The Grandmothers’ Farewell to Childcare Provision under China’s Two-Child Policy: Evidence from Guangzhou Middle-Class Families

As China’s one-child policy is replaced by the two-child policy, young Chinese women and their spouses are increasingly concerned about who will take care of the ‘second child.’ Due to the absence of public childcare services and the rising cost of privatised care services in China, childcare provision mainly relies on families, such that working women’s choices of childbirth, childcare and employment are heavily constrained. To deal with structural barriers, young urbanmothersmobilise grandmothers as joint caregivers. Based on interviewswith Guangzhoumiddle-class families, this study examines the impact of childcare policy reform since 1978 on childbirth and childcare choices of women. It illustrates the longstanding contributions and struggles of women, particularly grandmothers, engaged in childcare. It also shows that intergenerational parenting involves a set of practices of intergenerational intimacy embedded in material conditions, practical acts of care, moral values and power dynamics. We argue that the liberation, to some extent, of young Chinese mothers from childcare is at the expense of considerable unpaid carework from grandmothers rather than being driven by increased public care services and improved gender equality in domestic labour. Given the significant stress and seriously constrained choices in later life that childcare imposes, grandmothers now become reluctant to help rear a second grandchild. This situation calls for changes in family policies to increase the supply of affordable and good-quality childcare services, enhance job security in the labour market, provide supportive services to grandmothers and, most importantly, prioritise the wellbeing of women and families over national goals.


Shifting from Childbirth to Childcare under the Two-Child Policy
To increase the birth rate, which has declined to 1.047 in 2015 (Guo, Gietel-Basten, & Gu, 2018), China's birthcontrol policies have shifted away from the one-child policy to the two-child policy. Since 2016, all Chinese couples are allowed to have two children regardless of family background. However, four years after implementation, data issued by the National Bureau of Statistics of China (NBSC) shows that both birth number and the birth rate have continued to fall over the past three years (NBSC, 2019). Given that the new two-child policy has not achieved a satisfactory policy outcome, it suggests little association between reduced birth rate and strict birth-control policies. Researchers have recently explored other barriers to young Chinese couples' decisions to have children. Both case studies as well as regional and national surveys have suggested that more than half of young Chinese couples who already have one child do not intend to have a second one, and regard childcare capacity as their major concern (Shi & Yang, 2014). This means that how families understand their childcare capacity and the childbirth choices they make in the given socioeconomic and cultural environment also play a significant role in the effectiveness of policy. Therefore, issues of Chinese families' childcare arrangements and the impact on the choices whether to have children or not under the new policy have attracted growing interest from policy researchers. More importantly, the shift of focus from macro-level factors, such as urbanisation and birth-control policies, to micro-level family practices opens up a wider space for the academia to reflect upon the relationships between the state, family and Chinese women.
Studies have emphasised young Chinese women's harsh choices between childcare responsibilities and competitive labour market in the context of Chinese childcare reform (Cook & Dong, 2011). Family or childcare policies in China since 1949 have changed from a defamilialism model to currently, a familialism one (Ngok & Fan, 2018). During the socialist period , in order to rebuild post-war national economy and encourage women to work, the state minimised the caring role of the family and put up the slogan "women hold up half of the sky." Urban Chinese women were provided with free childcare services through danwei (work unit). By the end of 1956, there were about 26,700 different kinds of childcare facilities in urban areas, 260 times more than what existed before 1949 (Zuo & Jiang, 2009). Such defamilialised policies highly liberated married women from care duties, and urban women's labour participation rate increased up to 75% by 1988 (Meng, 2012).
During the economic transition period (1978-1990s), the state shifted its emphasis towards economic development and focused social expenditure on policy areas which cultivated an active labour force, such as higher education and health. As many state-owned enterprises which used to provide social services were shut down or reformed, publicly-funded childcare facilities shrunk rapidly. Meanwhile, the state loosened its regulations and encouraged the development of the private childcare market, leading to the rapid growth of private childcare centres from the mid-1990s onwards (Zhang & Maclean, 2012). The decline of public childcare facilities and the rising cost of private childcare services have implicitly redefined the family as the primary care provider.
With the advent of the new century, the state expanded social policies to enhance social harmony and political stability. A residual social security system has been established, but it only protects the most vulnerable and at-risk social groups, which includes orphans, disabled children, and those affected by poverty. For the rest of the population, the lack of affordable childcare has meant that the majority of childcare work still falls on the family alone. The decline in public childcare services has thus heightened tensions between work and caring duties for working mothers, particularly those who cannot afford to use private or commercialised care services (Du & Dong, 2013). This also contributes in part to the sharp decline of urban women's labour force participation rate, which had already fallen to 57% by 2009 (Hare, 2016).
Meanwhile, gender equality in private life has also not progressed substantially, with women continuing to shoulder more caring responsibilities than men. In fact, the birth control policy has resulted in each child becoming even more precious as they become the only hope of continuing the family line (Fong, 2002). While previous generations would worship their ancestors and place primacy over the needs of the elderly, today, the happiness and success of the youngest generation have become Chinese families' primary aim (Yan, 2018). To fulfil this goal, particularly with regards to fierce educational competition, Chinese families make unlimited investments for the benefit of their priceless only child. For Chinese mothers, it requires wholehearted devotion to childcare and child education (Yang, 2018).
It is true that some Chinese women are sceptical of the prevailing model trajectory of life and are eager to escape such a plight by remaining unmarried, practicing late marriage or delaying having children (Davis & Friedman, 2014). However, these are harsh choices afforded to women under the patriarchal structure and values of filial piety. Women in these circumstances are often discriminated, seen as being an "unfilial daughter" or "leftover woman" by their parents and partners . What's worse, those socioeconomically disadvantageous women are forced to leave the formal labour market, and some become housewives, leading to greater economic risk (Wu, 2014). The state also stresses the virtues of women who strive to build a greater nation through both labour participation and family building. This means that Chinese women should simultaneously be successful at work while also managing family duties (Tao, 2016). The state now places a higher demand on women than it did during the socialist period. Given these institutional constraints, married women are in a difficult position, reconciling tensions between work and childcare. To deal with these structural difficulties, then, grandmothers are often enlisted as joint child caregivers in both urban and rural areas.
As the state has retreated from providing childcare services, grandparents have gradually become secondary or even primary caregiver (Chen, Short, & Entwisle, 2000). A survey conducted by the China Research Centre on Aging in 2014 showed that 60-70% of Chinese children nation-wide between zero and two years of age were mainly taken care of by their grandparents, and among them, around 30% of children were cared for exclusively by their grandparents. Based on 2008 data by the China Health and Retirement Longitudinal Study, up to 58% of the elderly have once helped with grandchild caring for their adult children (Ko & Hank, 2013).
An intergenerational parenting arrangement, whereby grandparents and parents jointly take care of the child, is regarded as a family strategy that maximises resources for fulfilling the needs of both parents and children in contemporary China (Goh & Kuczynski, 2010;Zhang et al., 2019). Nevertheless, researchers have largely adopted a parental perspective, focussing on young mothers' struggles in the current socioeconomic and cultural context. Little attention has been paid to the lived care experiences of grandmothers, and whether such care arrangements are sustainable under the two-child policy and the reasons behind this. Thus, our work examines how grandmothers understand care provision, their responses to the demands of caring for a second grandchild, and the impact this response has on young couple's childbirth decisions.
While we do not deny that there are other factors that affect childbirth decisions, this article focuses on the variety of challenges experienced by grandmothers specifically and explores their emerging impact on the effectiveness of the two-child policy. Through the lens of grandparental involvement in childcare, we unveil structural constraints that have limited the choices and futures of Chinese women, both young and old. More broadly, our research questions relate to the greater role that grandparents are assuming in rearing the next generation worldwide in the era of increasing longevity, reduced fertility, and radical economic and family changes (Arber & Timonen, 2012;Buchanan & Rotkirch, 2018). In what ways can the case of China advance the understanding of grandparental involvement in childcare? What are the implications of China's family policies which currently, are highly familialised and gendered on meeting the needs of women and on enhancing their wellbeing?

Grandmothers in an Intergenerational Parenting Arrangement
A growing number of studies on grandparenting have explored grandmothers' care roles and their impact on life in old age, both in China and globally (Arber & Timonen, 2012;Chen, Liu, & Mair, 2011;Hayslip, Fruhauf, & Dolbin-MacNab, 2017). A study on the intergenerational parenting arrangement in the non-local families of Beijing shows that young mothers play a dominant role in childcare and make decisions in children's social and education activities, while the grandmothers mainly assist mothers with care work and household chores (Xiao, 2014). In this respect, the care role played by Chinese grandmothers are to some extent consistent with their Australian and Hong Kong counterparts (Kirby & Sanders, 2012;Leung & Fung, 2014). However, the division of care work and the power structure in the case of China are seen to result from the prevalence of modern parenting practices, and the decline of parental authority in the private sphere (Xiao, 2014). Moreover, studies have shown that variations in grandmothers' involvement in terms of time and care services have a clear association with family policies (Silverstein & Zhang, 2013), the availability of part-time jobs for women (Bordone, Arpino, & Aassve, 2016), and cultural values (Lee & Bauer, 2010;Low & Goh, 2015). Geographical proximity and a closer bond between two generations also increase grandparents' involvement in childcare, particularly maternal grandmothers (Zhang et al., 2019;Zhang, 2014).
Other studies have explored the impact of care responsibilities on grandmothers' retired lives. Regular care provision by grandmothers has been found to result in less leisure, less personal care, and less sleeping time (Li, Ju, & Huo, 2016). Full-time and long-hour caregivers also suffer higher subjective time stress, less life satisfaction, and worse health. Grandparenting also causes financial difficulties because extra expenses are incurred alongside less paid work participation due to caregiving (Bowers & Myers, 1999). Therefore, grandparenting generally results in mixed feelings of joy and exhaustion, but those who undertake a major care role feel significantly stressed (Goodfellow & Laverty, 2003;Leung & Fung, 2014). This is particularly the case for co-residential and maternal grandparents in low-income families and in places with a familial culture and limited childcare services (Burnette & Sun, 2013;Low & Goh, 2015).
Grandparenting can also cause family conflict between grandparents and parents in East Asia when grandparents play a teaching or disciplinary role rather than the non-interference supportive role expected by parents (Goh, 2013;Leung & Fung, 2014). Tensions are particularly intensified between in-laws in three-generation co-residential families (Thang, Mehta, Usui, & Tsuruwaka, 2011). However, studies suggest that the intergenerational bond remains strong in China (Shi, 2017). In rural China, the adult children (both daughters and sons) who migrate to cities for work reciprocate through emotional and financial support for their parents' involvement in grandchild care. Such exchange enhances intergenerational intimacy in China (Zhang & Silverstein, 2012).
Additionally, grandmothers' social participation is negatively affected due to their involvement in demanding and lengthy childcare responsibility. Chinese grandparents tend to have reduced participation in social activities and retire earlier to provide childcare for their adult children (He & Wang, 2015;Li et al., 2016). This not only weakens grandmothers' financial abilities, but also threatens their well-being as they feel isolated from peers and communities, particularly in areas where social services are limited (Ludwig, Hattjar, Russell, & Winston, 2007).
Existing studies have informed our study in three ways and, consequently, we aim to address three main questions. First, the caring experiences of Chinese grandmothers have to be understood in the specific socioeconomic and cultural contexts governing Chinese women across time and place. Given China's current familialised childcare policies and moral values, marketised child-care services, urban Chinese grandparents may undertake heavy care work. Therefore, how do Chinese grandmothers engage in care provision and how do they make sense of care provision to the lives of their daughters' and their own?
Second, whether an intergenerational parenting arrangement is likely to be formed or sustained is bounded by certain material conditions and financial capacities.
Thus, in what ways the childcare provided by urban grandmothers is precious for Chinese young women? Do women of the families which have more socioeconomic resources have better options?
Third, an intergenerational parenting arrangement involves a set of practices of intergenerational intimacy that is embedded in a given emotional and power relations between two generations. It means that the exchange of caregiving might be in various forms and go beyond monetary exchange. Thus, to what extent are young couples' decisions regarding childbirth affected by the response of grandmothers? If grandmothers hesitate or even resist requests to help rearing a second grandchild, under what circumstances do grandmothers actually agree to look after a second grandchild?

Method
To address these questions, we conducted in-depth interviews with grandparents and their adult children from ten middle-class dual-earner families in the city of Guangzhou during 2016 and 2017. Families who owned at least two flats in the city were identified as middleclass. In this main sample, the grandmothers acted as joint caregivers who had once helped or were now helping to raise at least one grandchild.
Middle-class families were recruited for two reasons. First, compared with low-income groups, urban mothers in these middle-class families are better able to use marketised care services (e.g., hiring a caregiver or sending their children to private childcare centres). Instead, they choose to seek help from grandmothers. Thus, it helps us further demonstrate women's struggles as well as the unique role of grandmothers in childcare when the affordability and availability of childcare is increasingly become a concern to Chinese women. Second, middle-class grandmothers who provide childcare may expect more non-material forms of return from their adult children. It thus helps us better understand how Chinese grandmothers make sense of the caregiving to maintaining their intergenerational relationship.
To maximise the diversity of the sample, we considered important variables that would affect a grandmother's carer role, such as living arrangements, lineage (maternal or paternal), number of adult children and grandchildren, gender of adult children/grandchildren, and places of origin (local or non-local) (Ikels, 1998;Zhang et al., 2019). To better demonstrate the complexity of family dynamics and diverse caring experiences, we also interviewed a small number of fathers and grandfathers in some families. To better investigate the impact of grandparenting on childbirth decisions, we also recruited two other middle-class families in which the grandmothers did not provide caregiving as deviant cases for this study.
Of the twelve families interviewed, most grandparents were in their 60s, although one was in their 70s (F9) and another in their 50s (F11). Most adult children were in their 30s. Seven families had one child, one mother was already pregnant with her second child, and four families had two children. Most children were under school-age.

Precious Informal Care Resources
Our findings reveal that the grandmothers provided considerable childcare regarding time and services for the working mothers, particularly when the grandchildren were pre-schoolers. Grandmothers' timetables closely matched their grandchildren's care needs and the demands of the mothers' employment. Grandmothers from all ten families who provided care did so at some point daily for ten hours, or full time, when the grandchildren were aged under three. They started care work when the parents went to work and finished when they returned home. When the grandchildren went to kindergarten, the grandmothers provided half-day/weekend part-time care. It is obvious that, given the unequal division of family responsibilities between men and women, working mothers mobilised a large amount of resources, such as time and care services, from grandmothers to meet the needs of labour participation.
The gender pattern continues in all ten families as the mothers and grandmothers were disproportionately involved in childrearing activities and household chores compared to fathers and grandfathers. In their own words: "Mothers educate, grandmothers care, fathers finance, and grandfathers stand by and watch." The mothers, particularly those in rural-urban migrant families, held greater decision-making power regarding important childcare and educational matters. This implies a challenge to traditional intergenerational power relationships and a decline in parental authority. The mothers clearly expressed that they made decisions about activities such as seeing a doctor, seeking a nanny, choosing a kindergarten/school, teaching with school assignments, etc. In contrast, the grandmothers engaged mainly in more physical work such as picking up/sending children to school, preparing meals, food shopping, cleaning, and washing, thus enabling mothers to focus both on educating a competitive child and attempting to ensure job opportunities.
The maternal grandmothers were more heavily involved in childcare than paternal grandmothers, given the changing family relations and intimate practices. The parents stressed that maternal grandmothers were their first choice for childcare because they were considered to be more reliable. The fathers further pointed out fewer family conflicts occurred or would be easier to handle when maternal grandmothers were involved. Instead of being married out, daughters now maintain a close bond with their natal families and practice filial piety after getting married (Shi, 2017). This has led to the redefinition of filial obligations and change in intimate practices in grandparenting. Maternal grandmothers showed more dedication to childcare in order to enhance emotional bonds and secure their daughter's employment and economic future. They acted more flexibly and were willing to compromise when conflicts occurred. In reality, however, our findings show that childcare choices were influenced by a range of factors, including the maternal grandmother's health condition (F12), post-retirement employment opportunities (F10), the maternal grandmother's relationship with her daughter (F8/11), and intergenerational living conditions, i.e., proximity and space (F2). Further, the intergenerational parenting relation sometimes involves two sets of grandparents. While one set provided care, the other provided financial support, for example hiring a nanny, or paying for grandchildren's commercial insurance or extra-curricular classes. We also found that practices of intergenerational intimacy in childcare are bounded by material conditions and families' financial capacity, e.g., the affordability to buy flats in the cities. Young couples were clearly aware of the significance of geographical proximity in achieving an efficient and sustainable intergenerational parenting arrangement. All young couples received financial support from their elderly parents to buy flats in the city. Five of the ten spouses lived together with the grandparents, and two families had shifted from co-living to nearby-living as the children grew up; the remaining three families lived nearby with the grandparents.
Compared to those living in a nearby-residence situation, grandmothers in the co-living condition provided more hours of care.
We also noted that with the decrease of public childcare services, access to available, affordable, and goodquality care and education for pre-school-aged children has become an imperative issue for all families interviewed. Only four out of the twelve families had hired or planned to hire a nanny as an alternative childcare choice. In Guangzhou, the current cost of a hired domestic caregiver for children is 5,000 to 6,000 RMB per month, accounting for half of the average monthly salary of employees in Guangzhou. None of the couples had chosen to send their children to private day-care centres before their children were able to attend public kindergarten at three years of age. As the mother in F1 said: "I wanted to put my child in the nearby kindergarten at first, as it offered very good services and they had native English teachers. But I gave up after I found out that it cost 8,000 RMB per month." The rising expense of childcare in the private sector constitutes a burden even for urban middle-class families, and the lack of access to affordable and good-quality childcare and education limits options of urban Chinese women so that they realise they have to rely on grandmothers for free childcare. We found that none of young couples paid the grandmothers specifically for childcare provision, though the couples that cohabited with the grandparents did pay for food and other living expenses related to childrearing. Moreover, compared with marketised childcare services, the parents considered the childcare provided by grandparents to be "safer" and "more reliable." The parents we interviewed complained that it was very difficult to find a reliable nanny, and that even if nannies were hired, grandmothers had to oversee their care work. In fact, the parents repeatedly mentioned news reports linking nannies and private care institutions to child abuse, child abduction, and child injuries.
Childcare choices are even fewer for those women who cannot rely on informal care, however. In the case of F11, neither maternal nor paternal grandmothers were available to provide childcare, as both sets of grandparents were already busy helping rear the grandchildren of their sons. To better care for her child, the mother chose to engage in flexible employment as a freelancer. This unfortunately meant that she had to stop doing interpretation work and now only took translation work. She expressed worries about her future job opportunities and income security during the interview, and voiced deep concern that current pension and health insurance programmes were unable to provide sufficient security for people like her who are self-employed or in other forms of precarious employment. This lack of social security for those in precarious employment makes women even more vulnerable.

Working Around the Clock
The theme of hard work, duties, and stress emerged often in the grandmothers' narratives. They often used phrases such as "working on duty" or "going to work" to describe caregiving. In contrast, fun-related words such as "enjoyment" and "fulfilment" were more used by parents when they described how the grandparents felt about caring experiences. The differences in emphasis between the two generations suggests that there are complex feelings surrounding childcare, but all interviewees agreed that caregiving has increasingly become hard work rather than being the leisure activity that they recalled from their past experiences. One father said: When I was a child, my parents didn't pay me too much attention…they just made sure I had enough food and nutrition, and paid my education fee….When they were busy, my grandmother occasionally looked after me, and sometimes I just stayed at home with my cousin. But now each child is cared for in detail. (F7, hairstylist) Long hours of service and intensive physical work were frequently mentioned by the grandmothers. Phrases like "I'm just like a non-stop spinning top," "I don't even have a second to sit down and take a rest," and "I won't have time to see a doctor until I die" vividly demonstrated their stress and the intensity of child care work. Although a few grandmothers felt their physical health condition was deteriorating, they were more concerned about losing the autonomy and freedom to arrange their old-age life. They even felt mental stress in relation to ensuring their grandchild's safety. As one grandmother said: I did wish to learn singing and drawing and to travel like my friends, but I didn't have time….We later hired a nanny to do the household chores and I took care of my grandson only….I had to keep an eye on the little boy, if he fell down, or the nanny might not be careful enough….I couldn't bear the responsibility. I couldn't sleep at night as I kept thinking what things I have to do tomorrow, and things went on and on in my head.
(F2, retired employee of a state-owned enterprise) We also found serious intergenerational conflicts between the mothers and grandmothers, particularly between in-law relations and in co-residential families. Differences of living habits and childrearing methods caused conflicts which were intensified by the crowded space. One father commented: We have been living with my parents all these years and the fights between my mother and my wife never stop. Quarrels start in the kitchen about what to eat tonight, in the living room about which TV show to watch, and in childrearing about whether it is OK to use diapers after a child is two years old. (F6, private company manager) Intergenerational conflicts in childcare demonstrate the tensions that exist between two different types of parenting knowledge. Young urban women feel empowered as their mothers increased investments since their grade school and higher education expanded during the reform era (Fong, 2002). Young mothers absorb parenting knowledge from peers and friends, on-line communities, and parenting guidebooks and experts. The past experiences of their parents do not constitute the sources of knowledge. However, they must rely on the grandmother's help with childcare to fulfil their needs. Thus, the conflicts in childcare emerge in knowledge-based forms and can be hard to reconcile.
Meanwhile, we found that when parents recognised the grandmothers' sacrifices and were willing to reciprocate in some ways, intergenerational conflicts could be reduced. Most mothers acknowledged that it was not the grandmothers' responsibilities to take care of the grandchildren. They were also aware that it was a tremendous sacrifice for the grandmothers to provide childcare. Most mothers increased material and emotional support to show their gratefulness to the grandmothers, particularly the maternal ones, such as buying gifts, paying for travelling expenses and arrange family activities. One mother said: During the school holidays I put my daughter with my in-law's side. And I ordered the tourist group for my parents and we paid for the fee, too. You have to give them some rest, just like recharging. (F9, university project officer)

The Grandmothers' Farewell
Most grandmothers actively helped their adult children with childcare under the one-child policy and the de-sire for childcare provision was initially both reciprocal and altruistic. Some grandmothers sought to strengthen the emotional bond and to secure more old-age support from their adult children in return. Additionally, some grandmothers confessed that they had to help, "otherwise they would be subject to gossip" (F8). This was particularly the case amongst the paternal and rural-urban migrant grandmothers. However, none of the grandmothers who had experienced socialist life and work questioned the legitimacy of women's labour participation. State discourses about women's liberation between 1949 and 1990 criticised housewives as being "unworking parasites" and praised working mothers (Jin, 2013), and this still exerts a profound impact today. The grandmothers also acknowledged young mothers' difficulties in taking good care of their children while managing their jobs. Thus, we see the interplay of childcare policy reform, the competitive labour market, and both traditional and modern cultural values in shaping the grandmothers' choices to provide childcare support in urban China.
Nonetheless, most grandmothers who had once provided childcare did not want to help rear a second grandchild. Except for F5 (who had twin granddaughters) and F4 (who thought little about this matter), the grandmothers from the other eight families explicitly or implicitly told their adult children that they were unwilling to help with childcare again at first. One grandmother said she told her daughter: "You are free to give birth, but I will not take care of the child again" (F1, retired kindergarten teacher).
This stance was common among the grandmothers who had provided substantial childcare support, whereas those who had provided much less or no childcare agreed to help their adult children with a second child. A similar situation was found among the grandfathers, who even looked forward to another grandchild. The one exception was F12. As the grandmother was in poor health, the grandfather (a retired manager of a state-owned enterprise) had taken care of the first grandchild. He refused to help a second time around, saying: "One time is totally enough! I cannot stand a second time.'' The grandmothers in multiple-children families also wanted to retreat, but this depended on which adult child they had previously helped with childcare. One father said: Before my mother came over to help, she had helped my elder sister take care of her child until she was nine years old, hoping that my sister would take care of her when she got older. But now my sister wants to have a second child. My mother worries that my sister will ask her to help again….She is very tangled. I don't want her to do so either….Three grandchildren one by one. Too terrible. (F8, newspaper editor) Shifting from active involvement to retreat suggests that these grandmothers are no longer willing to sacrifice themselves for their adult children. Most grandmothers confessed that as they approach 70 years of age, they are physically and mentally incapable of caring for a second grandchild. Meticulous childcare with its intensive workload and long duration was far beyond the grandmothers' limits.
Despite financial and other material support received from their adult children, the grandmothers expected more recognition and emotional support. Some complained that the adult children did not show adequate understanding and respect of their personal needs. When the grandmothers wanted to attend social activities (e.g., square dancing) to relax and "take a breath," the adult children sometimes worried about the childcare being affected. The grandmothers had to compromise and prioritise the needs of their grandchildren and adult children over their own. One grandmother said: While I am still able to walk around with my legs, I want to enjoy a free life in my later years. If I did it again, it would be another three to five years. How many five years do I have? (F10, retired clerk) The young couples stressed that it was their own decision whether or not to have a second child, and that it had nothing to do with their elderly parents. However, we found that the decisions were dependent on whether an intergenerational parenting relation could be sustained. Out of the ten families, there are three couples (F1, F6, and F9) who have had a second child since the state policy was relaxed. All three couples had successfully negotiated with the grandparents for a new intergenerational parenting arrangement before deciding to have the second child. In the case of F1, the mother had hired a nanny to share the care work with the maternal grandmother. In F6, the mother had bought a small flat with her husband and moved out of her parents-in-law's home so that her parents could come to Guangzhou and help with the second child. Thus, childcare work shifted from the paternal grandparents to the maternal grandparents. In F9, the couple successfully persuaded both sets of grandparents to help together and divided childcare work equally between them. Two of the couples interviewed (F10 and F11) are currently in the process of building up intergenerational parenting coalitions. In F10, both sets of grandparents were standing firm and refusing to help. The mother, a state-owned enterprise employee, said: "I have no choice but to have a baby first and then find a way to get their help." In F11, we observed how the grandmother's choice had changed a mother's childbirth decision. The mother, a freelance translator, was pregnant with a second child during the interview. It was an unplanned pregnancy, however, and she had first gone to the hospital for an abortion when she realised: I really couldn't do it. I called my mother from the hospital and asked her for opinion. She said, don't get the abortion, that she will find a way to help. I was relieved and I decided to keep the baby.
Thus, when young couples successfully rebuild an intergenerational parenting arrangement by enlisting grandmothers from both sides and/or supplementing that help with marketised means, they feel supported to have a second child. However, although a new intergenerational parenting arrangement can be built, extra expenses are often incurred due to new intergenerational living arrangements, and this also increases the uncertainties around having a second child. If an arrangement cannot be made, under current circumstances the young couples have to either seek marketised means, or the mothers may withdraw from the formal labour market. In these cases, the couples are very likely to give up their plans to have a second child.

Limitations
This article examines the role of grandparenting in childcare provision as well as in young Chinese couples' decisions around childbirth under the new two-child policy. We must acknowledge that there were a few limitations to our study. First, it does not explore the role of fathers in childcare, or how marriage dynamics affect decision-making regarding childbirth. We did observe that the fathers who were well educated or who brought in less income than their wives were more engaged in childcare work, and that men's participation also effectively decreased the grandmothers' workload. This suggests that childcare capacities and childbirth decisions may be bounded by even more complex family dynamics in childcare.
Second, other than grandparenting, our study does not consider other micro-level factors that may also influence young couples' childbirth decisions. Four couples had decided not to have a second child (F2, F3, F7, and F8). These couples, particularly the mothers, expressed great concern about the negative impact a second child would have on their personal career development, the effects on life quality of their first child, and the grandparents' old-age quality of life. Thus, future studies need to identify the micro-level mechanisms that explain the process of childbirth decision-making.
Third, we didn't examine the role of grandparenting in socioeconomically vulnerable families. For example, single mothers are often more in need of grandparental childcare in order to participate in the labour market (Posadas & Vidal-Fernandez, 2013). In these families, women spanning two generations might struggle more than the middle-class ones illustrated in our study, and they deserve more attention in future research.
Finally, change in birth rate is a dynamic and complex process. A longer observational period and more rigorous research are required to address the relationships between family behaviours and the birth rate trend in order to better inform policies that meet the needs of Chinese women and their family members.

Conclusion
Our study first and foremost illustrates Chinese grandmothers' considerable contribution to childcare provision to meet the needs of young mothers faced with institutional constraints such as a lack of public childcare, insufficient social security attached to flexible employment, persistent familialistic values and underregulated marketisation of childcare and education. The rising role of grandmothers in childcare has kept pace with the state retrenchment from childcare services since China's reform and opening-up in 1978. Without grandmothers' longstanding efforts, the labour force participation of young mothers-which is significant to China's economic growth-would probably decrease even more sharply given today's current circumstances. Thus, beyond the contribution to wellbeing of their daughters and families, those elderly Chinese women deserve social recognition from the state and Chinese society as a whole.
Our study also shows that intergenerational parenting involves a set of practices of intergenerational intimacy that are embedded in material conditions, practical acts of care and moral values. Though the young couples seldom give grandmothers a monetary payment explicitly for providing childcare, they reciprocate in the forms of emotional support (e.g., companionship, arrangements for family leisure activities), practical care and other implicit material means (e.g., buying gifts). However, little recognition of grandmothers' personal needs (e.g., live with a free retirement life) has made urban grandmothers reluctant to rear a second grandchild. It well notes the maternal grandmothers are increasingly involved in young women's decisions over childcare and childbirth. Redefined filial piety practiced by singleton daughters has enabled maternal grandmothers to contribute more in childcare, and act more flexible when family conflicts occur.
Furthermore, our study questions the rising status of Chinese women under the achievements described in the White Paper Equality, Development, and Sharing: Progress of Women's Causes in the 70 Years since the Founding of the PRC (State Council Information Office, 2019). The state has constantly instrumentally used the family-in particularly women-as an unpaid childcare workforce to serve its social and economic goals. We observed how grandmothers who jointly provide intensive childcare suffer from physical and mental stress as well as a lack of freedom in their retirement lives. Consequently, some of them are now beginning to withdraw from providing childcare under the two-child policy, and their reluctance is discouraging young mothers from having a second child. It is not until Chinese women collectively make decisions that conflict with the state's interests that the government will start to respond. The Chinese central government has recently issued instructions on developing childcare services, and initiated tax reductions for child education (CCP Central Committee, 2016). Some local governments, such as Shanghai, have initiated pilot programmes to develop early childhood care institutions. However, it remains uncertain whether these new measures indicate a new overall policy approach in China.
Second, intergenerational parenting reproduces gender inequality within families. To deal with structural disadvantages, young mothers enlist grandmothers to share tremendous amounts of care work required to raise children. The liberation, to some extent, of young women from care work occurs not because of a progress in gender equality in both the public and the private domains, but it takes place at the expense of older women by engaging them into unpaid housework. Although the caring experiences are mixed with joy, fulfilment, stress, and hard work, both mothers' and grandmothers' choices are still greatly limited by institutional barriers compared with their male partners.
Finally, the farewell of Chinese grandmothers to childcare with the introduction of the two-child policy shows that the persistent neglect of women's wellbeing by the state can by no means sustain families' childcare capacities. In fact, it has exerted negative influences on policy efforts to increase the nation's birth rate. The birth rate may continue to drop if the state insists with the current familialism model. Although some measures have been initiated, the state's policy stance is ambiguous and keeps swinging between a defamilialism model and a familialism model. Alternatively, since the late-1990s, most Western societies have chosen a supported-familialism model (Bordone et al., 2016;Daly & Lewis, 2000) that supports families with resources from public and private sectors, and employs various types of policy tools such as cash benefits, services, and parental leave. However, these caring policies often aim to relieve the pressures of young women as caregivers in nuclear families. As China has a high incentive to reduce social expenditure under the current economic slowdown, and as it faces strong public pressure and concerns about political stability, it is highly possible that the state may also embark on a supported-familialism model, but in a different way. It may enhance the caring role of families by providing more support to grandmothers in the extended families. Some time is required to see which policy model China adopts and they adapt it to Chinese conditions. It depends on how the Chinese government understands the values of the family-a policy means or a policy endand to what extent it recognises the equal rights that women should be entitled to.