Preference or Aversion?
Exploring Fertility Desires among China's Young Urban Elite
Although the VOC approach has been criticised for being theoretically underdeveloped, it is useful as it draws attention to three broad aspects shaping fertility behaviour: 1) the value of children to parents, 2) alternative ways in which those values can be attained, and 3) the financial and emotional costs of having children, the last two aspects explaining not only low fertility but also why some individuals or couples want no children at all. The first eight values listed in Table 1 have subsequently been conceptualised as values related to psychological-emotional wellbeing, while the latter has been conceptualised as an economical-utilitarian value, such as children contributing to the family economy through child labour, household work and old-age support.
Of relevance to this article, the VOC approach has been further developed in two important ways. First, by linking structural level factors in explaining how values are shaped and reshaped, and how macro and micro levels interact. Macro-level factors of immediate relevance are welfare provision and labour market opportunities, which both can replace in particular economical-utilitarian values of children to parents. This in turn can explain preference for fewer children, but also why some individuals want no children at all. Second, the role of intergenerational relations needs to be taken into consideration in understanding both the psychological-emotional wellbeing and economical-utilitarian value of children. Although all values listed in Table 1 are relevant to intergenerational relations between parents and their young children, there are two values of immediate relevance to intergenerational relations between adult children and their ageing parents. The first one is 'morality' in terms of ancestor worship and passing on the family line, which implies a form of transfer from adult children to their parents by way of ensuring grand-children. The second one is 'economic-utility,' in terms of old-age care and support provided by adult children. These two values may take on a special meaning for the one-child generation, since they do not have a sibling to share the responsibilities of securing the lineage and providing old-age support. Previous research has identified that as societies modernise, the economical-utilitarian value of children gets subdued in favour of psychological-emotional wellbeing values. Parallel to this shift, there has been a change in intergenerational relations where flow of resources has reversed from younger to older generations to older to younger generations. Yet, in an Asian context, research has found that the middle generation still plays a great role in transferring resources to the older generation. Hence, the middle generation constitutes a 'sandwich generation,' as they are also expected to invest increasingly in the younger generation.
The best would be to have a boy and a girl, because I think a child growing up alone is too lonely. You need a companion.
An older brother and a younger sister would be best.
This is because an older brother would care for the sister from young age. He would feel very responsible. Like a real man. And the sister could rely on her older brother, and grow up like a little princess.
Yamei, singleton woman
Young adults in today's China differ from previous generations in two major ways. First of all, they belong to the so-called 'one-child generation.' Second, and relatedly, young adults face a marked shortage of women among same-age peers. These two characteristics prompt questions related to their fertility desires; have young adults about to reach childbearing age internalised the one-child norm, something for which previous studies have found evidence? Have they abandoned son preference? Or do they even prefer daughters to balance out the skewed sex ratio in China, as a response to official and media discourses emphasising the negative consequences of surplus men?
Despite the fact that young adults represent a unique group in China for the reasons stated above, and that understanding fertility preferences among young adults is important for gaining cues about a society's future, fertility desires of young adults in China is largely understudied. Moreover, studies on fertility preference in the general population mostly rely on demographic data, analysing number of and sex of children and progression to higher parities, especially in the context of sex preference. However, these studies offer little in terms of qualitative understandings of the social embeddedness of fertility preferences and behaviour.
In light of this, in this article I aim to further the sociological understanding of how fertility desires are being shaped among young adults in higher education in China. Focusing on this group is motivated by the fact that an increasingly large proportion of Chinese youth proceeds to university and presumably constitute the future of the growing Chinese middle class. Moreover, previous research shows that those with higher education are the first to adopt notions of sex indifference. Hence, exploring the social embeddedness of how sex preference or sex indifference is shaped in this group is of special interest. I focus on how fertility desires are shaped in relation to sexuality, gender and intergenerational relations, both within families, but also in relation to structural level factors. I will not attempt to generalise findings and claim representability, nor do I strive for predictability, acknowledging that the extent to which preferences will translate into future practices is an open question. Rather, the overall goal is to foster a deeper understanding of the processes that shape fertility preferences through a case study of young adults in higher education.
Theoretical perspectives on fertility desires
Studies on fertility behaviour almost exclusively rely on quantitative research. Many of these studies are founded in the value of children (VOC) to parents approach, first developed by Lois Wladis Hoffman and Martin Hoffman in the 1960s, an approach which is useful for understanding both sex preference and preferred number of children. Ethnographic studies focusing on fertility behaviour have typically highlighted fertility among marginalised or high-risk groups, such as teenage mothers, and urban poor. These studies have kept a critical stance to the VOC approach, criticising it for being grounded in rational choice theory, where individuals and couples weigh pros and cons, as if they were aware of all the options, and the consequences of these. Yet, ethnographic studies have not offered any theoretical frameworks that help analyse fertility behaviour and desires. Therefore, while being mindful of the critique of the VOC approach, and of the fact that transferring a theoretical framework based on quantitative data to a qualitative material can pose limitations, this article will draw inspiration from the VOC approach, as well as the ways in which it has been developed in recent research.
Value of children to parents
Based on quantitative cross-national surveys, the VOC approach identifies nine values of children to parents, as indicated in Table 1.
Table 1. Value of children approach
Source. Lois Wladis Hoffman, 'The value of children to parents and the decrease in family size,' Proceedings of the American Philosophical Society 119(6) (1975): 430–38, p. 431.
Sex preference is of relevance as it tends to lead to higher fertility among couples who prefer one or the other sex, an effect which tends to be larger in low fertility contexts. It is also of importance as it may lead to sex-selection and discrimination of the unwanted sex. Data from world fertility surveys show that in most countries couples prefer sex composition of both sexes, even where son preference prevails, with some notable exceptions in South and East Asia.
As proposed by Elisabeth Croll, son preference is reproduced in a culture of gender reasoning, captured in the Chinese device zhongnan qingnü (emphasising male and understating female), meaning that 'practically and cogitatively daughters are reasoned to be secondary and a supplement, but rarely a substitute for sons.' From this follows that sons are considered primary and unsubstitutable. Gender reasoning presupposes that parents (and grandparents) have an essentialist understanding of appropriate gender identities and of what sons and daughters can be and do. The practical implications of gender reasoning need to be understood in their specific context, and in the Chinese case sons are typically valued higher for their passing on the family line, ancestor worship, patrilocal marriage patterns implying that women leave their natal families upon marriage to reside with or nearby the husband's family, and old-age support. Hence, the different values identified in the VOC approach do not necessarily apply equally to sons and daughters.
Despite documented son preference, some Chinese parents have long since also expressed preference for daughters. Already in the 1980s studies established that some urban families prefer daughters for being perceived as more caring and filial. Studies based on both surveys and ethnographic data have also established that having both a daughter and a son is the ideal for a complete family. Vanessa Fong has even argued for a 'daughter empowerment' effect of the one-child policy, as daughters in the absence of brothers experience unprecedented access to family resources and engagement.
Whether sex preference will attenuate with socio-economic development is another question left unanswered. Hilke Brockmann cautioned against assuming that 'modernisation' attenuates sex preferences, and suggested that sex preference and type of welfare regimes are interlinked. Indeed, a recent study has shown that son preference intensified in urban China the 1990s, coinciding with the dismantlement of welfare institutions. Yet, more recent studies in a western context suggest that as women's status increases, sex preference will convert to sex indifference. Similarly, following the assumption that women's status is positively associated with more daughter preference, daughter preference may be stronger among the more highly educated.
Methods and data
The findings of this study are part of a broader research project which aims at investigating family relations, with a focus on gender and intergenerational relations, union formation strategies and intimate relationships among young adults in urban China in the context of a shortage of women. A case study approach was adopted and semi-structured in-depth interviews were carried out with twenty-five young adults enrolled in higher education in Beijing from February to March 2014. The informants were recruited in two phases; half were recruited through a lecture open to all students at the university, where I briefly introduced the project to the 150 or so students present and asked interested students to contact me through WeChat, a social media forum. The other half were identified through snow-balling through the first round of informants, where informants with particular characteristics with reference to dating history, marital status and sexual identity were actively looked for in order to enhance the variety of the sample.
Among the informants, thirteen were women and twelve were men, and the age ranged from nineteen to twenty-four years. Five originated from rural areas, and the rest from urban areas. Five had never been in an intimate relationship, six were in a relationship, one was married with a child, and the remaining were single, but had experienced being in a relationship. Most identified themselves as heterosexual, but two women and three men explained that they were bi-sexual. Among these, two had never had an intimate relationship, two had had only same-sex relations, and only one had had intimate relations with both sexes. Among the twenty-five informants, six had siblings. Informants without siblings will be referred to a 'singleton' in the empirical analysis. In order to ensure anonymity, all names are fictive.
The data was analysed using NVivo, where the transcribed interviews were coded and recoded in order to identify different themes and concepts, as well as similarities and differences within the sample. In addition to seeking out what themes emerged from the interviews, I also paid attention to what was not said, something which is important to analyse in a case study research design. In this regard, I paid special attention to the 'population problems' of too many people and too few women.
Below follows an account of the findings of the study, where I relate the findings to previous research and concepts introduced in the theory section above. I start with presenting motivations for wanting (or not) parenthood, and then discuss preferred number of children, as well as sex preference and sex aversion—the latter a concept which emerged during the data analysis. Before concluding the article, I will account for the gender reasoning behind wanting both a boy and a girl.
In many cultures, parenthood is a marker of social adulthood, and is associated with certain status and position in society and community. Particularly in Chinese culture, parenthood has been regarded as one of the indispensable features of filial piety, where failing to produce an offspring would be one of the greatest harms to the older generation, as reflected in the Confucian proverb bu xiao you san, wu hou wei da (There are three things which are unfilial, and to have no posterity is the greatest of them). Hence, in Confucian China, wanting parenthood was not primarily about wanting, but a matter of obligation. However, a shift in the view of childbearing has been identified. As opposed to the generation of older Chinese, previous research has found that the younger generation frames the wanting of children as both an act of filial reciprocity or 'natural obligation,' and as an act of self-fulfilment or a 'consciously chosen desire.'
During my interviews wanting parenthood was a general feature among almost all informants, but not all. Four aspects emerged which were important for shaping aspirations for parenthood; social adulthood, intergenerational obligations, meaning of life and the expansion of the self. Indeed, parenthood was associated with adulthood and a marker of adult status and social identity, as predicted by the VOC approach. However, the value of children was also intertwined with the value of grandchildren and wanting parenthood related to fulfilling intergenerational expectations and obligations by ensuring grandchildren. Guoda has a younger brother but does not feel that the responsibility to continue the family line could be passed on to his brother. Rather, he explained how he has internalised his parents' wishes for him to have children:
Before I did not want to have children, but then I thought it is not very realistic in China.
There are many reasons, parents is one. Parents will for sure say that you need to have a child. Like my parents they will for sure think that I need to have not only one but two children. They will want me to have two children. One child is not enough.
I don't feel desperate for it, but it may change with age, for the moment it is not urgent.
Following my parents' wishes I want two. I feel like this because I have been subject to their wishes, so I also want to have two children.
Although the informants were below the mean age of first child, and therefore did not perceive any pressure to have children in the immediate future, several informants described how older siblings and cousins who were childless experienced pressure from their parents and other relatives. The pressure is particularly hard to juggle by bisexual persons, who are not legally entitled to marry or have children with someone of the same sex, and who may need to incur the 'cost' of not forming an intimate relationship with someone of the same sex in order to attain parenthood. The fact that parenthood is contingent on heterosexual marriage may explain why such a high number of homosexual men are married. In fact, all of the bi-sexual informants in this study believed parenthood was important, or at least that not wanting parenthood was problematic. Peng, a singleton man explained:
To me right now, to have a family is not that important because maybe, the possibility is not that big but maybe I won't get married–I think that's ok, but not ok to my family. We have a proverb saying 'There are three things which are unfilial, and to have no posterity is the greatest of them.' I'm not worried about not having children, I just think if you don't get married then you won't have children, and my parents will be very upset.
Wanting parenthood was also associated with meaning of life and the expansion of the self. Tingting, whose only love relationship has been with another woman, wants to get married with a man to have children, since meaning of life, having children, the notion of a family and heterosexual marriage are all inseparable from one another: 'I think a family living together is when a life is a real life. And I really want to have children.
I always wanted to get married and have children.' Meaning of life and the expansion of the self also had a gender dimension, where motherhood was felt as a 'natural calling.' Such an expansion of the self was not grounded in a 'natural obligation' to extend the family line as found by Evans, but rather regarded as an expansion of the female self. Feiyi, a single woman eager to marry and have children in a not too distant future, explains: 'I think that for women, compared to men, the most outstanding feature is that she can give birth to a child. I would not want to give up this right.' This finding echoes previous studies identifying parenthood as particularly important to women.
However, it should be noted that among the twenty-five young adults interviewed two women questioned whether they wanted children, while none of the male informants expressed such resistance, suggesting that fatherhood may be intrinsic to Chinese men. The reason for not wanting parenthood was rooted in a reluctance to get married, suggesting again the inseparability of marriage and childbearing, which indeed is also legal matter. Hence, the value of children did not compensate for the anticipated social cost of marriage. Worth noting is that social costs—not emotional or financial costs—were stated as reasons for not wanting any children. The social costs evolved around giving up freedom and apprehension about having a life-long commitment to someone, suggesting an understanding of the family as a life-long static unit, as indicated by Ruixiang, a singleton woman: 'Maybe forming a family will be better somehow. But I find it very hard to imagine having children. For now I tell my mother that I don't want children. And my mother says it's up to me. I think the thought of living together with the same person a whole life is very scary.' Similarly, Wenhui, another singleton woman, expressed that forming a family would entail loss of freedom: 'If you get married, you choose to become a partner with someone, and then you have to shoulder his whole social network, his social relations with his family, his friends and his colleagues. So you don't get married with only him, but you get married to his entire social network.' Hence, since marriage is a passage to parenthood, and since marriage is associated with losing freedom, some women were not keen on parenthood. This way of reasoning adds a cue as to why an increasingly large proportion of women in other parts of East Asia delay or forego marriage and childbearing, although far from all do so intentionally. Wenhui also explained that lack of parenting skills should be a reason for not wanting parenthood, even though in her view many couples become parents without having the appropriate skills.
In sum, wanting parenthood was not a unanimous wish among the informants, and there were diverting motivations, as well as exemptions. Intergenerational relations are important for shaping the intention for wanting parenthood, but not for all. Even though marrying and having children as a filial obligation may become intensified among the one-child generation, as those young adults do not have any siblings to share that obligation with, the findings of this study suggest that intergenerational relations are transforming in ways that allow the wanting of parenthood to be not only a merger of desire and obligation, but indeed a pure desire in its own right that is detached from intergenerational expectations and obligations, which—as illustrated in the case of Ruixiang—opens up to not wanting parenthood at all.
More than one child
Previous research has found that the one-child norm has been largely internalised in the population segment that this study is concerned with. However, interview data for this study generated a different picture. In fact, among the informants who wanted children, all but one wanted more than one child. Themes of loneliness and lack of a companion were often stated as motivations for wanting more than one child, which seems to suggest there is a social feedback loop among the one-child generation, where they transfer their own experience as singleton children to their own fertility desires. Hence preference for the number of children was not only rooted in the perceived value of children to parents, but also the perceived value of having siblings for the children themselves.
Preferred number of children is not only shaped in relation to the perceived values of children, but also associated costs, as suggested by the VOC approach. One important theme that emerged was that of affordability. Rihan, a singleton man, explained: 'I want one or two if I can afford it. Two may be nice, they can support each other. But I don't think I can afford that because nowadays keeping a baby is very expensive, to pay for school and hospitals. If I have a lot of money, I want to have two children.'
Feiyi, who, as quoted above, sees parenthood as an extension of the female self, shares the importance of companionship and affordability as two guiding principles in shaping preference for number of children, but added a third: to be enough as a parent.
Two children, like me and my little brother. I think like this it will not be so lonely. Three would also be too much, and for sure there would be issues of affordability. Like my aunts on my mother's side they all have three children and it is very tight.
there is always somebody who gets too little attention, [and have the feeling] that mum and dad don't love you enough. But with two it is easier to strike a balance.
In fact, not being enough as a parent and lack of 'parenting skills' was not only deterring young adults from having children, as Wenhui explained above, but also informed the preference for the number of children, indicating that being a good parent does not only require time and financial resources. Yet, the main motive for wanting more than one child was the need for companionship between siblings. However, it was also informed by the wish to have both a boy and a girl.
Sex preference or sex indifference?
When asked about sex preference of children about half of the informants expressed sex indifference, although most of the informants who wanted two children said that having a girl and a boy would be ideal, as discussed further on in the article. Among women who expressed sex indifference, an interesting pattern emerged. Several had grown up in a son preference context, experiencing being valued less by in particular grandparents. This lesser value was rooted in patrilineality and the notion that raising a (grand)daughter is like 'raising someone else's child.' Yet, this experience had served as a springboard for some female informants to 'prove' that sons are not unsubstitutable. Juhua, a singleton woman, explained:
Both my grandfather and grandmother were favouring my male cousins. In my childhood, it was a struggle to get by. Both my parents had just started to work, and at that time my grandmother lived with us. She was not very good to me. She would not wash my clothes, she would not make me a cotton-padded jacket, but she let my mother do everything. During the first month (yuezi) my mother had to clean my diapers, and my grandmother would not look after me. After two months my grandmother left and went back to the village. She told my parents that they should give me away and things like that
but my parents did not agree.
Reflecting on the experience of growing up in a son-preference context, Juhua believes she has grown stronger, and experienced what Fong describes as daughter empowerment:
When I was young I had to do everything very well, including in high school I studied two years of Taekwondo. I was thinking that if I did not have the same physical and intellectual capacities as a boy I would subconsciously think that I did not gain the same respect from other people. It affected me, but it was in a positive way. It made me chase an even more complete and fulfilling life, and made me perform even better.
It was a bit as if I wanted to compete with boys, wanted to be even better than boys, regarding both physical strength and knowledge.
Expressing sex indifference, Meimei, also a singleton woman, likewise exemplified how sons can be substitutable. Initiated by her father, who wanted to show respect towards his father-in-law, who had no son, Meimei had taken the surname of her mother. The name change suggested that both sons and daughters are able to pass on the family line. However, it should be noted that Meimei had three paternal uncles, who had already secured the patriline on her father's side.
As the cases of Juhua and Meimei both illustrate, parents of daughters had actively defied norms and practices of son preference in their families. Just like Evans in her study of mother-daughter relationships in China found, mothers and sisters were important sources of inspiration in the context of son preference, where gender disadvantage could actually function as a source of inspiration and motivation for such defiance. The current study found that also fathers are instrumental in providing such inspiration and motivation. Yet, there were several accounts of how young adults expressed sex preference, attributing certain values to sons and others to daughters. However, it is important to note that neither of the informants said they would be ready to terminate a pregnancy, but stated alternative medicine and continuing childbirth as two means to achieve desired sex composition of future children.
The value of sons
Values associated with son preference in the VOC approach are typically 'morality: religion, altruism, good of the group,' 'power, influence, effectance,' and 'economic-utility.' In the literature on China, it is well documented that these values are rooted in the practices of ancestor worship and passing on the patriline, as well as patrilocality and old-age support. An aspect that is less prominent in the literature is that of 'power, influence, effectance,' although it has been argued that the absence of sons in rural China affects the power and standing of families negatively.
Among the participants in this study, there were indeed expressions of son preference, aligning with previous research. However, in terms of the value of children related to 'morality,' the value of passing on the patriline (i.e. the good of the group) was more important than religion, as accounts of ancestor worship were absent from the interviews, although the two are interlinked. Bomei, who plans to marry her current boyfriend, explained how her future husband reasons:
My boyfriend prefers boys. He thinks that only boys can pass down his genes. He is influenced by traditional China a lot. In the traditional society, people think that only boys can extend the family line
His [the boyfriend's] father is the first son of his grandfather, he [the boyfriend] is the only boy in his family so he is influenced by this thought a lot.
If we can have two children, my boyfriend prefers a boy and a girl. But if we can only have one, he prefers a boy.
The value of sons was also associated with social standing and power, at least in rural settings. Kaiwen, a married man with a young daughter explains how his parents will have some regret not having a grandson, but that they will not show it:
They will still love this little baby, but in their heart they would rather want to have a grandson.
because where I come from [rural China] everybody for sure wants to have a son, they have son preference. From the talk of the neighbours you can sense that they rather wished for this child to be a boy. Even though they don't say so, you can sense that everybody thinks that having a son is a very good thing.
It's about old age support and the social standing in the village. Of course, a son can give you more things. So they want it for themselves. For example, if that family has five six sons, and these sons grow up, the family will gain certain power in the village. People don't dare to look down upon you. But if you have five six daughters, once the daughters get married, you are left on your own.
As this quotation implies, power and influence is closely linked with patrilocal marriage. Moreover, just like the experience of growing up in a son preference context can trigger sex indifference, such an experience can also reinforce the higher value of sons. Expressing son preference, Caohua explained how her parents lived with her brother, while they left her and her sister with the grandparents:
I can feel that they think my brother is much more important and want to educate him for themselves.
Because my father is the only son of my grandfather and he thinks only the boy can inherit something from them. When we marry to someone, we are not the member of the family.
My grandmother thinks it is very important to have my little brother
Even among individuals with sex indifference, gender reasoning and an essentialist understanding of what sons and daughters can be and do was evident, as exemplified by Meimei, who had changed to using her mother's surname: 'I feel boys can mantel the responsibility for the future of the family. Men are a bit more powerful. Maybe when you are older sons are more capable. They can deal with misfortune.' This suggests that there is an understanding that men are intrinsically more powerful—a notion which conceals gendered power structures that privilege men.
From the interviews it was evident that passing on the family line was only identified as a cue for son preference in the context of a man being the only son of a brother-less father. Other factors known for upholding son preference such as old age care and ancestor worship were not deemed important.
The value of daughters
Another well-established motivation for wanting children is that they bring values associated with 'primary group ties, affection' and 'stimulation, novelty, fun.' In the present study, those values were particularly associated with daughters. During the interviews, motives for wanting daughters revolved around daughters being 'lovely,' 'cute,' 'funny,' 'gentle' and 'soft.' Another theme associated with the value of daughters related to girls being perceived as more filial as in 'caring,' paying more attention to primary group ties, as expressed by Meimei: 'Girls are more intimate, and would keep you company often. A daughter may come home more often.'
Hanri, a singleton man, explained how his parents also have ideas about sex preference informed by the value and the experience of raising a (naughty) son:
My dad says he wants a granddaughter more and my mother has not expressed any wishes. He says daughters listen better, because when I was young I was very naughty. So my parents had a lot of headache because of me. My mother has said that if I have a daughter she can teach her how to dress up.
My mother would be able to accept both a girl and a boy, but my father rather wants a granddaughter.
Personally, I don't mind. I am like my mother.
The quotation suggests that some grandparents extend their fertility preference across generations to their own children, in order to have a child and a grandchild of different sexes, pointing at the importance of the value of grandchildren to grandparents. This may also explain why participants in this study who expressed sex preference mostly wished to have a child of the opposite sex to themselves.
The study did not find that daughter preference would stem from a shortage of women and that daughters would be valued higher as a consequence of the potential threat of a marriage squeeze for men. Worth noting is that this was not due to the informants believing they were 'immune' to the marriage squeeze due to their higher education status.
Sex preference was not always informed by the preference for daughters or sons, but rather the aversion against one or the other sex. Sex aversion is something which has been documented in India in the context of high dowries, which has instigated daughter aversion. Sex aversion is linked with concerns over social mobility both in terms of the parents and the children themselves.
Son aversion: The cost of children
Due to norms and expectation of patrilocality, a common practice in China is that young men and their parents provide housing when a young couple is about to marry. Not being able to provide housing can result in men not being considered a good match from a prospective wife and her family. Therefore, raising a son and ensuring his marriage incurs a heavy burden for parents, something which emerged as a main theme instigating son aversion. This is particularly so for the 'sandwich generation,' who in addition to investing in their children's future also mantel the sole responsibility for supporting their own parents at old age. This was exemplified by Andong, a singleton man planning to marry his girlfriend: 'My girlfriend wants a daughter, so we don't have to prepare for her house. If we have a son, we have to buy him a house.' Similarly, even though some informants expressed sex indifference, having a daughter was acknowledged as an advantage: 'Maybe daughter is fine because if both are boys, maybe we have to buy flats for them, which may be a burden. If we have a daughter, we don't need to spend that much money on her,' said Rihan, who as quoted above wants two children but thinks he can afford only one. Likewise, Feiyi acknowledged the burden placed on sons and their families: 'Sometimes I think of my brother and that it won't be so easy for him to get married.
If my brother wants to get married one day, our family needs to provide a flat and a car. Those are the minimum you need to provide, otherwise they [the bride's family] will not agree to the marriage.' Following a pattern of gender reasoning, quotations like this hint at a muted form of patrilocality in the sense that parents are only expected to provide for housing for their sons, and daughters are expected to 'marry out' and live with her husband.
Another theme informing son aversion was the perceived inability of sons to show affection and enter primary group ties in meaningful ways. Here, the ascribed competitive character of boys also informed son aversion, as explained by Tongxin, who wants three children, but not if they were all boys: 'I think three boys would be very noisy.
In the future they will all go their own way, and I think that maybe there would be disharmony and conflict between them.' Similarly, Bomei, who is in a relationship with a man who wants at least one son as accounted for above, explains why she attributes more value to girls: 'Some boys don't communicate, especially when they are old. Boys usually don't like to stay too close to their mother, maybe they just aren't good at communicating. They just feel uneasy when it comes to expressing themselves with their mother, especially when they are old, so I like girls more.'
Son aversion also stemmed from the expectations on men in terms of having a career and being successful in society. Nini, a singleton woman explained: 'Women have the advantage that if they lose their job, they can return to the home. If a man loses a job and returns home society will think he is a failure. We have more leeway.' Likewise, although expressing sex indifference, Qiuhua, thinks there are societal expectations on boys, and that gender essentialism has grown on her: 'From inside I predict there's no difference, but the society just changed me to say there are some differences. I've become more practical, for example I believe a boy has to be successful and it's better for him to have money.
For daughters there are no special notions.'
Hence, informants testified to son aversion stemming from the idea that men need to provide housing, something which is becoming increasingly expensive in China, in order to be able to attract a bride. Along the same lines, daughter preference was motivated by women having less pressure to perform outside of their future maternal/home-maker roles. Common for these ways of reasoning was the expectation that men are main providers, both in terms of providing housing and generating income, and that women get access to these resources through marriage, where hypergamy is a main route for social mobility for women.
Daughter aversion: The cost for children
Whereas son aversion was rooted in costs associated with providing housing to the newly wed, costs which are often incurred by the parents, daughter aversion was more associated with the cost for daughters themselves in terms of the hardship they may encounter in society. Disadvantages for daughters were understood to be both structural and individual, the latter something which notably only female informants expressed. Caohua, who prefers sons, and who has grown up in a son preference context, as elaborated on above, feels women as individuals are weaker, which is based on her personal experience and that of her female friends. Because of this weakness she believes sons grow up happier than daughters:
I think girls are disadvantaged in many aspects. We are sometimes too fragile. I think we are weak both mentally and physically.
We are more limited than them.
I think girls are afraid of many things like walking in the dark or going out alone, and for being attacked. But men can do more things. They don't have to think that much. They just have to do what they want to do.
I think they [my children] will be happier as boys.
The weaker character of girls was also reiterated by Yamei: 'If boys encounter some frustrations in their lives they can deal with it by themselves. With girls you have to worry a lot.' Sharing the sentiment that girls are weaker, Bomei stated risk of sexual harassment as a main reason. However, when asked if there are any other aspects that makes girls more vulnerable, Bomei explained how that feeling is generic and has been transmitted from her mother and media, although she herself does not have personal experiences of either sexual harassment or feeling weaker as a girl:
When I was young my mother always told me that [girls are weaker], but actually from very young to now I don't experience that I am less competitive than boys, but I just have that feeling. Thinking of girls I think of vulnerable images, maybe because girls are weaker, thinner, less strong, confronted with sexual harassment, maybe we are less powerful. As for other aspects, I can't think of any specific examples.
Although not framed as such by the informants, sexual violence can be understood as a structural level factor that instigates daughter aversion. Gender inequality was another structural level factor underpinning daughter aversion, as identified by Nini, who wants one boy and one girl, but two boys if she could only have two children of the same sex:
And there is a bit of gender inequality. Today in society, boys have a bit of an advantage for work. Also in school it used to be like that. The teachers in both lower and upper middle school had a saying like girls study until they can't study anymore, while boys who are clever get more energy the more they study. Before, we had a female head teacher. She was very fond of boys. Boys could be as naughty as they wished and she was still happy. She said male students were clever and good in school. And then she was excluding towards female students.
Worth noting is that gender discrimination at work and in school was not only mentioned by women, but also by men. Son preference was hence partly motivated by an understanding that women as individuals are weaker and partly motivated by structural level gender discrimination-induced daughter aversion.
Gendered reasoning for wanting an older boy and a younger girl
Another theme that emerged from the interview material was the importance attached to having two children of the opposite sex, where the older child was a son and the younger a daughter, as told by Bomei:
It will be more than wonderful if the boy is older than the girl. I think the brother usually takes care of his little sister and the sister will be protected. Maybe she [the sister] will learn something about boys from her brother. When they [grown-up daughters] are in love, they will have an unbiased view towards the relationship. They [girls] will understand boys more, they will become a wonderful girlfriend and then a good wife. As for the boy, learning how to take care of the girl from very young age is good for them to have responsibility, for them to take responsibility for the family. So I'd like to have a boy and a girl, and have them grow up together, I think they will learn a lot from the process.
The quotation implies that the family is an important unit to foster gender identities and to shape 'real men' and 'real women' in heteronormative ways. The preference for an older son and a younger daughter also stems from personal experience of growing up as a girl and the need for protection, as described by Tingting: '[as a child] I wanted the feeling that wherever I go there is someone to protect me. If I have a daughter she would for sure wish to have an older brother the same way I did.
then there would be no one to bully her.
I think little girls should have someone they can depend upon.'
There were also reasons for wanting an older son and a younger daughter that stemmed from some kind of awareness of unequal gender relations. In fact, Yamei, who was quoted in the introduction to this article, and seemingly had a gender stereotypical view on girls as princesses and in need of protection, also sees the danger of the 'bigger-sister-syndrome':
If it is an older sister and a younger brother she may be forced to take some responsibility. She would maybe need to mantel some of the responsibility a male would take. This may not be fair to her. And the little brother may also be spoiled to pieces, so it would not be ideal. My uncle has two older sisters and they have been very good to him and now he has no sense of responsibility. He just cares about his own happiness. This way of upbringing is also not good for boys. So I think it is better to have a boy first.
She [an older daughter] may slowly but surely learn how to care for a guy. Then maybe when she grows older she will have a personality that is prone to looking after her husband. And maybe her partner will have a lot of bad habits that she needs to fix, and she may endure all this in silence. That may be pretty hard.
Interestingly, preference for an older son and a younger daughter was something which only women expressed. Moreover, reasons ranged from being able to foster the most 'conducive' gender identities, to dealing with being bullied as a little girl, to avoiding fostering daughters into responsible women and sons into irresponsible men, traits which were considered incompatible with appropriate gender identities.
Drawing theoretically upon the Value of Children (VOC) approach, as well as some of its critique and refinement, in this qualitative case study I have explored fertility desires among one group of China's young urban elite, namely young adults in higher education. I have identified aspects shaping fertility preferences beyond what is encompassed in the VOC approach. I have also shown how gender and generation play important roles in shaping fertility preferences, and that the interlinkage between individual and structural level factors are important to take into account in order to get a more nuanced understanding as to how and why that is so. This goes to show that qualitative research can contribute more complex understandings, although the small sample means that findings must not be generalised beyond the present case.
I found that intergenerational relations are important for shaping young adults' wishes about becoming parents, and also for reinforcing a two-child norm. Hence, wanting parenthood is not only informed by the 'value of children' but also the 'value of grandchildren,' suggesting a merger of both individual desire and intergenerational obligation to have children. Attaining the value of children and grandchildren is particularly challenging to homosexual persons. Although the bisexual informants did not express it so explicitly, it is plausible that their homosexuality was suppressed in favour of bisexuality due to their wish to have children, something which requires heterosexual marriage since both same-sex marriage and out-of-wedlock births are illegal in China. Yet, the study also identified reluctance to parenthood (although not linked to sexuality). This suggests that intergenerational relations are transforming in ways that allow the wanting of parenthood to be not only a merger of desire and obligation, but indeed a desire detached from intergenerational obligations. An additional important finding is that economic-utility as a core value identified in the VOC approach was missing in the empirical material. This can be interpreted as a sign that the value of both children and grandchildren is shifting, where psychological-emotional wellbeing is becoming more important, at least among the (future) urban elite.
In terms of sex preference, a major contribution of this study is that it does not have to be informed by an explicit wish to have a daughter or a son, but can be shaped in relation to anticipated 'costs' or 'lack of value' associated with either sex. This is captured in the term sex aversion. Figure 1 summarises the main themes which emerged from the empirical data analysis, with reference to preference, aversion and (dis)incentives for having children in general.
Figure 1. Themes generated through the empirical analysis compared to the VOC approach*
Source. Author generated.
As Figure 1 illustrates, gender reasoning instigates sex aversion and sex preference in essentialistic and dialectic ways in relation to both individual and structural factors. At the individual level, daughter preference is motivated by the understanding that daughters are more caring, while son aversion is based on the perception that sons are less caring. Likewise, son preference is underpinned by the notion that men are strong and able, while daughter aversion is rooted in the idea that women are weak and vulnerable. Interestingly, women expressed a more internalised view of male domination, which may explain why the ideal of having an older son and a younger daughter was expressed by female informants only. At a more structural level, daughter preference is based on the idea that women have more leeway in life, while son aversion is related to men having too much pressure as main providers in a highly competitive economy.
Illustrating the importance of micro-macro interlinkages, son aversion is also related to the financial cost parents need to incur in providing for sons' housing. Families in today's China are expected to invest in the housing of sons to enable them to attract brides, something which has been intensified by the shortage of women and the so called male marriage squeeze. Hence, even though the shortage of women among young adults in today's China was not identified as a direct cause for daughter preference, the fact that the marriage squeeze intensifies the need for parents to invest in son's housing instigates son aversion. This indirectly causes daughter preference as daughters can gain in social mobility through marriage without much investment from parents.
Another important finding points to the qualitative difference between son and daughter aversion. While son aversion is mostly associated with financial costs for parents or lack of value for parents, daughter aversion is more associated with emotional costs (like worry) for parents as well as costs for daughters themselves, partly due to structural level factors pertaining to gender discrimination in school and on the labour market, as well as sexual harassment. Hence, although the VOC approach rightly draws attention to costs in understanding fertility behaviour, this study finds that costs extend beyond parents to (future) children.
Comparing the findings of this qualitative inquiry with the VOC approach, it was evident that two values, namely 'Achievement, competence, creativity' and 'Social comparison, competition' were not considered important in informing the value of children, although it can be argued that these values overlap with social adulthood. Also, missing from the VOC approach was the importance of sibling companionship as a cue to wanting more than one child; there seem to be a social feedback loop among the one-child generation, which may explain why they have not internalised the one-child norm to the extent previous studies have argued. The fact that most informants expressed a two-child norm suggests that China's population size is not a key issue for them on a personal level. Likewise, the 'problem' of sex ratio imbalance did not inform daughter preference or son aversion. This potentially suggests that population ageing and the well-being of their own future families override concerns of 'over population' and sex ratio imbalance. Rather, financial constraints were identified as disincentives for wanting more than one child. Moreover, the study identified not only financial and emotional costs of parents, as per the VOC approach, but also the social cost of marriage (for women) as reason for not wanting parenthood.
As illustrated in Figure 1, being 'lovely, cute and fun' was associated with the value of daughters only, while 'power and influence,' as well as 'passing on the family line' were associated with the value of sons only. However, passing on the family line was identified as a cue for son preference only in the context of a man being the sole carrier of the patriline. Moreover, some grandparents extend their sex preference across generations, and having a grandchild of the opposite sex of their grown child can make up for the 'missed opportunity' of having raised both a daughter and a son. This hints at the importance of sex composition of in-family generations as a cue for sex preference. The study also found that sex preference is formed in relation to personal life histories and trajectories, where informants testified to the experience of growing up in a son preference context as both instigating sex indifference and reproducing the higher values of sons. This finding concurs with Evan's study, but—in addition to the role of mothers—shows the importance of fathers in transmitting or breaking with son preference, such as bringing daughters into the function of passing on the lineage.
In terms of methodological considerations, it is important to view the findings of the study in light of the fact that the informants are at an early life-stage. For example, the lack of accounts about ancestor worship may be related to the informants' young age. Also, the case represents young adults belonging to the (future) middle class, which may explain why economic-utilitarian values of children and concerns over old-age support were absent from the empirical data. Moreover, even if fertility desires remain the same they may not translate into practice, and findings should not be interpreted as predicting future fertility behaviour. Rather, more studies investigating fertility desires among the one-child generation taking into account not only gender, sexuality and generation, but also class are needed to find out if—and if so why and how—fertility preferences will translate into practices.
The research on which this study is based was carried out during a Postdoctoral fellowship generously sponsored by the Department of Sociology at Lund University. I gratefully acknowledge the Linneus-Palme Foundation enabling the field work financially.
 While the estimated total fertility rate (TFR) was 5.81 in 1970 and fell to 2.75 in the late 1970s, it was reduced further through the so called one-child policy of 1979, to 2.17 in 1990 and 1.55 in 2000. See Susan Greenhalgh and Edwin Winckler, Governing China's Population: From Leninist to Neoliberal Biopolitics, Stanford: Stanford University Press, 2005.
 Since the mid-1980s, as sex-selective technology became widely available and number of children restricted through the one-child policy, sex-selective abortion became a means to de-select daughters in favour of sons. Sex ratio at birth increased from 111.3 in 1990 to 116.9 in 2000 and 117.9 in 2010 according to census data. See further: Lisa Eklund, 'Rethinking son preference – Gender, population dynamics and social change in the People's Republic of China,' Ph.D. dissertation, Lund University, 2011.
 While Greenhalgh and Winckler make a general claim for the internalisation of the one-child norm as closely intertwined with notions of 'modernity'; others have found that the one-child norm is adopted mostly in urban areas, where the policy was strict. See Greenhalgh and Winckler, Governing China's Population; cf. Giovanna Merli and Herbert L. Smith, 'Has the Chinese family planning program been successful in changing fertility preferences? Evidence from linked records in four counties in rural northern China,' Demography 39(3) (2002): 557–72.
 The Chinese state propagates the risk of surplus men, such as the marriage squeeze, to stop couples from de-selecting daughters, see: Lisa Eklund, '"Good citizens prefer daughters": Gender, rurality and the care for girls campaign,' in Women, Gender and Development in Rural China, ed. Tamara Jacka and Sally Sargeson, pp. 124–42, Cheltenham: Edward Elgar, 2011.
 Gisela Trommsdorff and Bernhard Nauck, 'Value of children: A concept for better understanding cross-cultural variations in fertility behavior and intergenerational relationships,' Journal of Cross-Cultural Psychology 41(5/6) (2010): 637–51.
 Gross enrolment in tertiary education was 29 percent for females and 26 percent for males in 2012 according to 'Education: Gross enrolment ratio by level of education,' UNESCO (accessed 2 March 2016).
 Karsten Hank and Hans-Peter Kohler, 'Sex preferences for children revisited: New evidence from Germany.' Population 58(1) (2003): 131–43.
 Lois Wladis Hoffman, 'The value of children to parents and the decrease in family size,' Proceedings of the American Philosophical Society 119(6) (1975): 430–38.
 Jacqueline Adams, 'Why do women want children? The case of Hong Kong, China: A lowest-low fertility context,' Marriage & Family Review, just accepted, 2016.
 Hank and Kohler, 'Sex preferences for children revisited'; Elizabeth Thomson, 'Value of children,' in International Encyclopedia of the Social & Behavioral Sciences, ed. Neil J. Smelser, Paul B. Baltes, vol. 8, pp. 1725–729.Amsterdam, 2002.
 Trommsdorff and Nauck, 'Value of children.'
 Trommsdorff and Nauck, 'Value of children.'
 Esther CL Goh, Bill YP Tsang and Srinivasan Chokkanathan, 'Intergenerational reciprocity reconsidered: The honour and burden of grandparenting in urban China,' this issue.
 Lisa Eklund, 'Marriage squeeze and mate selection – Analysing the ecology of choice and implications for social policy in China,' Economic and Political Weekly 48(35) (2013): 62–69.
 John C Caldwell, 'Toward a restatement of demographic transition theory.' Population and Development Review 2(3/4) (1976): 321–66.
 Elisabeth Croll, 'The intergenerational contract in the changing Asian family, Oxford Development Studies 34(4) (2006): 473–91; Kristina Göransson, The Binding Tie: Chinese Intergenerational Relations in Modern Singapore, Honolulu: University of Hawai'i Press, 2009.
 Hank and Kohler, 'Sex preferences for children revisited'; Lisa Eklund, 'Modeling prenatal son preference: Some new perspectives from China,' Population Review 50(2) (2011): 47–66.
 See for example Elisabeth Croll, Endangered Daughters: Discrimination and Development in Asia, London and New York: Routledge, 2000.
 Hank and Kohler, 'Sex preferences for children revisited.'
 Fred Arnold and Eddie CY Kuo, 'The value of daughters and sons: A comparative study of the gender preferences of parents,' Journal of Comparative Family Studies 15(2)(1984): 299–318.
 Croll, Endangered Daughters, p. 132
 Croll, Endangered Daughters.
 Arnold and Kuo, 'The value of daughters and sons.'
 Elisabeth Croll, Chinese Women Since Mao, Armonk: M.E. Sharpe, 1984; Margery Wolf, Revolution Postponed: Women in Contemporary China, Stanford: Stanford University Press, 1985.
 Thomas Scharping, Birth Control in China: Population Policy and Demographic Development. London and New York: Routledge Curzon, 2003; Kay Ann Johnson, Wanting a Daughter, Needing a Son: Abandonment, Adoption, and Orphanage Care in China, St Paul, MN: Yeong & Yeong, 2004.
 Vanessa L. Fong, 'China's one-child policy and the empowerment of urban daughters,' American Anthropologist 104(4) (2002): 1098–109.
 Hilke Brockmann, 'Girls preferred? Changing patterns of sex preferences in the two German states,' European Sociological Review 17(2) (2001): 189–202.
 Eklund, 'Modeling Prenatal son preference.'
 Michael S Pollard and Philip Morgan, 'Emerging parental gender indifference? Sex composition of children and the third birth,' American Sociological Review 67(4) (2002): 600–613.
 Hank and Kohler, 'Sex preferences for children revisited.'
 Alan Bryman, Social Research Methods, Oxford: Oxford University Press, 2008.
 Hoffman, 'The value of children to parents.'
 Song Zhifang, 'Flow into eternity: Patriarchy, marriage and socialism in a North China Village,' Ph.D. dissertation, ProQuest, 2008, p. 15.
 Harriet Evans, The Subject of Gender: Daughters and Mothers in Urban China, Lanham, MD: Rowman & Littlefield Publishers, 2007.
 Evans, The Subject of Gender, p. 179.
 Median age at first birth for women was 24.04 in 2010, a modest increase from 23.52 in 2000. See Fu Chonghui, Zhang Linghua and Li Yuzhu, 'Characteristics of the changes of population fertility in China based on the sixth census,' Statistical Research 30(1) (2013): 68–75, p. 71.
 Yang, Xueyan, Isabelle Attané, Shuzhuo Li and Bo Yang, 'Same-Sex sexual behaviors among male migrants in a context of male "marriage squeeze" results from an exploratory survey in urban Xi'an, China,' American Journal of Men's Health 6(6) (2012): 485–96.
 Evans, The Subject of Gender.
 Mardy S Ireland, Reconceiving Women: Separating Motherhood from Female Identity, New York: Guilford Press, 1993; E Nakano Glenn, 'Social constructions of mothering: A thematic overview,' in Mothering: Ideology, Experience, and Agency, ed. Evelyn Nakano Glenn, Grace Chang, Linda Rennie Forcey, pp. 1–31, London: Routledge, 1994; Adams, 'Why do women want children?'
 As also suggested by Xuan Li, 'The "nursing dad"? Constructs of fatherhood in Chinese popular media,' this issue.
 According to Chinese law, only married couples are allowed to have children, Eklund, 'Marriage squeeze and mate selection.'
 'Asian demography: The flight from marriage,' The Economist, 20 August 2011 (accessed 26 January 2016).
 Eklund, 'Marriage squeeze and mate selection.'
 As found by Evans, The Subject of Gender.
 Greenhalgh and Winckler, Governing China's Population; Merli and Smith, 'Has the Chinese family planning program been successful in changing fertility preferences?'
 As also found by Fu, Zhang and Li, 'Characteristics of the changes of population fertility.'
 Yuezi refers to the Chinese tradition that mothers during the first month after childbirth are supposed to take uttermost care and pay close attention to food, nutrition and health, and where others, typically mothers-in-law or the mother's own mother are instrumental in offering a conducive environment for the new mother. Among affluent women, it has recently become more common to employ staff to support the new mother during this month.
 Fong, 'China's one-child policy and the empowerment of urban daughters.'
 Evans, The Subject of Gender.
 See e.g. Croll, Endangered Daughters.
 Eklund, Rethinking Son Preference.
 Hoffman, 'The value of children to parents.'
 Lisa Eklund 'The sex ratio question and the unfolding of a moral panic? Notions of power, choice and self in mate selection among women and men in higher education in China,' in Scarce Women and Surplus Men in China and India: Macro Demographics Versus Local Dynamics, ed. Sharada Srinivasan and Li Shuzhuo, Place: Springer, 2016 (forthcoming).
 Patricia Jeffery, 'Supply-and-demand demographics: Dowry, daughter aversion and marriage markets in contemporary north India,' Contemporary South Asia 22(2) (2014): 171–88.
 Wei Shangjin and Zhang Xiaobo, 'The competitive saving motive: Evidence from rising sex ratios and savings rates in China,' Journal of Political Economy, 119(3) (2011): 511–64.
 Evans, The Subject of Gender.
 Wei and Zhang, 'The competitive saving motive.'
 Greenhalgh and Winckler, Governing China's Population; Merli and Smith, 'Has the Chinese family planning program been successful in changing fertility preferences?'
 In November 2015 the Communist Party announced that a two-child policy was to replace the one-child policy, indeed signalling that low fertility and population ageing are stronger concerns than population size.