‘Too many girls, too much dowry’: son preference and daughter aversion in rural Tamil Nadu, India

Abstract

The southern Indian state of Tamil Nadu has experienced a dramatic decline in fertility, accompanied by a trend of increased son preference. This paper reports on findings from qualitative interviews with women in rural villages about their fertility decision-making. Specifically addressed are the reasons behind increasing son preference and the consequences of this change. Findings suggest that daughter aversion, fuelled primarily by the perceived economic burden of daughters due to the proliferation of dowry, is playing a larger role in fertility decision-making than son preference. The desire for a son is often trumped by the worry over having many daughters. Women use various means of controlling the sex of their children, which in this study appear to be primarily female infanticide. It is important to distinguish between son preference and daughter aversion and to examine repercussions of low fertility within this setting.

Introduction

Although sex preference and sex-selective practices in India that disfavour women have received much scholarly attention, they are not yet fully understood. Indeed, in 2003 Amartya Sen commented that ‘Sex bias in natality calls for intensive research today in the same way that sex bias in mortality – the earlier source of “missing women” – did more than a decade ago …’ (Sen 2003, p. 1298). He argued for more research into the traditional cultural values that perpetuate female disadvantage. The study described here attempts to throw light on the complex interplay between traditional and modern beliefs that influence family fertility patterns in a region of one state in India (Tamil Nadu), but which may be illustrative of a larger phenomenon.

The southern state of Tamil Nadu has long been regarded as a demographic and social outlier in India. Much research in Tamil Nadu has documented the dramatic fertility decline that occurred over the last half-century alongside relatively high indicators of gender equality, including an apparent absence of son preference (Dyson and Moore 1983, Das Gupta et al. 2003). Although son preference still existed at low levels, it was not much emphasized in the literature because it was so eclipsed by northern Indian rates (Arnold et al. 1998, Basu 1999). In the last decade, Tamil Nadu – along with its neighbouring southern states – has achieved very low fertility. In the year 2005–2006, the Total Fertility Rate (TFR, the average number of children born to a woman over her lifetime) was estimated at 1.80, down from a TFR of 4.97 in 1974 (Dyson and Moore 1983, Guilmoto and Rajan 2001, International Institute for Population Sciences and ORC Macro 2006). A two-child norm has percolated across Tamil Nadu, due in part to strong government promotion, and has been mainly achieved via female sterilization. Adoption of such permanent family planning methods is heavily influenced by the gender balance of existing children, however (Borooah and Iyer 2004). Post-pregnancy methods of sex selection have also been documented including abortion and infanticide. Abortion has been legal in India since 1971 and is available at government hospitals; however, sex-selective abortions are illegal in all of India (Clark 2000, Arnold et al. 2002).

This impressive decline in fertility has been accompanied by worrisome evidence of the emergence of son preference in the past decade, which is reflected in increasingly disproportional sex ratios at birth and younger ages (Das Gupta and Bhat 1997, Basu 1999, Srinivasan 2005). In rural Tamil Nadu, the social, economic and religious benefits of having a son warrant ensuring at least one male in a family of few children (Arnold et al. 1998, Borooah and Iyers 2004). Son preference is generally believed to be dependent on the current gender distribution of children and parity in a family and therefore as family composition changes with new births, the strength of son preference changes within a family (Clark 2000, Mishra et al. 2004).

Not only do the factors that lead to son preference cause daughters to be unwanted, but female children also are seen as having added disadvantages. In patriarchal societies, women have less economic input into the family and are therefore valued less (Malhotra et al. 1995). Due to the green revolution, there has been a decrease in dependence on female agricultural labour, again changing the economic value of women (George 1997, Srinivasan 2005). Past literature has also suggested that in addition to at least one son, people often want one, but only one, daughter (Pande 2003). Pressure on couples to have at least one son becomes heightened at very low levels of fertility (Park and Cho 1995, Arnold et al. 2002, Das Gupta et al. 2003), even in contexts where boy and girl children are generally treated equally.

Despite empirical evidence of the emergence of son preference in Tamil Nadu, there has been little investigation into the motivations behind decision making regarding the number and sex composition of children in the region. In-depth anthropological research in northern India has shed light on the complex interaction between falling fertility, sex preference and fertility in this context (Khanna 1997) and it is essential to look beyond the numbers in Tamil Nadu as well. Not only has there been little qualitative research in Tamil Nadu on this subject to date, there has also been little qualitative fertility research in general. It is essential to understand this state’s unique transition towards increased son preference with lowered fertility as the rest of India’s fertility is also continuing to decline.

The goal of the present study, therefore, was to identify themes regarding fertility preference among the rural poor in Tamil Nadu, who make up most of the population. As work was conducted in one region only, it is primarily a hypothesis-generating exercise pursuant to further investigation through in-depth research with a larger sample of women representative of the state as a whole.

A series of interviews were conducted with women in Tamil villages to better understand the changing sociocultural context of their stated preference for boys and their behaviour to achieve these preferences. We examined respondents’ perceptions of the benefits of both boy and girl children, especially in light of the emerging necessity to provide large dowries in marriage for daughters. In addition, we explored how fertility decisions are acted upon with the use of traditional practices and newer reproductive technologies. Findings shed light on the complex decision-making that couples undertake regarding their fertility preferences and the implications these decisions have for future fertility trends and gender equality.

Methods

The study involved fifty-eight in-depth interviews conducted between May and July 2005 with women from nine villages in two rural districts of Tamil Nadu – Theni and Tirunelveli – in the south near the Kerala border. Interviews focused on fertility characteristics, family planning and sterilization, overall health, economics and family structure. Respondents were ever-married women over the age of eighteen, as non-married women are unlikely to experience childbirth in this setting. All respondents were of childbearing age or had recently stopped childbearing (three respondents were over 49 years). Villages were chosen based on accessibility to a central site and respondents were recruited through the head of the local women’s group in each village, creating a structured convenience sample suited to rapid assessment.

We do not believe that the sampling methods adopted (of either villages or individuals) led to a biased sample of any particular characteristics (such as autonomy or caste) but, without accurate population level data from the region, we cannot be sure. Again, our aim was to open up the issue in an informed way before moving to a more representative sample, with its attendant logistics, resource and time requirements. All of the women involved were currently or had been recently making childbearing decisions. The sample was therefore not formally representative of the entire population of married women in this region, but was broadly typical of the surrounding population as shown in Table 1.

Table 1.

Background characteristics of women in Tamil Nadu: comparison of respondents and the National Family Health Survey (NFHS) 2005–2006.

	NFHS 2005–2006	Respondents (n=58)
Total fertility rate	1.8	2.3
Mean ideal number of children	1.4	2.3
Women with 2 children wanting another (rural)	5.4	8.8
Any current use/ever used contraceptive method	62	17 (10)
Sterilized	57.7	67 (39)
Illiterate	30.6	28 (16)

Individual interviews were chosen as the research method due to the possibly sensitive nature of the questions, especially those pertaining to gender preference, abortion and family planning. In a few settings, it was not possible to interview the respondent alone, but all respondents were interviewed in a female-only environment and there were no men present during any of the interviews. Informed consent was obtained orally. All interviews were tape recorded and were carried out by four different interpreters, with the principal investigator being present throughout. The use of these methods ensured, to the best of our ability, that respondents were at ease during the interview. The fact that the research team was all female and that for the majority of the interviews the respondents were of the same age if not slightly older than the research team, led to comfortable discussion. The study was reviewed and approved by Brown University Human Subjects Review Board.

All interviews were translated by the interviewer, transcribed and entered into the database by the principal investigator and analyzed using Nvivo, version 2.0. As a check for consistency, a random selection (roughly 10%) of interviews was translated by someone other than the principal interviewer. The codebook was designed and tested with advice from experts in qualitative analysis at Brown University. The Nvivo coding structure highlighted themes of fertility history, fertility and gender decision-making, sterilization and family planning, knowledge of abortion and STDs, breastfeeding and spacing, economic decision-making and dowry and marriage decision-making. In the text that follows, the names of respondents have been changed to uphold confidentiality and to ensure anonymity.

Results

General background characteristics

Respondents lived in villages with populations less than 3,000 surrounded by agricultural fields growing mainly rice and lemons. They lived in huts made of mud with thatched roofs, most of which had electricity. Families cooked over an open fire using wood and dung for fuel. None of these villages had running water or modern sanitation. Residents had to travel to the nearest larger town by bus to sell their goods and buy most commodities.

One quarter of the respondents were illiterate and one-sixth had completed high school. The majority of respondents had married a relative and ranged in age from 12–31 years at time of marriage, with a median age of about 19 years. At the time of interview, respondents ranged in age from 20–65 years; however, only three of the respondents were not of childbearing age (over 49 years). Almost half of the respondents worked inside the home earning about US$1 a day making matchstick boxes or sewing clothes. The rest of the respondents primarily worked in agriculture or did not work for wages outside of the home. The majority of respondents came from lower castes and were generally very poor.

Low fertility and gender preference

Fertility decline in south India has been very rapid and, thus, the desired family size has been relatively small during the time that most respondents were childbearing. This is reflected in the fertility intentions voiced by respondents as well as in their completed family size. Almost all of the women interviewed stated that they wanted only two or three children and, of those who had completed childbearing, all of the respondents under the age of 40 (with one exception) had fulfilled this fertility desire (Table 2). Most of the women (69%) still of childbearing age who had stopped childbearing through sterilization had only two children.

Table 2.

Fertility preferences and other background characteristics of respondents.

	Respondents (n=58) (%)
Ideal family is one boy and one girl	48
Ideal family is two boys and one girl	17
Ideal family is four or more children	7
Ideal family other or no response	18
Actual family size: 0 children	2
Actual family size: 1 child	8
Actual family size: 2 children	59
Actual family size: 3 children	21
Actual family size: 4 children	6
Actual family size: 5 children	2
Actual family size: 6 children	2
Of those with one son and one daughter, wants no more children	100
Of those with two sons, wants no more children	92
Of those with two daughters, wants no more children	77
Had an abortion/knew someone who had an abortion	15/28
Ever heard of any contraceptive method (other than sterilization)	53
Dowry given in marriage	57

Economic factors were the primary concern for respondents in decision-making about number and sex composition of children. Respondents were quite poor and children were viewed as expensive, especially in a rapidly modernizing society. Respondents specifically spoke about the necessity and high cost of providing education and healthcare for their children. One respondent explained:

If we only have one or two we can give the best education, the best food and everything is best. Otherwise, within our income it is harder, so two is best. (Kalpana, aged 33 years, 3 girls, 9 years of schooling, sterilized)

Even with their limited incomes, by lowering their fertility, families were acting on their stated belief that it was important to place value on the quality of children, rather than their quantity.

Perceived benefits of boy and girl children

Almost half of respondents remarked that they specifically wanted one girl and one boy, many citing the national family planning slogan ‘One plus One’.

[We wanted] one boy plus one girl … I feel that both boys and girls are important, to avoid the feeling that ‘we did not have a boy or a girl’. Both are necessary. (Binita, aged 32, 2 girls, 12 years of schooling, not sterilized)

About one quarter of respondents thought that boys and girls required equal expenditure and that people should ‘… not discriminate or separate out between boys and girls’ (Rohini, aged 22, 1 boy, 1 girl, 6 years of schooling, sterilized).

Reasons cited for wanting a boy were similar to those commonly mentioned in the literature on son preference in India – namely economic and social benefits. In their comments, respondents focused on the benefits of having a son and no one mentioned specific additional costs of having one or more sons. However, respondents did note both benefits and costs of daughters. The major benefits of daughters were instrumental and emotional: emotional support in old age, help in the house, affection and maintaining a good family name:

Two boys and one girl is enough because two boys will support themselves, and the girl will be more useful to me. When I am old with problems, she will come to help me. (Priya, aged 47, 2 boys, 2 girls, no schooling, sterilized)

Even if girls have gone to the mother-in-law’s house they [girls] will think of their (own) mother and come back and be caring. (Sunita, aged 65, 3 girls, 2 boys, 10 years of schooling, not sterilized)

Respondents compared the economic benefits of boys with the emotional benefits of having girls:

I really wanted a girl because after marriage a boy will only look after wife and family, but a girl [will continue to be] very affectionate to her mother, that’s why I want a girl child. (Durga, aged 25, 1 boy, 1 girl, 10 years of schooling, sterilized)

I really need one boy and one girl, my husband needed only boys because he has three sisters, so that’s what the problem of arranging marriage and all the things for the girl child is very expensive. So that is why he has chosen that boys are good. But I feel that I need a girl child because the girl child is so caring compared to the boy. (Anjali, aged 26, 2 boys, 12 years of schooling, sterilized)

Dowry and daughter aversion

Despite the perceived benefit of having one daughter, respondents identified economic costs associated with girls, which lead to an aversion to having too many daughters. The majority of respondents who mentioned not wanting a girl cited dowry as their primary concern. The use of a dowry in marriage was perceived by the respondents to have risen dramatically, both in cost and prevalence, in southern India in the past few decades. While only about half of the respondents themselves had a dowry in their marriage, almost all expected to have to provide a dowry for their daughters and planned to demand a dowry from their future daughter-in-law’s family. One respondent explained:

During [the] olden days, there was no need for dowry and [the] cost of living was less, so women gave birth to more number of children. But now the condition is entirely changed. All the sophisticated items came to rural areas, so people … have to spend more money for each child. So more number of children are not possible and women are willing to be sterilized. (Priya, aged 47, 2 boys, 2 girls, no school, sterilized)

Another respondent explained that a dowry was an important symbol of prestige for a family and therefore its use was justified, particularly in rural areas where people are interested in raising their social and economic class.

Respondents mentioned that they felt compelled to give dowry in order to protect the wellbeing of their daughters. If they did not give the amount of dowry demanded by the husband’s family, their daughters could face hardships and even violence. As there were reportedly more marriages outside the family and the village, it was increasingly hard to ensure daughters’ welfare.

Although respondents seemed for the most part ready to demand and give a dowry, they had mixed feelings about the benefits of the practice overall. One younger respondent said:

I am ready to give dowry for my daughters because now men’s families demand dowries. But I feel that in rural areas, mother-in-laws are doing cruelty to daughter-in-laws because of dowry matters, they are demanding dowries and it should be changed. (Lakshmi, aged 26, 2 girls, 9 years of schooling, sterilized)

Respondents appeared to be aware of these potential abuses and, although many of them thought the dowry system should be abandoned, they nevertheless felt compelled to continue the practice because of societal pressure.

According to one study in South India, the average dowry equals approximately two-thirds of a household’s assets (Rao 1993). Dowry has a large impact, especially on a poor, low caste family that may use it in an attempt to secure upwards mobility. As one respondent explained:

We decided only two because we need to have more money to raise more than two. Dowry was a problem for us, it costs two lakh (two hundred thousand Rupees) for dowry for girls. We both do coolie (hired labour working in fields), so we can’t afford to have more than two. The government also suggests no more than two. It’s an incentive for people not to try for a boy. (Gayathri, aged 35, 1 boy, 1 girl, no school, sterilized)

About one-fifth of respondents cited the economic burden of dowry as the primary reason they did not want daughters:

My husband particularly wanted a boy child because there is no need to give dowry and all those things that are expensive after they grow up. (Asha, aged 25, 1 boy, 2 girls, 8 years of schooling, sterilized)

The economic condition decided that two children is enough. Fewer is better for a family because the cost of raising children is high and cost of living is high compared to old days, so giving birth to more children is not good to lead an economically good life. My doctor convinced me to stop giving birth after two girls. He said that in my community you have to give large dowry, so it is not good to give more number of children. So I stopped and was sterilized. (Jamuna, aged 31, 2 girls, 12 years of schooling, sterilized)

Each of the above cases provides evidence of daughter aversion because of economic factors, which are heightened for girls due to dowry. Respondents are forced to seriously weigh the costs and benefits of continuing childbearing at the risk of having another daughter.

Decision time: women with two or more daughters and no sons

In the following section, we will examine the experience of women who already have two or three girls and have no boys. These women are at an important decision point in their childbearing careers in terms of whether they will go on to try for a son at the risk of having a third or fourth daughter. It is at this point in the childbearing process that we find evidence of daughter aversion becoming significantly stronger and more influential.

In this study, the majority of the women who had two daughters and no sons were sterilized, suggesting that having two girls was a strong motivator to stop childbearing:

I feel that my economic condition will not permit me to have more number of babies. If we want to give proper education and healthy environment to the children it is not possible to have more number of children. However, if we have boys, the boy will support during the olden days, until death and so initially we were interested in boy children. But we have been convinced now. After the birth of the first girl child, both my husband and I wanted to have a boy child, but we had a second daughter. Because of family situation and economic situation, because my husband is a coolie – getting daily wages – we have been convinced. (Lakshmi, aged 26, 2 girls, 9 years of schooling, sterilized)

Everyone wants boys, people do not want girls. I am really scared because I have two girls. Boys have a chance to go out and earn a lot of money, girls I have to support. My mother-in-law said I needed to have a boy, she said mean things to me when I did not have a son. But I decided on a sterilization because I already had two kids. (Amrita, aged 27, 2 girls, 7 years of schooling, sterilized)

Each of the above respondents felt that the economic pressures outweighed the potential benefits of a son and so they were willing to stop childbearing even though they had two daughters. Some respondents who had had only girls were sterilized without their husband’s knowledge in order to be able to stop childbearing:

We have stopped because both the children are girls, and we have to give dowry to girls, so we stopped giving birth. My husband is interested in children, and he expected to have a boy child, but now he is convinced because … we are belonging to economically not rich family. I am not interested in more children, two is enough. I decided to have fewer children, but, because my husband is interested in a boy child, I went to have sterilization, and after that I informed my husband, not before. (Uma, aged 29, 2 girls, 6 years of schooling, sterilized)

Initially I wanted more, a boy child, but now I am convinced that two is enough, and that boys and girls are equal, there is no difference. The economic condition restricts me. My husband felt that he wants a boy. I was sterilized before I told my husband because he was very interested in a boy. After the sterilization I informed him … the doctor also convinced me that in my community you have to give large dowry, so it is not good to give more number of girl children. (Jamuna, aged 31, 2 girls, 12 years of schooling, sterilized)

Not all families with two daughters were comfortable with only having two girls, however, and some were weighing their options and deciding whether or not to continue childbearing in an attempt to have a boy. A few women had continued to try for a boy. One of them explained:

We have three girls. I wanted to have two kids, but there is no boy, so we decided to have one more. We really wanted to have a boy, he can help in all the ways for the family. After our marriage we decided that one is enough. The first was a girl. If the first one was a boy maybe we would have stopped, but first was girl. So we had another, tried again, it was a girl, so again we tried. We wanted a boy. (Kalpana, aged 33, 3 girls, 9 years of schooling, sterilized)

After the third girl, this same woman was sterilized.

Carrying out fertility and sex preferences

There were three main methods by which respondents could manipulate the sex composition of their children: by limiting the number of children before the next pregnancy; by using sex-selective techniques during pregnancy to control the sex of the live birth; and by selective neglect and infanticide post-birth.

Since sterilization was not stigmatized and was commonly used by respondents, women spoke openly about how they used sterilization to stop childbearing dependent on the sex composition of their surviving children. This study did not ask outright about sex-selective abortion, female infanticide or neglect, due to concerns about cultural sensitivity, although we did ask about abortion in general. However, in conversations discussing sex preference, women spontaneously brought up that other women use sex-based abortions and female infanticide in order to avoid having unwanted female children:

I know a lot of women who had scans [ultrasounds] and found out that it was a girl and then they got an abortion, in private hospital. Lots of people do that, I don’t know exactly, no one will tell outside… (Amrita, aged 27, 2 girls, 7 years of schooling, sterilized)

Respondents who spontaneously brought up the use of female infanticide not only knew about the practice, but were able to share details of the methods used:

… if they don’t like females they are putting paddy (rice on the stalk) in the baby’s mouth and it will die. Another one, they don’t feed the baby for 2–3 days and it will die. Or they give herbal plants, milk of some plant to kill. I know my neighbours and also other villagers. In this area it is very popular for female infanticide. (Sanjita, aged 25, 1 boy, 3 girls, no schooling, sterilized)

If women have two girls and want a boy they will give the female baby ‘stone milk’¹ and then the baby will die. (Amrita, aged 27, 2 girls, 7 years of schooling, sterilized)

The quote that follows sums up the difficult decisions women in this region have to make. They are pressured to lower fertility by their family and society and by their own economic circumstance. They are told that infanticide and using ultrasound to determine the sex of the foetus and then having a sex-selective abortion are illegal. However, they live in a world in which there is sex preference and, therefore, must secure their position as wives by producing sons and especially by not producing too many daughters:

There is lots of outside pressure from midwives and such. They come and if they see that you have a large family they get on your case, saying ‘How dare you have so many children? Do you think you can raise all these children?’ There is a lot of pressure (because) there is the dowry. Also when you have too many girl babies there is female infanticide. If too many girls, there are too many marriages, and too much dowry problems. Female infanticide is very common. If you have two boys then its fine, but if have two girls, government is on people’s cases for killing girls. You shouldn’t do it. The government says that women should raise girls as if they were boys. Now, if you are caught (using female infanticide) you have to go to prison for three years. They ask, ‘Why are you killing girls, why are you doing this?’ Women are saying ‘Because we don’t have enough money, that’s why we are killing our babies’. (Gayathri, aged 35, 1 boy, 1 girl, no schooling, sterilized)

Discussion and conclusions

The past literature on son preference has often focused on the benefits of male children with little acknowledgement that there may be competing benefits and costs to having daughters. The present study in rural Tamil Nadu supports the findings of past enquiry elsewhere in India in which women voiced the benefits of having girls as well as boys and, thus, a preference for one child of each sex (Pande 2003). However, as was found here, more than one girl is worrisome due to the economic burden of daughters and higher birth order daughters are more likely to be discriminated against (Pande 2003, Mishra et al. 2004). The present study also found evidence for the obverse to son preference: namely, daughter aversion. Indeed it seems possible that, in this setting, daughter aversion only appears after a family has two girls and it may play a larger role in fertility decision-making than son preference.

If strong daughter aversion is a prime driver of fertility limitation, it may put downward pressure on population growth, possibly helping to explain the recent trend toward very low fertility rates in Tamil Nadu. Although the phenomenon of daughter aversion has received less attention than son preference, there are a few past studies on it, primarily in Northern India, some which suggest that it may differ by religious group (Khanna 1997, Borooah and Iyer 2004). Our work here supports the idea that daughter avoidance, often described inaccurately as son preference, is increasing in rural Tamil Nadu and suggests that economic constraints, primarily due to increased use of dowry, are the cause.

The increased prevalence of dowry in Tamil Nadu is noteworthy because it was not long ago that the inverse practice, bride price (grooms giving money to the wife’s family), was routinely practiced (Rao 1993). The bride’s family now must give a dowry in order to increase or even maintain the family’s status. Previous studies in southern India and elsewhere have posed several possible explanations for the increase in dowry, including a desire for economic mobility, changing marriage patterns, an attempt to mimic the behaviour of higher castes and the belief that dowry can purchase better grooms (Caldwell et al. 1983, Dyson and Moore 1983, Malhorthra et al. 1995, George 1997, Basu 1999, Amin and Suran 2005, Sheela and Audinarayana 2003, Srinivasan 2005).

What makes the rise in dowry most troubling is that it is often the poorest families in the lowest castes who cannot afford the high prices of dowries and who will, therefore, be most affected by having to provide multiple dowries if they have more than one daughter. Even more worrying is anecdotal evidence of dowry-related abuse, including burning and domestic violence (Amin and Suran 2005), to which respondents in this study also alluded. Children are already seen as expensive and, therefore, the added cost of dowry for girls, in addition to their lower earning power, makes having girls exceedingly troubling to rural, poor families. Much past research supports our argument that the main disincentive for having daughters in India is dowry (George 1997, Arnold et al. 1998, Basu 1999, Das Gupta et al. 2003, Srinivasna 2005). The present research has shown the potentially negative consequences of the rise in dowry and large demand that dowry puts on people in this region – increased daughter aversion.

With the two-child norm now prevalent in Tamil Nadu, families have to weigh their options carefully, not only in terms of how many sons to have, but how many daughters they can afford. Past research has noted that the change to a more masculine sex ratio can be credited to a combination of slightly uneven ratios at birth combined with small differences in mortality starting at an early age, which persist over time (Murthi et al. 1995, Das Gupta and Bhat 1997, Arnold et al. 1998). When modern sex-selective techniques are combined with more traditional practices of sex-selective abortion, female infanticide and female neglect, women and families have a means of controlling the gender of their children, while still attaining the smaller family size desired. According to National Family Health Survey data (International Institute for Population Sciences and ORC Macro 1993), Tamil Nadu has the highest rate of induced abortions of any Indian state. In recent years, there has also been an increased prevalence of mobile private sex determination clinics using ultrasound technology, which often also provide abortion services (George 1997). These mobile clinics display advertisements such as “Pay 50 Rupees now to save 50,000 Rupees later” (Basu 1999).

Although this study was limited in that we did not ask directly about sex-selective techniques or female infanticide, findings suggests that female infanticide is practiced by some women in the study area. That respondents themselves mentioned female infanticide during interviews suggests that these practices are known about. Interestingly, female infanticide was mentioned more frequently than sex selective abortion, which could suggest either greater taboo against sex selective abortion or that female infanticide is practiced more frequently. There is evidence that female neglect (giving female babies and small children less breast milk, food and healthcare) is common and that it heightens with parity (Das Gupta and Bhat, 1997, Clark 2000).

Study findings suggest that the changing gender preference will continue to increase the uneven gender ratio as families resort to these sex-selective practices, not only because of son preference but also because of daughter aversion. With technological advances, these practices will become more readily available and affordable to rural families. A greater understanding of this phenomenon can give insight into the future impact of dowry on sex preferences, sex selection and fertility decision making in this region on the whole. Understanding these trends is important in order to provide informed recommendations for government policy and programmes working to discourage the practice of dowry.