian Population: A Fresh Exploration, Sage Publi-cations, New Delhi.Bardhan, P K (1974): ‘On Life and Death Questions’ in Economic & Political Weekly, Special Issue No 9:1293-1304.Bhat, Mari P N and S S Halli (1999): ‘Demography of Brideprice and Dowry: Causes and Consequences of the Indian Marriage Squeeze’ in Population Studies, Vol 53, No 2, July, pp 129-48.Berreman, G D (1975): ‘Himalayan Polyandry and the Domestic Cycle’ in P Uberoi (ed) (1993), Family, Kinship and Marriage in India, Oxford University Press, Delhi.Bourdieu, Pierre (1977): Outline of a Theory of Practice, Cambridge University Press, Cambridge.Caldwell, J C, P H Reddy and P Caldwell (1983): ‘The Causes of Marriage Change in South India’, Popula-tion Studies 37 (4), pp 343-61.Chakraborty, Tanika and Sukkoo Kim (2008): ‘Caste, Kinship and Sex Ratios in India’, Working Paper 13828, National Bureau of Economic Research, http:// www. nber. org/papers /w13828.Chowdhry, Prem (1994): The Veiled Women, Oxford University Press, New Delhi. – (2007): ‘Fluctuating Fortunes of Wives: Creeping Rigidity in Inter-caste Marriages in the Colonial Period’ in The Indian Historical Review, Vol XXX, No1, January, pp 210-43.Darling, Malcolm Lyall (1925): The Punjab Peasant in Prosperity and Debt, Manohar (1977 edition), Delhi.DasGupta, M (1995): ‘Fertility Decline in Punjab, India: Parallels with Historical Europe’ in Population Studies, 49, pp 481-500.Gough, Kathleen (1959): ‘The Nayars and the Defini-tion of Marriage’ in P Uberoi (ed), Family, Kinship and Marriage in India, 1993, Oxford University Press, Delhi.Hershman, Paul (1981):Punjabi Kinship and Marriage, Hindustan Publishing Corporation, Delhi. Jeffery, Roger and Patricia Jeffery (1997): Population, Gender and Politics, Cambridge University Press, Cambridge.Kaur, Ravinder (2004): ‘Across Region Marriages – Poverty, Female Migration and the Sex Ratio’ in Economic & Political Weekly, Vol XXXIX, No 25, pp 2595-603. – (2007): ‘Declining Juvenile Sex Ratios: Economy, Society and Technology, Explanations from Field Evidence’ in Margin – The Journal of Applied Economic Research, 1:2, pp 231-45, Sage Publica-tions, Los Angeles/London/New Delhi/Singapore. – (2007): ‘The Shortage of Spouses, Marriage Strat-egies and the Sex Ratio: An Examination of Cross-Region Marriages in Haryana’, unpublished paper presented at Jawaharlal Nehru University on September 3.Kodoth, Praveena (2006): ‘Producing a Rationale for Dowry? Gender in the Negotiation of Exchange at Marriage in Kerala, South India’, www. lse.ac.uk/collections/asia Research Centre.Miller, Barbara D (1981): The Endangered Sex: Neglect of Female Children in Rural North India, Cornell University Press, Ithaca.Mitra, Ashok (2001): ‘Implications of Declining Sex Ratio in India’ in Vina Mazumdar and N Krishnaji (eds), Enduring Conundrum: India’s Sex Ratio, Rainbow Publishers, Delhi.Minturn, Leigh (1993):Sita’s Daughters: Coming out of Purdah, Oxford University Press, New York.Oldenburg, Veena (2002):Dowry Murder: The Imperial Origins of a Cultural Crime, Oxford University Press, New Delhi.Palriwala, R (2001): ‘Transitory Residents, Invisible Workers: Rethinking Locality and Incorporation in a Rajasthan Village’ in K Sangari and U Chakra-varti (eds),From Myths to Markets, Manohar Publishers, New Delhi.Parry, Jonathan (1979):Caste and Kinship in Kangra, Routledge and Kegan Paul, London.Premi, M K and Saraswati Raju (1998): ‘Born to Die: Female Infanticide in Madhya Pradesh’ inSearch Bulletin, July-September 13, (3).Pettigrew, Joyce (1978): Robber Noblemen: A Study of the Political System of the Sikh Jats, Ambika Publishers, New Delhi.Shah, A M (1982): ‘Division and Hierarchy: An Overview of Caste in Gujarat’ inContributions to Indian Sociology (NS), Vol 16, No 1.Vella, Stephanie (2005): ‘Low Fertility and Female Discrimination in South India: The Puzzle of Salem District, Tamil Nadu’ in C Z Guilmoto and S Irudaya Rajan (eds),Fertility Transition in South India, Sage Publications, New Delhi. Vinayak, Ramesh (2003): ‘Grand Old Parents’ inIndia Today, December 22.Vishwanath, L S (2000):Female Infanticide and Social Structure, Hindustan Publishing Corporation, Delhi. – (2004): ‘Female Infanticide: The Colonial Experi-ence’ inEconomic & Political Weekly, May 29.