Intergenerational Financial Dependence, Perceived Fertility Benefits, and Low Fertility Intentions: Empirical Evidence from CFPS Data

Abstract

Maintaining an appropriate fertility level and population size is essential for Chinese-style modernization. Achieving this objective requires effectively addressing the current severe issue of low fertility intentions. Drawing on data from the China Family Panel Studies (CFPS), this paper examines the impact of intergenerational financial dependence on individuals’ fertility intentions. The results indicate that intergenerational financial dependence significantly reduces fertility intention. This conclusion remains robust after addressing endogeneity through instrumental variable methods and treatment effect models, and conducting a series of additional robustness tests. Mechanism tests indicate that the negative effect of intergenerational financial dependence on fertility intention is achieved by reducing perceived fertility benefits. The negative effect of intergenerational financial dependence on fertility intentions is primarily exerted on disadvantaged groups in market competition, such as non-Party members, low-income individuals, and those with low educational attainment. Furthermore, this negative effect is mainly driven by active intergenerational financial dependence. Overall, the findings of this paper explain the phenomenon of low fertility intentions from the perspective of perceived benefits.

Keywords:

intergenerational financial dependence

fertility intentions

perceived fertility benefits

Introduction

"The demographic dividend" provides favorable conditions for the economic growth of developing countries (Ssewamala, 2015; Taketoshi, 2020). However, as fertility rates continue to decline, the issue of low fertility has accelerated the loss of the "demographic dividend," which is not conducive to the economic transformation and upgrading of developing countries. As one of the world’s largest developing countries, China’s total fertility rate was only 1 in 2023, ranking second from the bottom globally. Low fertility has severely constrained China’s economic growth. To address this problem, China introduced the "selective two-child policy" in 2014 and the "universal two-child policy" in 2016, but these fertility adjustment policies yielded little effect. Data from the National Bureau of Statistics of China (NBSC) shows that the national birth population reached 16.55 million, 17.86 million, 17.23 million, 15.23 million, 14.65 million, and 12.00 million respectively during the period 2015–2020. The total birth population witnessed a slight increase only after the implementation of the "universal two-child policy," with the crude birth rate in 2020 dropping to a mere 0.852%. In 2021, China further implemented the "three-child policy," and in 2025, it launched inclusive childcare subsidy policies. Fertility intention, as people’s attitude toward childbearing, can predict future trends in fertility rates. Data from China’s National Bureau of Statistics shows that the fertility intention of women of childbearing age is only 1.8. Identifying the influencing factors and inherent causes of low fertility intention is crucial for effectively addressing China’s severe low fertility challenge.

Clan culture in China has a long history. As an informal institution, it supplements formal institutions (Li, 2025) and exerts a lasting influence on people’s behavioral decisions (Zhang, 2019). Under the influence of clan values, children are not only the continuation of the family bloodline (carrying on the family line) but also the economic source of support for parents in their old age (raising children for old-age security). However, fierce social competition has left young individuals unable to provide financial support to their elderly parents; instead, they are compelled to seek financial assistance from their parents. This phenomenon—where the financial support received from parents exceeds that provided to them by the individual—is defined as "intergenerational financial dependence," the core focus of this paper. From the perspective of intergenerational exchange theory, intergenerational financial dependence reflects an unbalanced state of intergenerational exchange within the family. It not only hinders the individual’s own development but also alters their fertility expectations: specifically, whether having children will provide financial security for their own old age or impose an additional economic burden in the future. Such considerations consequently influence individuals’ fertility intention. This constitutes the research question of this paper.

The "cost-utility" theory posits that fertility intention depends on the trade-off between the costs and utilities of childbearing (Leibenstein, 1974). Under the influence of traditional culture, children are obligated to support their elderly parents (raising children for old-age security). Guided by this notion, individuals can anticipate the economic benefits of having children. A balanced intergenerational interaction within the family is formed through the dynamic of "parents having children—children providing financial support for parents in their old age." In such balanced intergenerational interactions, having children yields corresponding and stable, predictable returns, thereby fostering higher fertility intention. Intergenerational financial dependence undermines this balanced intergenerational interaction within the family. It signifies that individuals can no longer obtain stable and predictable economic benefits through childbearing, challenging the "raising children for old-age security" concept rooted in clan culture. This disillusionment with the fertility expectation of "raising children for old-age security" reduces the perceived benefits of having children, ultimately lowering fertility intention.

However, empirically testing the causal relationship between intergenerational financial dependence and fertility intention is challenging. The main reason is that both low fertility intention and intergenerational financial dependence may be determined by individuals’ inherent traits (e.g., a laid-back mindset, lack of sense of responsibility). This gives rise to a selection bias issue, which is the primary confounding factor affecting the study’s conclusions. Additionally, individuals with low fertility intention are more likely to engage in intergenerational financial dependence, leading to a reverse causality problem. China’s strictly implemented one-child policy since 1980 provides a breakthrough for identifying the causal relationship between these two variables. On one hand, the strict enforcement of the one-child policy resulted in a large number of only children born after 1980. According to resource allocation theory, only children do not need to compete with siblings for family resources. This makes them more prone to lacking a competitive spirit (Yang, 2016) and thus engaging in intergenerational financial dependence. On the other hand, influenced by the traditional concept of "more children bring more blessings," people tend to prefer having more children. Therefore, most only children born after 1980 are essentially a consequence of the one-child policy.

In summary, using the identity of being an only child born after 1980 as an instrumental variable for intergenerational financial dependence can robustly identify the causal relationship between intergenerational financial dependence and fertility intention. To address the primary confounding factor of selection bias, this paper also employs a treatment effect model. In this model, the identity of being an only child born after 1980 serves as the exclusion restriction. After addressing endogeneity through the aforementioned methods, this study confirms a causal relationship between intergenerational financial dependence and fertility intention.

Compared with the existing literature, this paper extends the current research in the following three aspects:

First, this study enriches the literature on the determinants of fertility intention. Existing studies have examined the impacts of factors such as employment opportunities (Jensen, 2012), maternity leave policies (Farré & González, 2019), housing prices (Daysal, 2019; Liu et al., 2020), and household income inequality (Lin & Luo, 2025) on fertility intention and behavior. These factors primarily influence fertility through changes in childbearing costs (either direct or indirect costs), while few studies have focused on the impact of reduced childbearing benefits on fertility intention. This paper robustly identifies the causal effect between intergenerational financial dependence and fertility intention, exploring the determinants of fertility intention from the perspective of childbearing benefits and thereby enriching the existing research on this topic.

Second, it enriches research on the impacts of intergenerational support. Existing studies on intergenerational support have focused on its effects on the elderly’s well-being (Zhan et al., 2025), longevity (Bansak et al., 2025), and life satisfaction (Zhang et al., 2025), among other outcomes. As social competition intensifies, an increasing number of individuals still need to seek financial assistance from their parents after reaching adulthood—a phenomenon defined as intergenerational financial dependence. Under the influence of clan culture, "raising children for old-age security" is one of the key purposes of having children. Intergenerational financial dependence alters the notion of "raising children for old-age security," reduces the benefits of having children, and thereby lowers fertility intention. Few existing studies have paid attention to this topic. By investigating the relationship between intergenerational financial dependence and fertility intention, this paper supplements research on the impacts of intergenerational support from the perspective of intra-family intergenerational imbalance.

Third, it examines the heterogeneous impacts of intergenerational financial dependence on fertility intention across three dimensions: political identity, income status, and educational attainment. It also reveals the heterogeneous effects of different types of intergenerational financial dependence—active and passive intergenerational financial dependence—on fertility intention. This deepens the existing understanding of the causal relationship between intergenerational financial dependence and fertility intention, and provides important implications for addressing the issue of low fertility intention.

The subsequent structure is as follows: Section 2 presents the literature review and theoretical analysis; Section 3 describes the data sources and empirical design; Section 4 reports the stylized facts and empirical results; Section 5 discusses the moderating effects; Section 6 conducts the heterogeneity analysis; Section 7 offers the conclusions and policy implications.

Literature Review and Theoretical Framework

Literature Review

Literature closely related to this paper can be roughly divided into two categories. The first category focuses on fertility research. Drawing on the fertility decision-making theories proposed by Leibenstein, as well as Becker & Lewis, these studies explore the factors influencing fertility behavior and fertility intention from three dimensions: society, family, and the individual.

Social level: Housing prices affect fertility. High housing prices reduce women’s likelihood of having children (Liu et al., 2020). Specifically, the inhibitory effect of high housing prices on fertility intention mainly applies to rental households (Meng et al., 2023), while it has a promoting effect on the fertility intention of homeowners (Daysal, 2019). Public policies also significantly influence fertility intention. Studies have shown that government public spending (Zhang et al., 2022) and paid maternity leave (Girsberger et al., 2023) can both increase fertility intention. However, maternity leave policies do not boost fertility intention in all countries. Paternity leave in Spain increases the opportunity cost of having children and reduces men’s fertility intention (Farré & González, 2019). Additionally, maternity leave stigma is not conducive to improving fertility intention (Zhang, 2024). Cultural customs and social norms are also important factors affecting fertility (Zhang, 2019; Bethencourt & Santos, 2025). For instance, Zhang (2019) points out that the concept of "raising children for old-age security" embedded in clan culture is conducive to promoting fertility.

Family level: Zhang & Luh (2019) find that the health of grandparents has a sustained positive impact on family fertility. Lin & Luo (2025) show that household income inequality inhibits fertility intention by increasing childcare costs and reducing expectations of upward mobility. Chen & Xie (2025) indicate that the rising family childbearing burden is a key factor contributing to the decline in fertility rates.

Individual level: Jensen (2012) argues that increased employment opportunities lead women to enter the labor market or pursue higher education, delaying marriage and childbearing. Nie et al. (2023) demonstrate that internet use inhibits fertility by reducing marital satisfaction and changing attitudes toward "gender roles," among other pathways. Zhang & Zhao (2023) find that improved educational attainment among women increases their income and alters their fertility preferences, thereby delaying childbearing.

The other strand of literature focuses on intergenerational support, with a primary emphasis on its social effects—specifically, the impact of intergenerational support on the lives of older adults. Bansak et al. (2025) demonstrate that the "late marriage, sparse childbirth, and fewer children" policy increased the likelihood of parents co-residing with their children, enabling parents to receive more intergenerational support and thereby extending the lifespan of older adults. Zha et al. (2025) find that intergenerational support improves the health status of older adults, which in turn promotes their travel consumption. Zhang et al. (2025) point out that different types of intergenerational support exert heterogeneous impacts on the life satisfaction of older adults: emotional intergenerational support has a significant effect on the life satisfaction of rural older adults, while instrumental intergenerational support significantly influences that of urban older adults.

Based on the above analysis, existing studies have two shortcomings. First, although researchers have achieved fruitful results in exploring the factors influencing fertility and fertility intention, few studies have investigated these factors from the benefit perspective—specifically, they have not focused on the impact of declining fertility benefits on fertility intention. Second, as market competition intensifies, young people have become increasingly dependent on their parents’ financial support. Intergenerational financial dependence has thus evolved into a prevalent social phenomenon. This phenomenon has disrupted the intergenerational balance within families under the traditional model, leading to intra-family intergenerational imbalance. Consequently, fertility benefits have decreased, which in turn may affect individuals’ fertility decisions. However, existing literature has not yet addressed this issue. Therefore, using micro-level longitudinal household survey data (CFPS), this paper employs rigorous econometric models to explore the causal relationship and underlying mechanisms between intergenerational financial dependence and fertility intention. Furthermore, it examines the heterogeneous impacts of intergenerational financial dependence on fertility intention across three dimensions: political identity, income status, and educational attainment.

Theoretical Framework

Existing views generally hold that the reason people are unwilling to have children lies in the high cost of childbearing. Childbearing costs mainly consist of two components: direct costs and indirect costs. Direct costs include expenses for raising children, their education, and establishing their own families and careers, among others. Indirect costs primarily refer to the opportunity costs incurred due to having children. Following the logic of the cost theory, the key to addressing the "low fertility" phenomenon lies in reducing childbearing costs. Therefore, the main measures to mitigate low fertility derived from the cost perspective are as follows: first, lowering direct costs by increasing people’s income, providing financial subsidies to pregnant couples, and reducing child-rearing and education expenses; second, establishing a supportive policy system for childbearing—including postnatal employment support and vocational training—to reduce indirect costs. While the cost theory can explain some cases of low fertility intention, it struggles to account for the phenomenon where high-income groups are unwilling to have more children or even choose not to have any. The "quantity-quality trade-off theory of children" argues that high-income groups prioritize children’s quality over quantity (Becker & Lewis, 1973). However, this theory focuses on decisions regarding the number of children under the premise of existing fertility intention—addressing the question of "how many to have"—whereas low fertility intention is characterized by individuals’ reluctance to have children at all, making the theory inadequate in explanation. Thus, the cost theory cannot fully explain the current low fertility intention phenomenon. Furthermore, the logic of the cost theory leads to a paradox: the higher one’s income, the higher the opportunity cost of having children, and the less affordable childbearing becomes. This makes it difficult to achieve targeted governance of low fertility intention from the cost perspective.

Leibenstein’s "cost-utility" theory posits that factors influencing family fertility decisions include two aspects: costs and utility (Leibenstein, 1974). When considering the number of children to have, cost factors dominate; when deciding whether to have children at all, utility factors take precedence. Thus, this paper argues that the decline in fertility intention stems from reduced fertility utility—specifically, a decrease in the benefits derived from having children. In traditional societies, the benefits of having children can be summarized into two categories: first, providing additional labor. In societies with low productivity, a family’s production capacity is closely linked to its size, and having children means acquiring more labor. Second, ensuring a sense of security and economic support in old age, i.e., "raising children for old-age security." The logic of intergenerational exchange theory aligns with the "raising children for old-age security" concept in traditional Chinese fertility culture. This theory suggests that exchange relationships exist within intergenerational family dynamics: parents’ financial support for their children is essentially an intra-family intergenerational exchange, with the goal of obtaining more care, solicitude, and financial assistance from their children in old age (Zhang, 2019). Therefore, when individuals anticipate receiving greater feedback from their children in old age (i.e., higher fertility benefits), they tend to have children, manifesting as higher fertility intention. However, the "raising children for old-age security" concept has faced significant practical challenges in recent years. On one hand, commercial insurance-based old-age support has undermined this traditional notion; on the other hand, the prevalence of intergenerational financial dependence has increased. Intergenerational financial dependence has disrupted the balanced intergenerational exchange relationships within families, giving rise to a large number of unbalanced intra-family intergenerational exchanges. Such unbalanced relationships violate the "feedback model" rooted in traditional Chinese familism culture, reducing the economic benefits of having children for parents. Additionally, some cases of intergenerational financial dependence lead to value conflicts between parents and children, as well as unmet parental expectations of their children. This causes parents to experience a strong sense of "frustration," thereby reducing the non-economic benefits of having children (Liu, 2016). In summary, unbalanced intra-family intergenerational relationships caused by intergenerational financial dependence have diminished the utility of having children. Based on this, under the influence of the expectation effect, individuals in unbalanced intergenerational exchange relationships anticipate that they may face similar situations in the future. Consequently, they adjust their own behaviors in response to their current realities (He, 2009). It is evident that when intergenerational relationships are characterized by a "one-way" obligation-based ethics, individuals are more inclined to forego having children when making fertility decisions (Caldwell, 1976), thus exhibiting lower fertility intention. In conclusion, this paper proposes the following research hypothesis:

Intergenerational financial dependence reduces fertility intention.

The underlying mechanism through which intergenerational financial dependence reduces fertility intention is the decline in fertility benefits.

Data Sources and Empirical Design

Data Sources

The data used in this paper are derived from the China Family Panel Studies (CFPS) database. The survey covers 25 provinces/municipalities/autonomous regions in China, excluding Hong Kong, Macao, Taiwan, Xinjiang, Tibet, Qinghai, Inner Mongolia, Ningxia, and Hainan. A probability sampling method was adopted for sample selection, ensuring good national representativeness. This study utilizes CFPS data from 2011, 2016, and 2018. Specifically: Fertility intention, intergenerational financial dependence, and individual characteristic variables are obtained from the 2018 CFPS adult survey data; Household characteristic variables are sourced from the 2018 CFPS household economic survey data; The number of minor children is extracted from the 2018 CFPS household relationship survey data; Intergenerational financial dependence with a two-period lag is derived from the 2016 CFPS adult survey data; The number of siblings is obtained from the 2010 CFPS adult survey data.

Based on research needs, the data were further processed as follows: Excluding samples of individuals under 16 years old or over 49 years old; Removing samples with "don’t know" or "refuse to answer" responses; Eliminating samples with missing values.

Variable Design

Dependent Variable

The dependent variable in this paper is fertility intention. Since the 2018 CFPS surveyed individuals’ "desired number of children," this variable from the 2018 CFPS questionnaire is selected to measure fertility intention. The use of the "desired number of children" to operationalize fertility intention has also been widely adopted in existing literature (Nie et al., 2023).

Core Independent Variable

The core explanatory variable of this paper is intergenerational financial dependence. Reflecting the phenomenon of intra-family intergenerational imbalance, intergenerational financial dependence is operationalized using two variables from the 2018 CFPS questionnaire: "the average monthly amount of financial support provided to parents" and "the average monthly amount of financial support received from parents."

Individuals are defined as having intergenerational financial dependence if the monthly financial support they provide to their parents is less than the amount they receive from their parents. Conversely, those whose monthly financial support to parents exceeds the amount received are defined as not having intergenerational financial dependence.

Control Variables

To exclude the interference of confounding factors on the research conclusions, this paper draws on the approach of existing studies (Nie et al., 2023) and selects a set of control variables at both the individual and household levels. Control variables at the individual level include age, age squared divided by 100, household registration type, gender, years of education, health status, traditional values, marital status, political identity, and endowment insurance. At the household level, the control variables include the number of minor children, household per capita income, housing value, and housing property rights. In addition, provincial fixed effects are controlled for in the model to eliminate the interference of macro-level factors.

Based on the above variable design, Table 1 provides a detailed report of the names and definitions of all variables.

Table 1

Variable Names and Definitions
Variable Type	Variable Name	Definition
Dependent Variable	Fertility Intention	Desired number of children
Core Independent Variable	Intergenerational Financial Dependence	Dummy variable: 1 if the amount of financial assistance provided to parents is less than that received from parents; 0 otherwise
Control Variables	Age	2018 minus year of birth
	Age Squared / 100	(2018 minus year of birth) squared, then divided by 100
	Household Registration	Dummy variable: 1 for urban household registration; 0 for rural
	Gender	Dummy variable: 1 for male; 0 for female
	Years of Education	Number of years of education completed by the respondent
	Health Status	5 = Very healthy; 4 = Relatively healthy; 3 = Average; 2 = Relatively unhealthy; 1 = Very unhealthy
	Traditional Values	5 = Continuing the family line is very important; 4 = Relatively important; 3 = Average; 2 = Relatively unimportant; 1 = Unimportant
	Marital Status	Dummy variable: 1 for married (remarried) or cohabiting; 0 for unmarried, divorced, or widowed
	Political Identity	Dummy variable: 1 for Communist Party of China (CPC) member; 0 otherwise
	Endowment Insurance	Dummy variable: 1 for participating in endowment insurance; 0 otherwise
	Number of Minor Children	Number of children under 16 years old
	Household Per Capita Income	Natural logarithm of total household income divided by household size
	Housing Value	Natural logarithm of the current market value of the house
	Housing Ownership	Dummy variable: 1 if family members own full or partial property rights; 0 otherwise

Model Specification

Benchmark Model

To examine the relationship between Intergenerational Financial Dependence and fertility intention, this study constructs the following benchmark regression model:

child _i = β₀ + β₁kenglao_i + γcontrol_i+prov_i+ɛ_i (1)

child _i denotes individual’s fertility intention, and kenglao_i is a dummy variable indicating whether individual i engages in Intergenerational Financial Dependence. control_i represents the set of control variables, including individual and household characteristic variables. prov_i denotes provincial fixed effects, and ɛ_i is the random error term.

This study focuses primarily on the sign of coefficient β₁: if β₁ > 0, it indicates that Intergenerational Financial Dependence promotes fertility intention; if β₁ < 0, it suggests that Intergenerational Financial Dependence inhibits fertility intention; and if β₁ = 0, it implies no relationship between Intergenerational Financial Dependence and fertility intention.

Two-Stage Least Squares (2SLS) Model

As analyzed earlier, the relationship between Intergenerational Financial Dependence and fertility intention may suffer from endogeneity issues caused by selection bias and reverse causality. To address this endogeneity, this study employs a two-stage least squares (2SLS) model. The specific models are specified as follows:

First Stage:

kenglao _i = β₀ + β₁IV_i + γcontrol_i+prov_i+ɛ_i (2)

Second Stage:

child _i = β₀ + β₁kenglao_i^’ + γcontrol_i+prov_i+ɛ_i (3)

kenglao _i ^’ denotes the predicted value of Intergenerational Financial Dependence; IV_i represents the instrumental variables (IVs) for Intergenerational Financial Dependence—being an only child born after 1980, being an only child born between 1980 and 1995, and being an only child born after 1995.

The definitions of all other variables are consistent with those in the benchmark model.

Treatment Effect Model

As analyzed earlier, endogeneity arising from selection bias constitutes the primary confounding factor in identifying the causal effect between Intergenerational Financial Dependence and fertility intention.

Therefore, on the basis of the instrumental variable approach, this study further employs a treatment effect model to address this issue. The specific models are constructed as follows:

First Stage (Selection Equation):

kenglao _i = β₁IV_i + γcontrol_i + ɛ_i (4)

Second Stage:

child _i = β₀ + β₁kenglao_i + β₂imr_i + γcontrol_i+prov_i+ɛ_i (5)

In the first stage (selection equation), a Probit model is used to estimate the Inverse Mills Ratio (imr). Furthermore, the imr is incorporated as an additional control variable into the second-stage model to correct for selection bias. Except for the imr, the definitions of all other variables in the treatment effect model are consistent with those in the benchmark regression.

Stylized Facts and Empirical Results

Stylized Facts

First, a descriptive statistical analysis of the key variables in this paper is presented in Table 2. It can be observed that the mean value of fertility intention is 1.8620, with a median of 2, both leaning toward the minimum value, indicating that the overall fertility intention in China is relatively low. The mean value of intergenerational financial dependence is 0.4146, meaning 41.46% of young individuals engage in intergenerational financial dependence, which suggests that this phenomenon has become a relatively prevalent social issue. Regarding the distribution characteristics of the control variables: the gender ratio in the sample is 1:1; the proportion of urban samples is 53.29%, which is close to China’s permanent population urbanization rate in 2018 (59.58%); the mean years of education is 9.955, which is comparable to the per capita years of education at the end of the "13th Five-Year Plan" period (10.8 years) and the average years of education of the national labor force in 2018 (10.4 years) as released by the Ministry of Education. The distribution characteristics of the aforementioned control variables indicate that the sample selection in this paper is reasonable.

Table 2

Descriptive Statistics
Variables	Sample Size	Mean	Std. Dev.	Maximum	Median	Minimum
Fertility Intention	16373	1.8620	0.7172	10	2	0
Intergenerational Financial Dependence	8450	0.4146	0.4927	1	0	0
Age	18730	33.38	9.5790	49	33	16
Age Squared / 100	18730	12.06	6.45	24.01	10.89	2.56
Household Registration (Urban = 1)	16871	0.5329	0.4989	1	1	0
Gender (Male = 1)	18730	0.5014	0.5	1	1	0
Years of Education	15997	9.955	4.292	23	9	0
Health Status	18520	3.312	1.127	5	3	1
Traditional Values	16401	3.954	1.14	5	4	1
Marital Status (With Spouse = 1)	16590	0.7271	0.4455	1	1	0
Political Identity (CPC Member = 1)	18730	0.0640	0.2448	1	0	0
Endowment Insurance (Participated = 1)	16245	0.1655	0.3716	1	0	0
Number of Minor Children	17605	0.6716	0.8153	7	0	0
Household Per Capita Income	18332	9.726	1.021	15.55	9.731	0
Housing Value	18223	3.114	1.24	8.517	3.045	0
Housing Ownership (Owned = 1)	18313	0.8318	0.3741	1	1	0

Data Source

Calculated based on the 2018 China Family Panel Studies (CFPS2018) data.

Second, we compared the differences in fertility intention between individuals with and without intergenerational financial dependence, with detailed results presented in Fig. 1. It can be observed that the mean fertility intention of individuals with intergenerational financial dependence is 1.6928, while that of individuals without intergenerational financial dependence is 1.9293. The difference between the two groups is -0.2365, which is statistically significant at the 1% level. Evidently, individuals with intergenerational financial dependence have lower fertility intention. This preliminary analysis verifies that intergenerational financial dependence reduces fertility intention.

Fig. 1

Differences in Fertility Intention by Heterogeneity of NEET Dependent Behavior

Note

The sample size of individuals with NEET dependent behavior is 3,470, and that of individuals without such behavior is 4,924. The inter-group difference is statistically significant at the 1% level.

Finally, to further explore the relationship between intergenerational financial dependence and fertility intention, this paper plots the trends of both variables from 2016 to 2018, with detailed information shown in Fig. 2. It can be observed that the trends of intergenerational financial dependence and fertility intention exhibited a "scissors gap" pattern between 2016 and 2018. This suggests a potential negative correlation between intergenerational financial dependence and fertility intention, indicating that intergenerational financial dependence may be one of the important factors contributing to low fertility intention. However, whether a causal relationship exists between the two requires further verification.

Fig. 2

Trends in NEET Dependent Behavior and Fertility Intention

Note

The proportions of NEET dependent behavior in 2016 and 2018 are derived from CFPS 2016 and CFPS 2018, respectively. The fertility intention data in 2018 is from CFPS 2018. Since there is no fertility intention data available in CFPS 2016, the 2016 fertility intention is approximated by the mean value of fertility intention from CGSS 2015 and CGSS 2017.

Benchmark Regression Results

Table 3 presents the results of the benchmark regression. In Column (1), without controlling for any variables, the regression coefficient of intergenerational financial dependence on fertility intention is -0.2365, which is statistically significant at the 1% level. In Columns (2) to (4), after sequentially controlling for provincial fixed effects, individual characteristics, and household characteristics, the coefficients of intergenerational financial dependence on fertility intention change to -0.2067, -0.0376, and − 0.0395 respectively, all of which are statistically significant. These results indicate that intergenerational financial dependence reduces fertility intention.

As discussed earlier, the relationship between intergenerational financial dependence and fertility intention may suffer from endogeneity issues caused by selection bias, reverse causality, and other factors. Therefore, the conclusions of the benchmark regression should only be taken as a reference, and the interpretation of the specific causal effect shall be based on the results after addressing the endogeneity problem.

Table 3

NEET Dependent Behavior and Fertility Intention
Variables	(1)	(2)	(3)	(4)
Intergenerational Financial Dependence	-0.2365*** (0.0154)	-0.2067*** (0.0147)	-0.0376* (0.0195)	-0.0395** (0.0193)
Age			-0.0005 (0.0071)	-0.0030 (0.0072)
Age Squared / 100			0.0009 (0.0101)	0.0075 (0.0103)
Household Registration			-0.0458*** (0.0161)	-0.0371** (0.0171)
Gender			0.0431*** (0.0148)	0.0405*** (0.0149)
Years of Education			-0.0145*** (0.0026)	-0.0092*** (0.0026)
Health Status			0.0000 (0.0078)	-0.0037 (0.0078)
Traditional Values			0.0967*** (0.0078)	0.0910*** (0.0079)
Marital Status			0.2327*** (0.0266)	0.2142*** (0.0269)
Political Identity			0.0417 (0.0303)	0.0536* (0.0311)
Endowment Insurance			-0.0548** (0.0230)	-0.0417* (0.0237)
Number of Minor Children				0.0676*** (0.0107)
Household Per Capita Income				-0.0586*** (0.0103)
Housing Value				-0.0034 (0.0082)
Housing Ownership				-0.0311 (0.0226)
Constant	1.9293*** (0.0096)	1.9170*** (0.0092)	1.4892*** (0.1147)	2.0478*** (0.1422)
Provincial Fixed Effects	NO	YES	YES	YES
Observations	8394	8393	7547	7037
Adj. R²	0.0277	0.1030	0.1764	0.1966

Note

***, *, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. Robust standard errors are reported in parentheses. Unless otherwise specified, the same notation applies to all subsequent tables.

Endogeneity Treatment

The causes of endogeneity issues can be mainly summarized into several aspects, such as sample selection bias, omitted variables, reverse causality, and measurement error. As analyzed earlier, the primary endogeneity issue that may interfere with the research conclusions of this paper is selection bias, and reverse causality will also exert an impact on the conclusions. Therefore, this paper focuses on addressing endogeneity problems arising from selection bias and reverse causality, while also considering endogeneity issues from other aspects.

Instrumental Variable (IV) Approach

Table 4 presents the empirical results of the instrumental variable (IV) method. Columns (1) and (2) report the results using the only-child status after 1980 as the instrumental variable. It can be observed that the F-statistic is 9.959 (close to 10), and the p-value of the Anderson-Rubin test is 0.009 (less than 0.01), indicating that the instrumental variable in this paper does not suffer from the weak instrument problem. The LM statistic is 9.843, which is statistically significant at the 1% level, rejecting the null hypothesis of underidentification. The regression coefficient of IV1 (only-child status after 1980) on intergenerational financial dependence is 0.0862, significant at the 1% level, which is consistent with theoretical expectations, confirming the validity of the instrumental variable. In the second stage, the regression coefficient of intergenerational financial dependence on fertility intention is -1.0361, significant at the 5% level, which is consistent with the benchmark regression results. This indicates that the conclusion of this paper remains robust after addressing endogeneity issues using the instrumental variable method.

Furthermore, this paper constructs two additional instrumental variables for intergenerational financial dependence: only-child status during 1980–1995 and only-child status after 1995. The rationale is that on April 20, 1994, China officially connected to the Internet with full functionality, marking the country’s entry into the Internet era. The social environment and lifestyle in the Internet era have undergone profound changes, which have exerted a significant impact on the mindset and behavioral patterns of individuals who grew up entirely in the Internet environment (i.e., those born after 1995). Columns (3) and (4) present the results using these two instrumental variables. It can be seen that the LM statistic is 11.509, significant at the 1% level, passing the underidentification test. The F-statistic exceeds 10, indicating the absence of the weak instrument problem. The Hansen J statistic is 0.193 with a p-value of 0.660, passing the overidentification test, which confirms the exogeneity of the instrumental variables. The regression coefficients of IV2 (only-child status during 1980–1995) and IV3 (only-child status after 1995) on intergenerational financial dependence are 0.0852 and 0.2621, respectively, both significant at the 1% level and consistent with theoretical expectations. Meanwhile, the first-stage regression coefficients of the instrumental variables show that only children born after 1995 are more likely to engage in intergenerational financial dependence than those born during 1980–1995, which is consistent with practical realities and further corroborates the validity of the instrumental variables. In summary, the instrumental variables used in this paper are effective. In the second stage, the regression coefficient of intergenerational financial dependence on fertility intention is -1.0714, significant at the 5% level, which is consistent with the benchmark regression results. This reconfirms the robustness of the research conclusions.

Table 4

Endogeneity Test: Instrumental Variable Approach
Variables	First Stage	Second Stage	First Stage	Second Stage
Variables	(1)	(2)	(3)	(4)
Intergenerational Financial Dependence		-1.0361** (0.4952)		-1.0714** (0.4955)
IV1	0.0862*** (0.0273)
IV2			0.0852*** (0.0274)
IV3			0.2621*** (0.0417)
Constant	2.4074*** (0.1175)	4.2968*** (1.1985)	2.4070*** (0.1175)	4.3800*** (1.1993)
Individual Characteristic Variables	YES	YES	YES	YES
Household Characteristic Variables	YES	YES	YES	YES
Provincial Fixed Effects	YES	YES	YES	YES
LM Statistic & p-value	9.843*** [0.002]		11.509*** [0.003]
Kleibergen-Paap rk Wald F Statistic	9.959		23.199
Anderson-Rubin Wald Test	0.009		0.026
Hansen J Statistic & p-value			0.193 [0.660]
Observations	7037

Note

IV1 denotes being an only child born after 1980. IV2 denotes being an only child born between 1980 and 1995. IV3 denotes being an only child born after 1995. p-values are reported in square brackets. The same applies to the following tables.

Sample Selection Bias Mitigation

To mitigate endogeneity issues arising from sample selection bias, this paper employs the PSM-OLS (Propensity Score Matching - Ordinary Least Squares) method for verification. The specific steps are as follows:

First, samples with intergenerational financial dependence are designated as the treatment group, while those without are assigned to the control group. All characteristic variables from the benchmark regression, along with occupation type and personal income, are used as matching covariates. A one-to-one nearest neighbor matching approach is adopted to identify suitable control group samples for the treatment group. Second, empirical tests are conducted using the successfully matched samples, with results presented in Column (1) of Table 6. It can be observed that the regression coefficient of intergenerational financial dependence is -0.0671, which is statistically significant at the 5% level. This indicates that after excluding the interference of sample selection bias, intergenerational financial dependence still exerts a suppressing effect on fertility intention, consistent with the findings of the benchmark regression. Thus, the research conclusions of this paper are robust.

The reliability of the conclusions derived from the PSM-OLS method depends on the validity of PSM, which in turn hinges on whether the matched treatment and control groups satisfy the balance test. Therefore, this paper performs a balance test on the PSM results, with detailed findings reported in Table 5 and Fig. 3.

From the balance test results in Table 5: Before matching, there are significant differences in characteristic variables—including age, age squared divided by 100, household registration type, years of education, traditional values, marital status, number of minor children, housing value, housing property rights, occupation type, and personal income—between the treatment and control groups. After matching, however, the differences in all characteristic variables between the two groups are no longer statistically significant. This indicates a good matching effect, which enhances the credibility of the research conclusions.

Table 5

Balance Test Results
Variables	Pre-Matching			Post-Matching
Variables	Treatment Group	Control Group	T-Statistic	Treatment Group	Control Group	T-Statistic
Age	30.531	36.085	-17.70***	30.562	31.129	-1.49
Age Squared / 100	10.042	13.7	-16.81***	10.06	10.4	-1.37
Household Registration	0.609	0.561	2.59**	0.608	0.625	-0.79
Gender	0.594	0.600	-0.32	0.595	0.615	-0.92
Years of Education	11.744	10.5	8.69***	11.739	11.792	-0.34
Health Status	3.288	3.263	0.64	3.290	3.269	0.45
Traditional Values	3.884	3.986	-2.39**	3.888	3.845	0.83
Marital Status	0.600	0.792	-11.78***	0.602	0.624	-1.02
Political Identity	0.097	0.116	-1.62	0.097	0.090	0.54
Endowment Insurance	0.263	0.244	1.16	0.264	0.296	-1.60
Number of Minor Children	0.525	0.676	-5.16***	0.527	0.498	0.89
Household Per Capita Income	10.009	9.970	1.34	10.008	10.036	-0.79
Housing Value	3.339	3.231	2.33**	3.339	3.349	-0.18
Housing Ownership	0.870	0.843	2.00**	0.870	0.876	-0.40
Occupation Type	1.137	1.206	-3.17***	1.138	1.148	-0.44
Personal Income	9.849	10.325	-8.33***	9.879	9.757	1.18

Based on the standardized bias of covariates before and after matching presented in Fig. 3, the standardized bias of all covariates falls within 10% after matching, which is acceptable. Overall, the samples of the treatment group and the control group after matching satisfactorily meet the balance test requirements, further verifying the validity of the PSM results.

Fig. 3

Standardized Bias of Covariates Before and After Matching

Treatment Effect Model

This paper also employs a treatment effect model to correct for selection bias that may interfere with the research conclusions, with the empirical results presented in Column (2) of Table 6. It can be observed that the regression coefficient of the inverse Mills ratio (IMR) is significantly positive at the 5% level, indicating that the research conclusions of this paper are indeed affected by selection bias. After correcting for selection bias, the regression coefficient of intergenerational financial dependence on fertility intention is -1.0793, which is statistically significant at the 1% level. This confirms that intergenerational financial dependence still reduces fertility intention even after addressing the interference of selection bias.

Omitted Variable Bias

Considering that both occupation type and personal income are correlated with intergenerational financial dependence, and they also affect fertility intention. However, due to the high proportion of missing values in occupation type and personal income, including them in the benchmark regression model may lead to biased research conclusions, while omitting these two variables may result in omitted variable bias. To prevent the interference of omitted variable bias on the research conclusions, in the endogeneity test, this paper sequentially incorporates occupation type and the logarithm of personal income into the regression model for control, with the results presented in Columns (3) and (4) of Table 6.

It can be observed that after sequentially controlling for occupation type and the logarithm of personal income, the coefficients of intergenerational financial dependence change to -0.0379 (statistically significant at the 10% level) and − 0.0515 (statistically significant at the 5% level) respectively. The research conclusions remain unchanged after controlling for these variables, thus eliminating the interference of omitted variable bias.

Reverse Causality

Reverse causality may affect the research conclusions of this paper. The rationale is that individuals with low fertility intention are more likely to exhibit a "Buddhist-style" mindset in work, employment, and other aspects, thereby being more prone to engaging in intergenerational financial dependence. To rule out the impact of reverse causality, this paper reconstructs the intergenerational financial dependence variable using data from the 2016 CFPS and re-examines its effect on fertility intention, with the results presented in Column (5) of Table 6.

It can be observed that the regression coefficient of intergenerational financial dependence lagged by two periods on fertility intention is -0.0409, which is statistically significant at the 5% level. After accounting for reverse causality, the research conclusions of this paper remain unchanged, thus eliminating the interference of reverse causality. Meanwhile, this indicates that the suppressing effect of intergenerational financial dependence on fertility intention has a sustained impact in the short term.

Measurement Error

Since the core explanatory variable—intergenerational financial dependence—is closely related to two variables: the amount of financial support individuals provide to their parents and the amount they receive from their parents. Both variables are derived from survey responses, so the research conclusions of this paper may be affected by measurement errors (e.g., memory bias, misreporting). To eliminate the interference of measurement errors on the conclusions, this paper draws on the approach of existing studies (Zhou & Zhang, 2021) by excluding samples where the respondents’ comprehension ability score is less than 4 (with scores ranging from 1 to 7) and conducting empirical tests using the remaining samples. The results are presented in Column (6) of Table 6.

In addition, considering that individuals who have withdrawn from the labor market or are unemployed are more likely to misreport, leading to measurement errors, this paper also excludes samples of individuals who have withdrawn from the labor market or are unemployed to mitigate such errors, with the results shown in Column (7) of Table 6. It can be observed that regardless of the method used to exclude measurement errors, the regression coefficient of intergenerational financial dependence on fertility intention remains significantly negative at least at the 10% level. The above results indicate that the research conclusions of this paper remain valid after minimizing the impact of measurement errors.

Table 6

Endogeneity Tests
Variables	PSM-OLS Model	Treatment Effect Model	Adding Control Variables		Lagged by Two Periods	Excluding Samples with Low Comprehension Scores	Excluding Samples Who Exited the Labor Market
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Intergenerational Financial Dependence	-0.0671**(0.0301)	-1.0793*** (0.4099)	-0.0379* (0.0213)	-0.0515** (0.0234)		-0.0407** (0.0194)	-0.0412* (0.0217)
L2. Intergenerational Financial Dependence					-0.0409** (0.0201)
Constant	1.8914***(0.3684)	2.4225*** (0.2115)	2.2160*** (0.1964)	2.1063*** (0.2403)	2.2525*** (0.1832)	1.9708*** (0.1424)	2.5572*** (0.2081)
imr		0.6488** (0.2558)
Individual Characteristic Variables	YES	YES	YES	YES	YES	YES	YES
Household Characteristic Variables	YES	YES	YES	YES	YES	YES	YES
Provincial Fixed Effects	YES	YES	YES	YES	YES	YES	YES
Occupation Type	YES	YES	YES	YES	NO	NO	NO
Personal Income	YES	YES	NO	YES	NO	NO	NO
Observations	1367	7037	5084	3400	5433	6724	4980
Adj_R²	0.1867	0.1972	0.1931	0.1726	0.1941	0.1915	0.1928

In summary, the research conclusions of this paper remain robust after addressing endogeneity issues caused by selection bias, reverse causality, and other factors. This validates the causal effect between intergenerational financial dependence and fertility intention, as well as the proposed research hypothesis H1.

Robustness Checks

Model Replacement

Based on the characteristics of the data, this paper employs the following two models for robustness tests:

First, fertility intention is treated as a count variable and estimated using a Poisson regression model, with the results presented in Column (1) of Table 7. Second, fertility intention is regarded as censored data (greater than or equal to 0) and estimated using a Tobit model, with the results shown in Column (2) of Table 7.

In Column (1), the regression coefficient of intergenerational financial dependence on fertility intention is -0.0213, which is statistically significant at the 5% level. In Column (2), the corresponding coefficient is -0.0404, also significant at the 5% level. The conclusions from the robustness tests using alternative models are consistent with those of the benchmark regression, indicating that the research conclusions of this paper are robust.

Excluding Student Samples

Considering that some individuals in the sample are still in school and thus face difficulties achieving economic independence, classifying them as part of the intergenerational financial dependence group would lack fairness and may lead to an overestimation of the research conclusions. Therefore, we excluded samples of individuals who are still in school for a robustness test, with the results presented in Column (3) of Table 7.

It can be observed that after excluding the in-school samples, the regression coefficient of intergenerational financial dependence on fertility intention is -0.0509, which is statistically significant at the 5% level. This result is consistent with that of the benchmark regression, indicating that the research conclusions of this paper are robust.

Redefining the Dependent Variable

Based on the data, the total proportion of individuals with a fertility intention of 3 or higher is less than 10%. Additionally, fertility intention reflects people’s attitudes toward childbearing—when individuals’ fertility intention exceeds 3, there may be no substantial difference in their attitudes toward childbearing. Therefore, this paper redefines fertility intention for a robustness test: a fertility intention of 0 is categorized as low fertility intention (coded 0); 1 as relatively low fertility intention (coded 1); 2 as moderate fertility intention (coded 2); and 3 or higher as high fertility intention (coded 3). An ordered logit model is employed for empirical testing, with the results presented in Column (4) of Table 7.

It can be observed that the regression coefficient of intergenerational financial dependence on the redefined fertility intention is -0.2245, which is statistically significant at the 1% level. The consistency between this result and the benchmark regression conclusion—even after redefining the dependent variable—indicates that the research conclusions of this paper are robust.

Table 7

Robustness Checks
Variables	Poisson Regression	Tobit Regression	Excluding Student Samples	Ordered Logit Model
Variables	(1)	(2)	(3)	(4)
Intergenerational Financial Dependence	-0.0213** (0.0106)	-0.0404** (0.0192)	-0.0509** (0.0202)	-0.2245*** (0.0697)
Constant	0.6290*** (0.0897)	1.8985*** (0.1662)	2.3898*** (0.1948)
Individual Characteristic Variables	YES	YES	YES	YES
Household Characteristic Variables	YES	YES	YES	YES
Provincial Fixed Effects	YES	YES	YES	YES
Observations	7037	7037	5557	7037
Pseudo R²/ Adj_R²	0.0187	0.1064	0.1934	0.1348

In summary, after the aforementioned series of robustness checks, the research conclusions of this study remain robust, verifying their reliability.

Mechanism Tests

After robustly establishing the causal effect between intergenerational financial dependence and fertility intention in the preceding analysis, this paper proceeds to explore the underlying mechanism through which intergenerational financial dependence influences fertility intention. Based on the theoretical analysis earlier, the intrinsic mechanism by which intergenerational financial dependence reduces fertility intention lies in its negative impact on fertility benefits. Specifically, fertility benefits are mainly reflected in two aspects: first, having children can provide security for one’s old age (i.e., "raising children for old-age support"); second, having children can increase household labor output, namely providing financial assistance to the family (hereinafter referred to as "financial support"). Therefore, this paper characterizes fertility benefits from the above two dimensions. Considering data availability, the aforementioned quantitative indicators are derived from the 2020 CFPS data. The rationale is as follows: as a household panel survey, the 2020 CFPS is relatively close in time to the 2018 wave, with a lower sample turnover rate compared to other years, which can minimize missing values after matching. Meanwhile, using the 2020 CFPS data helps avoid the interference of reverse causality. Additionally, the widespread influence of traditional concepts such as "raising children for old-age support" and "more people mean greater strength" on individuals in China can, to a certain extent, mitigate the impact of selection bias on the research conclusions. In summary, using the 2020 CFPS data to measure fertility benefits enables a reliable identification of the causal relationship between intergenerational financial dependence and fertility benefits. Among the indicators: "raising children for old-age support" is measured using the questionnaire item "Having children is to have someone to help me when I am old"; "financial support" is operationalized with the item "Having children is to provide financial assistance to the family". Respondents’ answers are coded from 1 to 5, ranging from "strongly disagree" to "strongly agree".

Furthermore, the regression model constructed to test Research Hypothesis H2 is as follows:

M _i = β₀ + β₁kenglao_i + β₂imr_i + γcontrol_i+prov_i+ɛ_i (6)

Where M_i denotes the mechanism variables—"raising children for old age security" and "financial support"—and the definitions of all other variables are consistent with those in the benchmark regression.

Table 8 presents the empirical results of the mechanism test. After excluding the interference of confounding factors, it can be observed that the regression coefficient of intergenerational financial dependence on "raising children for old-age support" is -0.0527, which is statistically insignificant. In contrast, the regression coefficient of intergenerational financial dependence on "financial support" is -0.0843, statistically significant at the 5% level.

These empirical results indicate that intergenerational financial dependence significantly reduces individuals’ perception that children can provide financial assistance to the family, but there is no empirical evidence to support that intergenerational financial dependence significantly weakens individuals’ belief in "raising children for old-age support." This finding also suggests that the reduction in the immediate benefits of having children caused by intergenerational financial dependence is the primary mechanism underlying the decline in fertility intention. In summary, Research Hypothesis H2 is supported.

Table 8

Mechanism Test Results
Variables	Raising Children for Old Age Security		Financial Support
Variables	(1)	(2)	(3)	(4)
Intergenerational Financial Dependence	-0.2157*** (0.0329)	-0.0527 (0.0377)	-0.2516*** (0.0332)	-0.0843** (0.0382)
Constant	3.6524*** (0.0177)	4.2357*** (0.3935)	2.9310*** (0.0194)	4.4129*** (0.3956)
Individual Characteristic Variables	NO	YES	NO	YES
Household Characteristic Variables	NO	YES	NO	YES
Provincial Fixed Effects	NO	YES	NO	YES
Observations	4668	3938	4663	3933
Adj_R²	0.0367	0.1330	0.0478	0.1869

Heterogeneity Analysis

As noted earlier, intergenerational financial dependence is associated with individual characteristics, and thus differences in individual traits may lead to heterogeneous effects of intergenerational financial dependence on fertility intention. Therefore, this paper examines the heterogeneous impacts of intergenerational financial dependence on fertility intention from the dimensions of individual characteristics such as political identity, income status, and educational attainment. In addition, considering that differences in the causes of intergenerational financial dependence may result in varying consequences, this paper also explores the heterogeneous effects of different types of intergenerational financial dependence on fertility intention.

Heterogeneity by Political Identity

This paper divides the full sample into Party member and non-Party member subsamples based on political identity and conducts subgroup regression analysis. The empirical results are presented in Columns (1) and (2) of Table 9.

It can be observed that in the non-Party member subsample, the regression coefficient of intergenerational financial dependence on fertility intention is -0.0349, which is statistically significant at the 10% level. In contrast, the corresponding coefficient in the Party member subsample is -0.0866, statistically insignificant. This indicates that intergenerational financial dependence significantly reduces the fertility intention of non-Party members. The rationale is that political identity serves as an indicator of social capital—individuals with Party membership to a certain extent possess higher social capital. Higher social capital reduces the likelihood of individuals engaging in intergenerational financial dependence, thereby resulting in an insignificant negative impact of intergenerational financial dependence among Party members.

Heterogeneity by Income Status

This paper defines individuals with personal income above the median as high-income individuals and those with income below the median as low-income individuals, followed by subgroup regression analysis. The empirical results are presented in Columns (3) and (4) of Table 9.

It can be observed that in the low-income group, the regression coefficient of intergenerational financial dependence on fertility intention is -0.559, which is statistically significant at the 10% level. In contrast, the regression coefficient between intergenerational financial dependence and fertility intention in the high-income group is -0.0533, statistically insignificant. These results indicate that intergenerational financial dependence significantly reduces the fertility intention of low-income individuals. The rationale is that compared with high-income groups, low-income groups are more likely to rely on financial support from their families, making them more prone to engaging in intergenerational financial dependence. Consequently, they are more significantly affected by the negative impacts of intergenerational financial dependence.

Heterogeneity by Educational Attainment

This paper uses college education (junior college) as the cutoff point for educational attainment, defining individuals with a college degree or higher as the high-education subsample and those with education below college level as the low-education subsample. Subgroup regression analysis is conducted, with the empirical results presented in Columns (5) and (6) of Table 9.

It can be observed that in the low-education subsample, the regression coefficient of intergenerational financial dependence on fertility intention is -0.0425, which is statistically significant at the 10% level. In contrast, the corresponding coefficient in the high-education subsample is -0.0397, statistically insignificant. These results indicate that intergenerational financial dependence reduces the fertility intention of low-education groups. The rationale is that in the labor market, higher educational attainment signals positive attributes, enabling individuals to gain an advantage in fierce social competition. In contrast, individuals with lower educational attainment struggle to secure such advantages in intense competition, thereby being compelled to seek financial support from their parents and engage in intergenerational financial dependence. Consequently, they are more severely affected by the negative impacts of intergenerational financial dependence.

In summary, intergenerational financial dependence significantly reduces the fertility intention of non-Party members, low-income individuals, and those with low educational attainment—indicating that intergenerational financial dependence exerts a significant negative impact on these three groups. Notably, all three groups share a common characteristic: they are at a disadvantage in fierce market competition. This result suggests that fertility benefits are a crucial consideration for disadvantaged groups when making fertility decisions.

Table 9

Heterogeneity Test Results
Variables	Political Identity		Income Status		Educational Attainment
	Party Member	Non-Party Member	Low-Income	High-Income	Low-Education	High-Education
	(1)	(2)	(3)	(4)	(5)	(6)
Intergenerational Financial Dependence	-0.0866 (0.0830)	-0.0349* (0.0198)	-0.0559* (0.0319)	-0.0533 (0.0349)	-0.0425* (0.0229)	-0.0397 (0.0255)
Constant	2.5479*** (0.5767)	2.0161*** (0.1465)	1.9336*** (0.3075)	2.3577*** (0.4240)	2.2269*** (0.1575)	1.8120*** (0.4464)
Individual Characteristic Variables	YES	YES	YES	YES	YES	YES
Household Characteristic Variables	YES	YES	YES	YES	YES	YES
Provincial Fixed Effects	YES	YES	YES	YES	YES	YES
Observations	516	6521	1816	1583	5681	1356
Adj_R²	0.1067	0.2050	0.1634	0.1774	0.2034	0.1449

Heterogeneity by Type of Intergenerational Financial Dependence

In addition to discussing the heterogeneous impacts of intergenerational financial dependence on fertility intention arising from differences in individual traits, this paper also explores the effects of different types of intergenerational financial dependence on fertility intention. Considering data availability, this paper classifies the types of intergenerational financial dependence based on the frequency of individuals’ face-to-face interactions with their parents per week: those who engage in intergenerational financial dependence and meet their parents three or more times a week are defined as active intergenerational financial dependence, while the rest are categorized as passive intergenerational financial dependence.

Furthermore, subgroup regression analysis is conducted for empirical testing, with the results presented in Table 10. After excluding the interference of confounding factors, it can be observed that in the active intergenerational financial dependence group, the regression coefficient of intergenerational financial dependence on fertility intention is -0.0702, which is statistically significant at the 5% level. In contrast, the corresponding coefficient in the passive intergenerational financial dependence group is 0.0163, statistically insignificant. These results indicate that active intergenerational financial dependence reduces fertility intention.

The rationale is that individuals with active intergenerational financial dependence are more likely to directly perceive that having children cannot provide financial assistance to the family, thus being more significantly affected by the decline in fertility benefits.

Table 10

Heterogeneous Effects of Different Types of Intergenerational Financial Dependence on Fertility Intention
Variables	Active Intergenerational Financial Dependence		Passive Intergenerational Financial Dependence
Variables	(1)	(2)	(3)	(4)
Intergenerational Financial Dependence	-0.2130*** (0.0253)	-0.0702** (0.0291)	-0.1666*** (0.0215)	0.0163 (0.0318)
Constant	1.9130*** (0.0187)	1.8529*** (0.2314)	1.8721*** (0.0133)	1.6748*** (0.2044)
Individual Characteristic Variables	NO	YES	NO	YES
Household Characteristic Variables	NO	YES	NO	YES
Provincial Fixed Effects	NO	YES	NO	YES
Observations	2700	2422	3794	2985
Adj_R²	0.1316	0.2319	0.0703	0.1668

Conclusions and Policy Implications

Conclusions

Maintaining an appropriate fertility level and population size is a key measure for high-quality population development to support Chinese-style modernization, and the issue of low fertility has become one of the important concerns of the Party and the country. As people’s attitude toward childbearing, fertility intention can reflect the future trend of fertility rates. How to effectively address the problem of low fertility intention urgently requires in-depth exploration by researchers. Therefore, against the realistic backdrop of comprehensively advancing the great rejuvenation of the Chinese nation through Chinese-style modernization, identifying the causes of low fertility intention holds significant theoretical and practical significance.

Another social phenomenon coexisting with low fertility intention is the annual expansion of the "intergenerational financial dependence group." Is there a correlation between the growth of this group and the decline in fertility intention? Does a causal relationship exist between intergenerational financial dependence and low fertility intention? These questions warrant in-depth investigation.

Using data from the China Family Panel Studies (CFPS), this paper examines the impact of intergenerational financial dependence on fertility intention. The main conclusions are as follows:

(1) Intergenerational financial dependence reduces fertility intention, and this conclusion remains valid after a series of robustness tests.

(2) Mechanism tests reveal that the underlying mechanism through which intergenerational financial dependence lowers fertility intention is the reduction in the benefits of having children.

(3) Heterogeneity analysis indicates that the suppressing effect of intergenerational financial dependence on fertility intention primarily acts on disadvantaged groups in market competition, such as non-Party members, low-income individuals, and those with low educational attainment. Additionally, this suppressing effect is mainly driven by active intergenerational financial dependence.

The research conclusions of this paper provide a potential explanation for the decline in fertility intention from the perspective of child-related benefits, offering important insights for effectively addressing the severe "low fertility" phenomenon in China at the current stage.

Policy Implications

Based on the above conclusions and combined with China’s actual national conditions, this paper puts forward the following policy recommendations to address the decline in fertility intention: First, establish a pension and retirement system linked to the number of children. This study finds that the decline in fertility intention is caused by the reduction in child-related benefits. Therefore, by establishing a pension and retirement system tied to the number of children, we can connect retirement and elderly care with childbearing, promote the integration of social elderly care and family elderly care, and drive the return of elderly care responsibilities to families. This will endow individuals with reasonable rights to benefits from childbearing, thereby enhancing people’s fertility intention.

Second, improve the institutional mechanisms in the labor market. Heterogeneity analysis shows that intergenerational financial dependence significantly reduces the fertility intention of groups at a disadvantage in market competition. Therefore, we can alleviate the competitive disadvantage of these groups by providing employment training, smoothing their access to employment information, offering employment assistance, and promoting the integration of education and industry. This will help mitigate the negative impacts of intergenerational financial dependence.

Third, unblock channels for social mobility. Based on the heterogeneous effects of different types of intergenerational financial dependence on fertility intention, it is evident that the negative impact of intergenerational financial dependence on fertility intention is mainly driven by active intergenerational financial dependence. In China’s current society, an important reason for active intergenerational financial dependence is the lack of smooth social mobility channels and the prevalence of involution. Therefore, it is necessary to unblock social mobility channels to avoid the negative impacts of active intergenerational financial dependence. Meanwhile, we should conduct positive public opinion guidance using internet technology, foster correct values, and eliminate active intergenerational financial dependence at the ideological level.

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reason able request.

Author Contribution

Li Shun: Conceptualization, Original draft. Xu Lu: Editing, validation, and final draft.

Acknowledgement

Funding was provided by Excellent Youth Project of the Education Department of Hunan Province (24B0673) and the Youth Project of the Natural Science Foundation of Hunan Province (2025JJ60470).

References

Bansak C, Dziadula E, Wang SX (2025) The effect of family size on intergenerational support for elderly parents and their life expectancy: Evidence from China’s Later, Longer, Fewer campaign. China Econ Rev, 102528

Becker GS, Lewis HG (1973) On the Interaction between the Quantity and Quality of Children. J Polit Econ 81(2):S279–S288

Bethencourt C, Santos D (2025) Gender-role identity in adolescence and women fertility in adulthood. J Econ Behav Organ 240:107250

Caldwell JC, Toward A (1976) Restatement of Demographic Transition Theory[J]. Popul Dev Rev 2(3):321–366

Chen S, Xie B (2025) The abolition of People’s Communes and fertility decline in rural China. J Dev Econ 172:103375

Daysal NM, Lovenheim M, Siersbæk N, Wasser DN Home Prices, Fertility, and Early-life Health Outcomes[J]. J Public Econ, 2019(198): 1–17

Farré L, González L Does Paternity Leave Reduce Fertility?[J]. J Public Econ, 2019(172):52–66

Girsberger EM, Hassani-Nezhad L, Karunanethy K, Lalive R (2023) Mothers at work: How mandating a short maternity leave affects work and fertility. Labour Econ 84:102364

He XF (2009) On Intergenerational Relationships in Rural China: With a Discussion on the Value Foundation of Intergenerational Relations [J]. Soc Sci Res, (05): 84–92

10.

Jensen R (2012) Do Labor Market Opportunities Affect Young Women's Work and Family Decisions? Experimental Evidence from India[J]. Q J Econ 127(02):753–792

11.

Leibenstein H (1974) An Interpretation of the Economic Theory of Fertility: Promising Path or Blind Alley? J Econ Lit 12(2):457–479

12.

Leibenstein H (1974) An Interpretation of the Economic Theory of Fertility: Promising Path or Blind Alley?[J]. J Econ Lit 12(2):457–479

13.

Li M, Li C, Li Q (2025) From tradition to transition: The effect of clan culture on energy poverty. Energy Policy 206:114777

14.

Lin S, Luo N (2025) Widening Income Inequality and the Decline in Fertility Intentions: Micro Evidence from Chinese Households. International Review of Economics & Finance, p 104723

15.

Liu J, Xing CB, Zhang Q House Price, Fertility Rates and Reproductive Intentions[J]. China Econ Rev, 2020(60): 1–16

16.

Liu W (2016) Intergenerational Affection and Family Solidarity in Transitional Period: A Comparative Study of Two Types of Parent-Dependent Families in Shanghai [J]. Sociol Stud 31(04):145–168

17.

Meng L, Peng L, Zhou Y (2023) Do housing booms reduce fertility intentions? Evidence from the new two-child policy in China. Reg Sci Urban Econ 101:103920

18.

Nie P, Peng X, Luo T (2023) Internet use and fertility behavior among reproductive-age women in China. China Econ Rev 77:101903

19.

Shipman JE, Swanquist QT, Whited RL (2017) Propensity Score Matching Acc Research[J] Acc Rev 92(1):213–244

20.

Ssewamala FM (2015) Optimizing the demographic dividend in young developing countries: The role of contractual savings and insurance for financing education. Int J social Welf 24(3):248–262

21.

Taketoshi K (2020) How and to what extent has the demographic dividend affected China’s economic growth? Int J Economic Policy Stud 14(2):337–350

22.

Yang F, Yu L (2016) With or without siblings: Sorting into competition in the experimental labor market. China Econ Rev 41:284–298

23.

Zha J, Yan L, Wang Y, He L, Xu D, Yu Y (2025) Demystifying intergenerational support from adult children in senior tourism consumption: Insights from large-scale household level survey data. Tourism Manage Perspect 59:101398

24.

Zhan Y, Zhan X, Wu M (2025) From Financial Assistance to Emotional Care: The Impact of Intergenerational Support on the Subjective Well-Being of China’s Elderly. J Happiness Stud 26(2):1–28

25.

Zhang C (2019) Family support or social support? The role of clan culture. J Popul Econ 32(2):529–549

26.

Zhang H, Yang Y, Sun S (2024) It is not a vacation of leisure: The effect of maternity-leave stigma on fertility intentions in China. Acta Psychol 248:104413

27.

Zhang J, Li J, Li C, Wang J (2025) The impact of intergenerational support on older adults’ life satisfaction in China: Rural-urban differences. Healthc Rehabilitation 1(1):100011

28.

Zhang J, Li X, Tang J (2022) Effect of public expenditure on fertility intention to have a second child or more: Evidence from China's CGSS survey data. Cities 128:103812

29.

Zhang YY, Luh YH (2018) Grandparents' Health and Family Fertility Choice: Evidence from Taiwan[J]. China Econ Rev, (51): 294–308

30.

Zhou S, Zhang WT (2021) Analysis of the Subjective Well-being Effect of Internet Usage [J]. Econ Res J 56(09):158–174

Yes

Individuals with zero siblings are defined as only children and coded 1; otherwise, they are classified as non-only children and coded 0.

Due to data limitations in other waves of the CFPS, either data on fertility intention or intergenerational financial dependence is unavailable. Therefore, the empirical analysis of this paper primarily relies on data from the 2018 CFPS.

While both the 2010 and 2014 waves of the CFPS provide information related to siblings, the 2014 CFPS only includes data on whether siblings are deceased, their occupation types, and educational attainment. In contrast, the 2010 CFPS offers more detailed information such as the number of siblings and their ages. Therefore, the number of siblings in this paper is derived from the 2010 CFPS data. The use of sibling number data from the 2010 CFPS is reasonably justified for two main reasons. First, China implemented a strict one-child policy in 1980, which was not adjusted until after 2014. Second, among the 2010 CFPS sample (individuals aged 16–49), over 90% of their parents were over the childbearing age (49 years old). Consequently, the likelihood of individuals in the sample having new siblings between 2010 and 2018 was relatively low.

⁴

The null hypothesis of the Anderson-Rubin test is that "the sum of the endogenous regression coefficients equals zero." Rejecting the null hypothesis indicates that the instrumental variable has a strong correlation with the endogenous explanatory variable, and there is no weak instrument problem.

⁵

Occupation type is measured using the variable "current main job/most recently ended job type" from the CFPS questionnaire, with the following coding: 1 for non-agricultural casual work or wage employment, 2 for agricultural wage labor, and 3 for self-employment or family agricultural production and operation. Personal income is operationalized using "total taxable wage income from all jobs (main job + part-time jobs) in the past 12 months" from the CFPS questionnaire, and the natural logarithm of personal income is taken for data processing.

⁶

The reason for including occupation type and personal income as matching covariates here is to match the treatment group with control group samples that have as similar characteristics as possible, thereby eliminating sample selection bias caused by observable factors.

⁷

The reason for including occupation type and personal income as control variables in this model is to prevent changes in the sample structure caused by variations in matching covariates from confounding the research conclusions (Shipman et al., 2017).