Factors affecting childbearing age

Childbearing age is affected by many factors, such as the economic situation and the stage of social development in a community (Balbo et al., Reference Balbo, Billari and Mills2013). However, childbearing is essentially a family-related and personal issue. With social development and the gradual improvement of social security, the impact of familial factors on individual childbearing age is slowly decreasing. At the same time, individuals can increasingly make independent decisions about their affairs, and the impact of personal factors on childbearing behaviour is increasing (Mitchell and Gray, Reference Mitchell and Gray2007). Catherine Hakim put forward a similar argument in her preference theory: ‘in prosperous modern societies, preferences become a much more important determinant, maybe even the primary determinant of woman’s behaviour’ (Hakim, Reference Hakim2003, p. 361). Some people contrast the fertility behaviour of people with a ‘traditional’ family orientation and those with a ‘modern’ family orientation. The traditional family orientation emphasises that families should be formed as soon as possible and that families with children are complete families (Freedman et al., Reference Freedman, Coombs, Chang and Sun1974), whereas the modern family orientation emphasises people’s feelings and the realisation of self-worth. On the face of it, this would seem to suggest that a traditional family orientation should advance the onset of childbearing and a modern family orientation should delay it. However, some authors argue that a modern family orientation induces people to advance their marriage and childbearing so that they can escape at a young age from an unequal family environment and so that once childbearing is over their work and life will not be interrupted by childbirth. People with a traditional family orientation may also cherish the value of children and prefer more preparation time to provide their children with the best environment, leading to delayed childbearing (Feng et al., Reference Feng, Kugler and Zak2000). Therefore there is no consensus on the impact of family orientation on the age at childbearing. In this paper we shall examine this issue.

There is also an association between the number of years spent working (which we refer to in this paper as ‘working years’ and childbearing age. Here, the findings in the literature are rather more consistent for women who have reached the end of their childbearing years: the more years worked, the later the childbearing age. However, there is no agreement on the mechanism behind this association. Work experience has a crowding-out effect on childbearing in that, the more work experience a woman has, the greater the opportunity cost of having children. This opportunity cost consists of a current opportunity cost element (the salary foregone during periods devoted to child care) and a future opportunity cost element (the future accumulation of human capital abandoned because of career interruption). Economists have long argued that human capital plays an important role in people’s decisions about the timing of childbearing (Gustafsson, Reference Gustafsson2001). Generally, they hold that the earlier the birth-related interruption to working life occurs, the greater the future opportunity cost (Cigno and Ermisch, Reference Cigno and Ermisch1989), and some studies have provided evidence of this (Amuedo-Dorantes and Kimmel, Reference Amuedo-Dorantes, Kimmel, Gustafsson and Kalwij2006; Correll et al., Reference Correll, Benard and Paik2007; Miller, Reference Miller2010).

Others have explained delayed childbearing from a gender equality perspective. The decline of fertility has been associated with increased female labour force participation accompanied in some countries by a lack of institutional factors which make combining childbearing and work easier (Brewster and Rindfuss, Reference Brewster and Rindfuss2000; Billari and Kohler, Reference Billari and Kohler2004). The decline in fertility was most pronounced in these countries with traditional familial social policies (García-Manglano et al., Reference García-Manglano, Nollenberger and Sevilla2014). McDonald (Reference McDonald2000) observed that in countries with high levels of gender inequality in individual-oriented institutions (such as education and the labour market) and family-oriented institutions, fertility is low. Fertility transition from high to low is accompanied by a conflict between personal aspirations and a woman’s expected role in the family. An important idea is that, as the overall level of gender equality in a society increases, fertility initially declines, since more women choose activities that compete with childbearing, but then increases again, as the society modifies its institutions to allow women to combine childbearing with these activities (Esping-Anderson et al., Reference Esping-Anderson and Billari2015). Therefore the relationship between the fertility rate and gender equality is U-shaped.

There are some societies, however, with considerable gender equality in the public sphere where fertility remains low because of an imbalance of gender equality between the public and the private spheres. In these societies, women are becoming equal to men in the public labour market but the degree of equality in the family is lower. The increase in fertility which accompanies gender equality at the macro-level requires gender equality to increase in both public and private spheres. Not every family accepts gender equality at the same time, so equality in the private sphere can lag behind that in the public sphere. Family support for an egalitarian lifestyle, so that a balance is achieved within the household between the financial contributions of men and women and their domestic responsibilities can play an important role in promoting fertility (Mörk et al., Reference Mörk, Sjögren and Svaleyrd2013).

In addition to opportunity costs and gender equality, other objective factors affect childbearing age. People who have difficulty conceiving or bringing pregnancies to term, for example, may have a high average age at childbearing. This paper will not consider such factors, but will focus on the associations between the number of years spent working and the age at childbearing, as mediated by the economic position within the household and opportunity costs.

To be clear, we hypothesise that, among women who are approaching, or who have reached, the end of their childbearing years, a greater number of years spent working will be associated with delayed childbearing because of increased opportunity costs and because working strengthens women’s awareness of equality issues and this raises tensions if, within the family, their position is subserviant. We have no a priori hypothesis about the direction of the association between traditional or modern family orientation and the age at childbearing, for reasons already given.

In China, changes in the age at childbearing have an important influence on fertility. In theory, the fertility rate is determined by the proportion of women who have children and the average number that they have. In China, cultural factors mean that the willingness to have at least one child is high, so the effect of the proportions of women who have children on variations in the fertility rate is muted. For both traditional and modern family oriented people, children have great significance, although the nature of that significance might be different (Balbo and Arpino, Reference Balbo and Arpino2016). Not to have children is considered unusual and those who do not have children are sometimes referred to as ‘leftover men’ and ‘leftover women’ (Wang, Reference Wang2012; Evans, Reference Evans2015). Most people, therefore, want to reproduce. At the same time, Chinese women believe that they should complete childbearing before the age of 30 years (Xing et al., Reference Xing, Meng, Lin and Qin2019), regardless of the number of births they have, because of the risk of reproduction at older ages and their children’s health. Therefore the age at which childbearing starts affects the number of children women have (the earlier you start, the more children you can have before the age of 30 years). Fertility intentions are highest at ages 26-30 years and decline markedly after age 30 (Gao, Reference Gao2000).

Data and definitions of variables

Data were obtained from the China Family Panel Study (CFPS) 2014. The various waves of the CFPS are aimed at gathering data at the individual, family and community level to reflect the changes in China’s society, economy, population, education and health and provide a basis for academic research and public policy analysis. The CFPS pays attention to both the economic and the non-economic welfare of Chinese residents. It is a national, multidisciplinary project. Its sample covers 25 provinces, cities and autonomous regions and its target sample size is 16,000 households. All family members in these households are considered as respondents. In the 2014 wave the CFPS interviewed 37,147 people and screened 12,144 people who had given birth and were in the age range 17-49 years. However in our analysis we restricted attention to 3,069 men and women aged 40-49 years in order only to consider those who were unlikely to have another child (Li and Zhang, Reference Li and Zhang2021).

The dependent variable is the childbearing age, calculated by subtracting the age of the respondent’s eldest son or daughter from his or her own age.

Family orientation was characterised as ‘traditional’ or ‘modern’ based on responses to a battery of six questions: (1) ‘Despite how badly parents treat their children, children should still be kind to them’; (2) ‘Children should give up their personal aspirations and achieve their parents’ wishes’; (3) ‘Sons should live with their parents after marriage’; (4) ‘People should have at least one son to carry on the family line’; (5) ‘People should do something to honour their ancestors’; and (6) ‘Children should go home to visit their parents even if they work away from home’. The scoring was on a five-point scale from ‘strongly disagree’ (1 point) to ‘strongly agree’ (5 points). The study conducted principal components analysis on the responses to extract two independent principal components, one measuring the intensity of traditional family orientation and the other modern family orientation (Table 1).

Table 1.

Principal component analysis of family orientation: correlations between answers to six questions and traditional and modern family orientation

Cumulative sum of squares 55.7 %.

Source: China Family Panel Study 2014.

See https://opendata.pku.edu.cn/dataverse/CFPS?language=en [accessed 31 October 2022].

The variable ‘working years’ cannot be derived directly from the CFPS questionnaire. We therefore computed it as the respondent’s age minus years of education minus 6. The rationale behind this is that compulsory education law in China requires that children begin school at the age of six years. Accordingly, the age at leaving education is equal to the years of education plus 6. Since both men and women in China generally start working straight after leaving education the current age minus the age at leaving school can be regarded as the number of years spent working. Clearly this method of computing working years means that current age, years of education and working years form a linear combination. However, the correlation between years of education and working years was low.

Drawing upon the literature, the study also controlled for other variables that may affect age at childbearing, such as gender, educational level, personal income level, family income level, region, hukou, ownership of the enterprise in which the person worked and the type of work done. The family income variable was derived by taking the logarithm of the answer to the question ‘How much money did your family earn last year?’ (Vignoli et al., Reference Vignoli, Minello, Bazzani, Matera and Rapallini2021). The personal income variable was based on the logarithm of answers to the question ‘What was your total income last year?’. The type of work was classified as ‘agricultural’ and ‘non-agricultural’, as previous studies have found that non-agricultural workers have a later childbearing age (Wang and Chi, Reference Wang and Chi2017). The ownership of the enterprise was classified as ‘public’ or ‘private’. Previous studies have found that publicly-owned enterprises follow the government’s wishes and implement policies regarding marriage and childbearing, which may affect childbearing age (Cheng et al., Reference Cheng, Ma, Qi and Xu2021). The hukou is classified as ‘urban’ or ‘rural’, as previous studies found that people with an urban hukou have a later childbearing age (Mu et al., Reference Mu, Yeh, Zhang, Wang and Lin2022). Region is divided into ‘eastern region’ (Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong and Hainan), ‘central region’ (Shanxi, Inner Mongolia, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei, Hunan and Guangxi) and ‘western region’ (Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia and Xinjiang). Previous studies have found that people in central and western China have an earlier childbearing age than people in eastern China.

According to the income standard issued by the National Bureau of Statistics in 2014, personal income is classified into four bands: ‘low’, ‘middle’, ‘relatively high’ and ‘high’. We used this categorisation in our analysis. Family income is divided into ‘low’, ‘middle’ and ‘high’, where the dividing lines are the 30 per cent and 70 per cent quantiles of the distribution. We also computed a variable denoting the ratio of personal to family income (PTF), or personal income as a proportion of family income divided into three categories: ‘less than one quarter’, ‘one quarter to half’ and ‘more than half’.

Analysis procedure

The study used the Statistical Package for the Social Sciences (SPSS) 22.0 for the principal components analysis, correlation analysis and regression analysis. The principal components analysis has already been described. Regression was used to explore the effects of family orientation, number of working years and their interaction on the age at childbearing. We took childbearing age as the dependent variable and analysed the impact of family orientation and the number of working years on childbearing age for males and females (Models 1 and 2) and both sexes together (Model 3). We also examined the interaction of these two variables (Model 4).

To examine the effect of family income on the relationship between family orientation and childbearing age we divided the sample into three groups according to family income and estimated separate regression models for each group. To examine the impact of opportunity costs we estimated additional regression models for three family income groups, and for three personal income categories. If opportunity costs are important to the explanation we would expect the impact of working years on childbearing age to be greater among richer individuals.

Finally, we looked at the effect of traditional family orientation and modern family orientation on childbearing age for different levels of the ratio of personal to family income, as a measure of the respondents’ position within their households.