Childbearing intention and childbearing behavior in low fertility society: evidence from Shanghai

Exploring the patterns of population change under low fertility is a historical mission for demographic research in the 21st century. Taking Shanghai as an example, this paper discusses the new patterns and characteristics of people’s childbearing intention and behavior under low fertility settings. The results show that Shanghai, as a typical society with ultra-low fertility, people’s childbearing intention and behavior demonstrate the characteristics such as “downward inconsistency” between childbearing intention and childbearing behavior, “convergence” of childbearing intention and childbearing behavior between different sub-groups, “inversion” in childbearing intention and childbearing behavior with education and income, and the “decoupling” between marriage and childbearing. The childbearing intention and behavior of low fertility society as seen in the case of Shanghai show the patterns as somehow different from what was observed in a high fertility society. It suggests that many of the childbearing patterns shown in the high-fertility era may not necessarily be applicable in low fertility settings, and urges more research to be carried out in this regard.


Introduction
China has witnessed the dramatic fertility decline since the 1970s, and become a lowfertility country since the early 1990s. The total fertility rate (TFR) has been below the replacement level and continues to go downward over the last three decades. The seventh national census showed that the TFR has been as low as 1.3 in 2020. With the recent relaxations of fertility policy in China, from the selective two-child policy in 2014 to universal two-child policy in 2016 and then to the policy for three children in 2021, the period of fertility control is gradually phasing out from the historical stage, and a new population situation is looming large. For a long time, in order to accelerate the decline of fertility rate, demographers have carried out a variety of researches to understand the determinacies and consequences of fertility changes in the high fertility period. Can those learnings be immutably applied to the understanding of fertility change under low fertility or even ultra-low fertility? Under the low fertility rate, does it show the same patterns as observed in the high fertility period? Exploring the patterns of population change in the era of low fertility and its policy implications will undoubtedly become the mission of demographic research in the 21st century.
Shanghai is one of the regions with lowest fertility in China (Chen & Gu, 2018;Chen, 2021). The 20 years from 1950 to 1970 witnessed rapid fertility decline in Shanghai as indicated by the TFR from 5.6 to 2.28, which further dropped below replacement level in 1971, indicating that Shanghai has already completed the classic demographic transition and been in the era of low fertility for 50 years since then, which was much earlier than China as a whole. Just as the situation of many European countries after completing the classic population transition, the total fertility rate in Shanghai has not been maintained at the replacement level, rather continued to go downward to a very low level of 1 or even below 1 as so-called "lowest low" (see Fig. 1). The objective of this paper is to take Shanghai as an example to discuss the new changes of childbearing intention and behavior under low fertility rate in Shanghai based on the available fertility statistics, relevant social surveys and authors' previous studies (Chen & Gu, 2014;Chen 2018;Chen & Gu, 2018;Chen 2020a;Chen 2020b;Chen & Gu 2020;Chen, 2021).

Data and methods
This paper uses the fertility statistics of Shanghai since 1949, and data from 27 surveys about childbearing intention in Shanghai conducted between 1981 and 2017. The cross-sectional historical meta-analysis was used to link the results of those surveys to examine the change of childbearing intention in Shanghai over the last three decades, and to explore the relationship between childbearing intention and childbearing behavior. Meanwhile, the paper makes comparison of childbearing intention and childbearing behavior between different sub-groups, including the urban and rural residents with Shanghai hukou, residents with and without Shanghai hukou, only child and not only child, and people with different incomes and education levels, etc.
In addition, the data from the "Survey on the Status and Needs of Two Children Families in Shanghai" we conducted in the six districts of Shanghai from August 2017 to August 2018 (hereinafter referred to as the 2018 Survey) are also used as another major source for this paper. The 2018 survey gathered information about the demographic, social and economic characteristics, as well as the motivations for having a second child, and the cost of childrearing for couples who had given birth to a second child with Shanghai hukou. The details of the survey are available in the authors' previous papers (Chen & Gu, 2018;Chen, 2021).

The downward inconsistency between childbearing intention and childbearing behavior
Studies have shown that childbearing intention and childbearing behavior are not always consistent. However, the inconsistency does not mean that childbearing intention and childbearing behavior are irrelevant to each other. On the contrary, there is a strong correlation between them. That is, people's actual number of children is often higher than their reported childbearing desire in the early stage of demographic transition, but the relationship is reversed in a low fertility society at the end of population transition (Bongaarts, 2002). According to the data released by National Bureau of Statistics, the number of children intended of women of childbearing age was 1.8, while the actual fertility level was 1.3 in 2020, indicating that China has entered a stage when the fertility level is lower than the childbearing intention, with a gap of about 0.5 child.
For Shanghai, which has entered an era of low fertility as early as in 1971, there has been a situation that people's actual childbearing behavior is lower than what they may desire since the 1980s, but the changing trend of the two appears generally consistent in a virtually parallel pattern (see Fig. 1) 1 . There are three implications. Firstly, people's actual fertility level is indeed lower than the reported fertility intention in a city with long-term low fertility level, which is largely because in big cities such as Shanghai, both the direct living cost and opportunity cost of childbearing have been enormous, preventing some couples from fully realizing their childbearing desires. Secondly, taken into account of the possible gap between the two, the long-term follow-up childbearing intention survey can still provide a basis for understanding and predicting people's actual fertility level. That is, when the childbearing intention is weakening, the childbearing behavior is likely to be more depressed. Thirdly, high quality longitudinal survey, supplemented by reasonable quantitative models, can to a great extent measure the impact of childbearing intention on childbearing behavior. The surveys in Shanghai have shown that the gap between childbearing intention and actual childbearing behavior is 0.379 on average, that is, the actual fertility level is about 0.4 lower than the reported intention in number of children, a bit smaller than that at the national level (0.5) as suggested above. It should be noted that the underlying causes of the gap in the 1980s as well as in the 2010s may be considerably different. Given Shanghai began implementing the One-child Policy in 1979, in the 1980s, the gap between the two indicators was largely due to the birth control policy. With the recent relaxations of fertility policy in China, the role of birth control policy has been greatly diminising, and socioeconomic factors are playing an increasingly significant and even decisive role.

The convergence of childbearing intention and childbearing behavior between different sub-groups
With the acceleration of urbanization and the deepening of urban-rural integration, the ideal number of children of residents with Shanghai hukou has shown a decreasing trend since the early 1980s, which was roughly 2 in the early 1980s, to approximately 1.5 in the 1990s, and around 1.3 after 2000 (see Fig. 1). Moreover, while among them the ideal number of children of rural residents was higher than urban residents in the early 1980s, the gap was narrowed in the 1990s and gradually converged after 2000. The difference in gender preference for children between urban and rural residents has also gradually disappeared. Nevertheless, some differences in childbearing desire still exist between residents with and without Shanghai hukou, though the ideal number of children of residents without Shanghai hukou has been also decreasing to less than two children, and the son preference is weakening as well, especially among the young people. As the younger generation of floating population live longer in Shanghai, their childbearing intention may become more and more assimilated toward their peers with Shanghai hukou. In addition, although there are some differences in childbearing intention between couples born as only child and not only child, and among people with different incomes and education levels, the average number of children intended to have is still less than 2 among all the subgroups in Shanghai (Chen, 2021).
In terms of actual number of children, according to fertility statistics of Shanghai, more than 90%, even 95% of the annual newborns by the residents with Shanghai hukou were the first child from the early 1980s to 2012. With the gradual relaxations of fertility policy in recent years, the proportion of the first child among the annual newborns has been decreasing meanwhile the proportion of the second child has been increasing since 2013. In 2017, 66.53% of the newborns were the first child which is the lowest in proportion since 1980, and 32.54% were the second child. This reflects the shift from the vast majority of families having only one child to an increasing number of families having two children (see Fig. 2), though it is still extremely rare to have three or more children since the early 1980s. Consistent with the trend of the ideal number of children, the actual number of children of urban and rural families has been also converging. In the 1970s, the actual number of children of rural families was higher than that of urban residents. The two had become relatively close by the mid-1980s and almost the same since the mid-1990s(see Figs. 3 and 4). While the proportion of women without Shanghai hukou who having two children is always higher than women with Shanghai hukou, the proportion of women without Shanghai hukou who having three or more children remains very low. Only 5.9% of women without Shanghai hukou who had given birth to a child in 2019 which is her third child even it is the highest since 2004 (see Fig. 5).
In summary, despite some differences in terms of childbearing intention and behavior among people with different socioeconomic characteristics, the convergence is obviously more prominent than the differences between them. Whatever childbearing intention or childbearing behavior, to have one or two children is the actual choice of the vast majority of the families, with very few couples for three or more children. The convergence of childbearing intention and childbearing behavior by different sub-groups may be one of the mostly remarkable characteristic we should pay attention as far as the low fertility is concerned. 2

The inversion in childbearing intention and childbearing behavior with education and income
Previous studies have found that the decline of high fertility occurs first among the better educated groups, and in particular, that the upgrading of education for women has played a significant role in leading fertility decline (Zheng, 2019). The five fertility surveys from 2013 to 2017 found however, that among the people with Shanghai hukou those with higher education tend to be more likely to have a second child. Among the 2,009 women aged 15-49 years with Shanghai hukou interviewed in the 2017 National Fertility Survey (Zhuang et al., 2018), the average number of children intended by the women with below primary education, junior high education, senior high and technical secondary education, college education, undergraduate education and graduate education was reported at 1.09, 1.07, 1.08, 1.16, 1.17, 1.36, respectively, demonstrating an upward tendency. In addition, among the couples who had a 2 The paper mainly compares the childbearing intention and childbearing behavior between different subgroups. However, from the perspective of life course, people's childbearing intention and childbearing behavior may also change with their age, marital status, number and gender of children they already had. second child interviewed in the 2018 Survey, nearly 90% of the wives and husbands had a college degree or above, more than 60% of the wives and husbands had a bachelor's degree or above, and the proportion of husbands with a master's degree or above (14.44%) was higher than that of their wives (8.90%).
The surveys conducted in 2009, 2012 and 2013 also found that the proportion of intending to have two children goes upward along with the increase of the income level among residents with Shanghai hukou (Chen 2020 a ; Chen 2020b). The survey conducted by Chen and Wu in Minhang District of Shanghai in 2009 showed that childbearing intention of higher income groups is relatively higher than that of lower income groups as well (Chen & Wu, 2009). In terms of actual childbearing behavior, there were 40.8% and 18.2% families with two children interviewed in 2018 Survey having a total annual post-tax income of more than RMB 200,000 and 300,000, respectively. It may suggest that families with higher income are more likely to intend to and actually give birth to a second child. Furthermore, 86.99% of these surveyed families took the income of husbands as the primary financial source.
According to the classic fertility theory, fertility tends to be negatively related with education and income (Weeks, 2014;Poston et al. 2017). For people with higher education, the opportunity cost of childbearing and childrearing tends to be higher than that of those with less education, so they are more likely to have fewer children compared with couples with less education. However, the Shanghai case suggests the otherwise, i.e. in low-fertility societies where couples tend to have less than two children on average, people with higher education are more willing and more likely to have a second child than those with a lower level of education 3 . This is largely due to the high costs for living and childcare in Shanghai; people with higher education levels are more likely to have higher incomes and are less subject to financial constraints. As a result, people with higher education and income tend to have the ability to put their childbearing intentions into action.
It is true that not only in Shanghai, but also in some Northern and Western European countries, it is observed that better educated women tend to have higher childbearing intention than less educated ones (Testa, 2007;Lanzieri, 2013). Therefore, the relationship that exists between women's education level, family income and childbearing desire, and actual childbearing behavior during the fertility transition stage tends to turn inverted in the low fertility setting. This once again reminds us that many of the childbearing patterns observed in the high-fertility era may not be necessarily applicable in low-fertility settings.

The decoupling between marriage and childbearing
Previous fertility studies have shown that childbearing tends to be followed sooner or later after marriage in one's life. Getting married is the premise and prelude of having children, and giving birth is a quite certain follow-up life event after getting married.
It is precisely because of the close relationship between the two life events that China had vigorously advocated late marriage for late childbirth, so as to achieve fertility reduction during the past several decades. However, it seems evident that marriage is decoupling itself from childbearing under low fertility scenario.
The relationship between marriage and childbearing has been changing as already observed in some European countries under the second demographic transition. The connection between sexual behavior, marriage and childbearing has been diminishing, which means marriage and childbearing are no longer closely linked to each other, and people may have sex even neither in marriage nor for pregnancy and childbearing. On the one hand, non-marital childbirth is becoming more and more common and recognized. On the other hand, giving birth to a child is no longer the inevitable follow-up event after getting married, which means people may not have children immediately after getting married or even do not give childbirth at all. In addition, marriage patterns have changed: fewer and later marriages, more and more divorces, and even more for non marriages. And the phenomenon of unmarried cohabitation has been booming rapidly, jeopardizing the stability of traditional marriage and family system. Unmarried cohabitation has gradually become a "quasi marriage" (Lesthaeghe, 2014).
Similar new phenomena can be observed in Shanghai as well. For example, the marriage rate has been decreasing from around 15 per thousand in 1985 to 6 per thousand in 2020, while the divorce rate keeps an increasing trend from less than 1 per thousand in 1985 to more than 4 per thousand in 2020 (See Fig. 6), late marriage and unmarried cohabitation have been becoming more and more popular. Unlike in Europe however, even with unmarried cohabitation so common in Shanghai, it has not yet become an alternative form of mainstream marriage, but more likely function Fig. 6 The marriage rate and divorce rate of residents in Shanghai, 1985-2020 as a trial marriage and a premarital transitional period. If getting pregnant during cohabitation, people will mostly go for a shotgun marriage along with childbirth or simply perform an induced abortion. The phenomenon of non-marital childbirth that can be actually observed is still very rare, since it is not well accepted by the society and the law. We also observed that the connection of marriage and childbearing has become weaker and weaker, and even been in a trend towards decoupling each other. Not only do more and more people choose not to get married or get married later, but even if they get married, they may not give birth to a child immediately or even decide to have no children as observed in many countries with low fertility. According to the monitoring data of Shanghai Municipal Health Commission in 2019, among the couples who are both first married in 2014, only 7.46% of those couples gave birth to the first child in the year of getting married, 22.15% in the second year after marriage, 22.98% in the third year, 11.31% in the fourth year, and 6.88% in the fifth year, that is nearly one-third of the couples still have no children even five years after marriage.

Conclusion and discussion
Based on the empirical data, this paper finds that there are a series of new characteristics and trends in the childbearing intention and behavior of the residences in Shanghai as a typical city with ultra-low fertility, which are different from as seen in the era under high fertility: 1) The long-term trend between actual fertility level and childbearing intention is consistent, with still a gap between them, but the pattern is reversed from what was observed in a high fertility setting, that is, actual fertility tends to be lower than what intended under the low fertility scenario.
2) A trend of weakening differences and increasing convergence in childbearing intention and childbearing behavior is observed between different subgroups.
3) Couples with higher education and income were more likely to not only intent but actually have a second child, which suggests the relationship of women's education level, family income with childbearing intention and actual childbearing behavior in the low fertility setting tends to be contrary to the negative correlation revealed by the traditional fertility theory.
4) The relevance between marriage and childbearing, the two life events ought to be closely related, is becoming disconnecting to each other, and even appearing a trend of decoupling.
All in all, the study suggests that many of the patterns in childbearing and marriage observed in the high-fertility era may not be necessarily applicable in lowfertility settings, which calls for in-depth research that may deepen the understanding of people's childbearing intentions and behavior under low fertility. This study represents a preliminary effort in this regard, based on the situation in Shanghai only. It is necessary to test the findings of this research with additional studies to determine whether the characteristics of childbearing intention and behavior in low fertility society found by this study are of universal significance for the whole country, and even some countries and regions in East and Southeast Asia with the same cultural tradition as China.
Furthermore, while many characteristics of marriage and childbearing behavior observed in the European countries experiencing the second demographic transition have gradually been appearing in Shanghai, there are still some differences. Some characteristics have already been very obvious and even more prominent than those of European countries, such as the extremely low fertility level, but some others are not so much shown such as non-marital childbirth since it is yet recognized by the law or the culture, which have been also observed in many East and Southeast Asian cities. Will Shanghai's demographic transition move towards closer to the European countries? Or will Shanghai and even China go on a different pattern of the second demographic transition from European countries due to the difference between eastern and western cultures? There is also to be answered along with more empirical studies in this regard in the era of low fertility.

Conflict of interest
The author has no conflicts of interest with respect to the content of this article.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/ licenses/by/4.0/.