Introduction

Since the 1960s, global fertility rates have halved, grabbing the attention of researchers. Low fertility rates are now common in both developed and developing countries [1–4]. The situation has raised the question of whether fertility behavior adequately reflects people’s preference for the number of children they would like to have; the discrepancy between ideal and actual number of children known as the fertility gap [5].

On January 1, 2016 China relaxed its family planning policy and adopted a universal two-child policy to actively address the country’s aging trend. However, the new policy has failed to have any significant effect on the fertility rate of many provinces [6, 7], and in fact the birth rate decreased from 13.0 births/1,000 women in 2016 to 12.3 births/1,000 women in 2017 [6]. In light of the country having the highest sex ratio at birth in the world and the huge burden of the aging population [8], improving the fertility rate is an urgent priority in China. Motivated by the above scenario, an understanding of the determinants of fertility intention will help policy-makers tackle these challenges.

Although Inner Mongolia is an Autonomous Region of China which has relaxed the one-child policy for Mongolian people, the birth rate of Inner Mongolia has been very low for many years. In 2015, the birth rate in Inner Mongolia was 7.72‰ which ranks it 26^th among the 34 provinces of China; at the same time the birth rate was 12.07‰ in the whole country [9]. This study aims to determine the psychosocial determinants on a couple’s fertility intention that influences the ‘fertility gap’ and to understand how couples make their fertility intention within the opportunities and constrains provided by the societal structures in which they are embedded. The discrepancy between ideal number and planned number of children is largely unexplored and calls for further research on fertility decision-making among couples.

Wives’ age at marriage is one of the most important determinants of child-birth in a family. Under the Chinese law regarding marriage and birth, the postponement of marriage means the postponement of childbirth. In addition, the out-of-wedlock birth is in conflict with traditional Chinese culture, so any level of fertility in a population can always be traced by the percentage of women at first age of marriage [10, 11].

Minorities in China have had low fertility rates since the 1990s [12], even being supported by the relative liberal one child policy in China. Inner Mongolia has the lowest fertility rate in all Autonomous Regions in China [13], in both urban and rural areas. Compared with Han, the Mongolian ethnic group have fewer sons and a multi-child preference [4].

Most research studies combined demographic characteristics, such as education, occupation, and income, as socioeconomic factors linked with fertility [14–16]. Previous research showed a significant negative correlation between fertility and wives’ socioeconomic status [17], but the negative relationship disappears or becomes even positive after accounting for the endogeneity of schooling [18, 19]. A study from Norway showed that husbands who had a higher education delayed their fatherhood, yet fewer remained childless, and the rates of second and third births increased with their educational attainment [20]. Among all published studies, the impact of husbands’ socioeconomic status on fertility decision has largely been overlooked. Similar situations were reported in studies from China [21–23].

The Theory of Planned Behavior (TPB) is an extensively validated psychosocial theory useful for understanding couples’ fertility decision [24]. It is a general psychological theory concerning the link between attitudes and behaviors and has also been applied to explain fertility decision-making [25–27]. Based on the TPB, three determinants influence behavior intention which is a proxy measure of the behavior (in our case, a decision to have a child): 1. attitudes toward the behavior which refer to the individual’s positive and negative feeling of the behavior and outcome of performing the behavior; 2. subjective norms, which relates to the individual’s perception of the social environment surrounding the behavior; and 3. perceived control over the performance of the behavior [24].

The objective of this research was to investigate determinants of fertility decision-making among couples in Inner Mongolia. We developed our initial conceptual model based on the Theory of Planned Behavior [27] (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Conceptual model based on the Theory of Planned Behavior.

https://doi.org/10.1371/journal.pone.0221526.g001

Material and methods

Results

Demographic characteristics

A total of 1,513 couples were asked to join the study, of which 1,399 (92.5%) agreed to participate and completed the whole questionnaire and whose data were included in the analysis. The mean age at marriage of the wives was 26.3 years (standard deviation (SD) = 3.1). The majority of the families (74.3%) belonged to the Han ethnicity and most had a middle level education. About one-third of the couples (34.2%) were both self-employed. Over 80% of the couples had monthly income less than 10,000 Chinese yuan (equal to 1,440 US dollars) (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic characteristics of the study sample.

https://doi.org/10.1371/journal.pone.0221526.t001

Descriptive analysis of construct variables of fertility decision model

Summary statistics of all variables of the three dimensions in the fertility decision model are shown in Table 2. The mean overall score was highest for the Perceived behavior control domain (2.52), followed by Attitudes (2.34), and Subjective norm (1.89). The means of all individual items were in the middle range (1.86–2.68), indicating that most respondents placed the value of each factor as “of little importance” to “very important”. Cronbach’s alphas of all domains exceeded the minimum acceptable level of 0.70, indicating high internal consistency of the items measuring the same construct.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Summary of behavioral constructs and their subdomains.

https://doi.org/10.1371/journal.pone.0221526.t002

Comparison of the total scores of the fertility decision model by socio-demographic characteristics

Table 3 shows the total scores of all manifest variables of the fertility decision model by social-demographic variables. Higher scores indicate a higher perception of the importance of perceived behavior control, social norm and positive attitudes on the decision to have a (another) child. The results showed that the scores of fertility decision were not different between couples belonging to minority and Han ethnic groups. However, significantly higher score were seen among more educated couples, couples with a higher monthly family income and those who were both employed, compared to their counterparts. Wives who married between the ages of 24–35 years had higher mean scores compared with those married at an earlier or later age. Based on these univariate results, education, employment status, monthly income and wife’s age at marriage were included in the initial SEM of the fertility decision model. Although ethnicity was not significantly associated with the fertility decision in this univariate analysis, it was deemed to be an important factor for the fertility decision (the fertility policy has been regulated differently between Han majority and other ethnic minority populations), it was thus kept in the initial model.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Comparison of the total scores of the fertility decision model by socio-demographic variables.

https://doi.org/10.1371/journal.pone.0221526.t003

Predictors of fertility decision model based on Theory of Planned Behavior (TPB) by structure equation model

Model fit indices.

The overall fit indices for the final structural equation model were as follows: the relative Chi-square = 3.868 (χ²/df), RMSEA = 0.045 (90% CI: 0.041, 0.050), AGFI = 0.949, and CFI = 0.966. All indices indicated that the final model fitted the data well. Fig 2 shows the structural model of the association between fertility decision with demographic characteristics.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Final model of fertility intention based on Theory of Planned Behavior.

PC: Perceived Behavior Control; PC1: The couple is ready to sacrifice the time and freedom for the baby; PC2: A suitable babysitter is available when the couple works outside; PC3: A lot of money will be spent to raise a (another) child; PC4: The couple has sufficient materials for child rearing; PC5: The couple’s family members can help them take care of the baby; PC6: The couple has a stable income; SN: Social Norms; SN1: The family should have a boy (or girl); SN2: His/her partner prefers to have a boy (or girl); SN3: The relatives and friends around the couple have already had two children; AT: Attitudes; AT1: The couple enjoys having a big family; AT2: The couple enjoys the fun of raising a child; AT3: Only healthy parents can have a child; AT4: Having a child can maintain a good relationship between the couple; e in a circle denotes measurement error of that latent variable.

https://doi.org/10.1371/journal.pone.0221526.g002

Standardized regression weights.

Table 4 shows standardized regression weights of all variables in the fertility decision model, based on the theory of planned behavior, obtained from the SEM. The strongest determining domain was perceived behavior control (β = 0.92, 95% CI: 0.89, 0.98), followed by attitudes (β = 0.83, 95% CI: 0.78, 0.87) and subjective norm (β = 0.64, 95% CI: 0.59, 0.69). Within each domain, perceived importance of having a stable income (PC6) (β = 0.77, 95% CI: 0.74, 0.79), subjective norm on the partner’s sex preference of the newborn child (SN2) (β = 0.75, 95%CI: 0.72, 0.80) and the attitude that only healthy parents can have a child (AT3) (β = 0.82, 95% CI: 0.78, 0.85) were the most important manifest variables.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Standardized regression weights of parameters.

https://doi.org/10.1371/journal.pone.0221526.t004

Ethnic group was significantly associated with the subjective norms towards fertility decision, with ethnic minority couples having less weight than Han couples (β = -0.13, 95% CI: -0.18, -0.06). Education was significantly associated with perceived behavior control and attitudes, but not with subjective norms. Couples with higher education had more perceived behavior control (β = 0.20, 95% CI: 0.14, 0.27) and attitudes (β = 0.83, 95% CI: 0.78, 0.87) than those with lower education. Employment status also had significant and positive relationship with all three domains of the fertility decision model.

Discussion

This study followed the Theory of Planned Behavior (TPB) with the aim of advancing the understanding of determinants of fertility behavior among couples in Inner Mongolia, China. The strongest determinants of the fertility decision among Chinese couples were found to be in the domain of attitudes, which were that only healthy parents can have a child (AT3) (β = 0.82, 95% CI: 0.78, 0.85) and that a child can maintain a good relationship between the couples (AT4) (β = 0.79, 95% CI = 0.75, 0.82) reflecting that having a child cannot only affect the relationship between a couple, but also link with the health status of the couple. Infertility in China is very low. If young couples do not have children, there will be pressure from family members and friends. One study [35] showed that both positive and negative partner relationships had a negative effect on the timing of first as well as second and third births. Another study [36] suggested that childlessness might result in a decrease in quality of life and increase in marital discord and sexual dysfunction. Childlessness could also impose physical, psychological, emotional, and financial burdens. Our findings thus emphasize the role of children in Chinese families and call for public support for older couples who are childless.

Among the perceived behavior control variables, the perceived importance of having a stable income (β = 0.77, 95% CI: 0.73, 0.79) and having family members to help take care of the child (β = 0.75, 95% CI: 0.71, 0.78) had the highest effects on fertility decision. This indicates the high concerns on the costs for raising a child and awareness of the burden of childcare by couples. The cost of raising a child does not only depend on the financial costs, but also on the value of child in their parent’s view, which changes with time. Children became increasingly valued for their emotional worth rather than for their economic contribution in modern society of China, with compulsory education and schooling replacing work as the child’s primary social obligation [37]. Because of the inadequate infant-care social welfare system, it is common that the grandparents provide childcare for the third generations in China. Such situation makes the couples at childbearing age hesitate to have a (another) child [38]. Similar findings were found even in the context of a country with strong institutional support for childrearing such as Norway and Japan [26, 27, 39].

The perceived importance of having a suitable babysitter (PC2) shows the change in child-rearing practices of a family where young parents need others to help take care of their baby. In traditional Chinese culture, caring for the elders is the responsibility and obligation of the whole family, especially the children. However, in the Chinese modern society, the family structure relationship has shifted from the older-centered intergenerational relationships to the next generation as the center [40]. Older parents provide childcare for the three generations fostering a better relationship between generations [10]. This can explain why the couples in our study raised the issues of having a babysitter and family member to take care of their baby as the important determinants for their fertility behavior.

The strong effect of perceived importance of self-sacrifice of time and freedom for their baby (PC1) shows that with the development of society and an increase in people’s education level and self-consciousness, young couples no longer passively accept the birth of a child. Rearing a newborn baby needs a lot of time and energy by the parents; this not only increases the economic pressure on the entire family, but more importantly, devoted love of children needs enormous efforts from parents. Another reality is that the time allowed during statutory maternity leave has been found to be insufficient for taking care of a newborn child in China [41] and public kindergartens do not accept children less than three years of age [42]. Therefore, to have a child may have a great impact on a couple’s career advancement, so a couple’s career type and aspirations may have a major influence on their fertility intentions.

The sex preference of the newborn child was measured by two subjective norms (SN1: the preference of the respondent as representing the couple and SN2: the preference of his/her partner), which had significant effects on fertility decision. In contemporary China, gender preference is a sensitive issue, which is often described in the public and popular discourse as ‘‘traditional’, ‘backward’ and ‘feudal’ [10, 43]. In fact, preference for a son, a desire common among couples in India, Nepal, and Bangladesh [43, 44], is also common in China. China’s one-child policy, implemented in 1979 and recently abolished in January 2016, contributed to the country’s high sex ratio, the highest in the world. A higher proportion of males in their reproductive age was estimated at around 30 million in 2010 [45]. Preference for a son has changed in the modern Chinese society. Raising a son entails huge economic pressures, and people now realize that daughters, more than sons, can provide better emotional support for their parents. The advantages of having a daughter are increasingly being recognized by more couples, and more young couples have shifted their attention from reproducing a particular sex to the quality of child care [10]. In this study we did not specify the preference for a boy or girl in the questionnaire, but simply a sex preference, however our findings indicate that sex preference (SN1 and SN2) still matters in the decision by a couple to have a child and, in fact, had a greater impact than their relatives and friends already having two children (SN3), which, since the adoption of the universal two-child policy, has become more popular.

Apart from the above-mentioned determinants, ethnic minority couples had less weight than Han couples (β = -0.13, 95% CI: -0.18, -0.06). The fertility culture of Han Chinese prefer to have at least a son and multi children [46]. Compared with the fertility culture of Han, the minority Mongolians do not have such preferences [47].

Couples who were both employed perceived more behavior control compared with couples who had at least one self-employed. People’s increasing educational attainment and labor force participation has contributed to fertility decline in most developing countries [48]. The family role conflict theory asserts that there is a contradiction between work and family life, which are time conflicts, stress conflicts and behavior conflicts [49]. In the modern society of China, young couples put high values on educational attainment and career promotion, so they may worry that having a (another) child would be an obstacle and be an opportunity cost.

Although this study showed reasonable similarities with previous studies both internationally and in China, some limitations should be acknowledged. Our study aimed to describe the fertility situation in Inner Mongolia, therefore the generalizability of the findings to the whole China is limited. Although cross-sectional designs cannot reveal causal relationships, they can provide some points for further studies. Also, the discrepancy of fertility intentions between the first and second child was not specified.

Conclusions

This study supports the assumption that use of the theory of planned behavior can explain the fertility decision-making model, which includes perceived behavior control, subjective norms, and attitudes. Due to the culture, minority couples are less affected by subjective norms. With improved education and employment, these couples may be more affected by perceived behavior control, subjective norms and positive attitudes towards to having a child in the future. Result of this study can help the Chinese government improve the fertility intention of its residents and the fertility rate of the country. We have the following recommendations. First, the government should provide equal access to basic public service as the expansion of basic education and infant care. Second, the government should provide adequate provisions for maternity protection and parental leave as an essential policy. Third, more consideration should be given to older ages couples who are involuntarily childless. Finally, the government should implement a supporting policy to reduce the time pressure and opportunity cost for couples who are conflicted by working and child-rearing. Sufficient parental leave for couples should be considered.

Supporting information

Acknowledgments

The authors would also like to thank all participants who were enrolled in this study.