Economic Uncertainty, Cultural and Ideational Transition, and Low Fertility

Abstract

Increased objective uncertainties, upward movement in the hierarchy of needs, and associated cultural and ideational transition are inherent to modern societies. These factors were previously treated as independent macro-shocks and studied separately, without regard for their interactions. In this paper, we provide an all-around framework to interpret fertility behavior and low fertility in developed economies, to compensate for the isolation of economic uncertainty from a cultural and ideational transition in previous empirical studies. In this regard, we conduct an empirical analysis of panel data of 34 OECD countries from 2000 to 2018, to discuss the impact of economic uncertainty on the fertility rate and the moderating effect of cultural and ideational transition on that impact. Below are our findings: (1) economic uncertainty significantly inhibits the fertility rate, and such an inhibiting effect is found to be underestimated after endogeneity is controlled; (2) according to heterogeneity analysis, the inhibiting effect of economic uncertainty on the fertility rate is stronger after the 2008 financial crisis and among low-income economies and countries where Confucianism is practiced; (3) a significant negative moderating effect of cultural and ideational transition on the relationship between economic uncertainty and fertility rate is observed, indicating that the inhibiting effect of structural dimensions that combine objective and subjective factors regarding the fertility rate may be self-reinforcing; and, (4) further tests show that economic uncertainty and cultural and ideational transition affect the fertility rate by means of the effect of delayed parenthood, the substitution of cohabitation for marriage, and fertility preferences. We find that fertility behavior is cumulatively affected by both economic uncertainty and cultural and ideational transition. This implies that reducing economic uncertainty and fostering a culture that encourages marriage and fertility are fundamental for increasing the fertility rate in China, a country resorting to the third-child policy to promote a fertility rebound.

1. Introduction

According to the 7th National Population Census, China’s TFR (total fertility rate) was 1.3 in 2020, and the number of newborns hit a record low of 10.62 million in 2021, as revealed by the National Bureau of Statistics, drawing wide concern regarding the dropping fertility rate in China.

The decrease in China’s fertility rate is believed to be the result of family planning policies and socioeconomic development. The decrease in the number of new births is attributable to factors such as a decline in the number of women of childbearing age, delayed entry into parenthood, and the COVID-19 pandemic. Adjustments to family planning policies since 2014 are undeniably unsatisfactory in increasing the overall fertility level. This indicates that the key to increasing the fertility rate lies in socioeconomic drivers rather than in China’s family planning policies, which no longer work.

It underlines the value of interpreting China’s low fertility and the reasons behind it, in light of general trends and international practices. Reasons for China’s low fertility rate were generally classified into economic and institutional changes that impose objective restrictions and economic pressures on young people when making decisions regarding marriage and childbearing, on the one hand, and ideological or ideational changes that were deemed to be external shocks, independent of economic fluctuations, on the other hand. It has been assumed that the change in individuals’ needs induces an individualism that shapes fertility values and behaviors in return. Regrettably, these factors were isolated in previous studies, where neither a unified theoretical framework nor a holistic perspective was found. Thus, they did not convincingly explain the low fertility in the new era.

For this reason, in the context of increased economic uncertainty and cultural and ideational transition, we draw our findings from 34 OECD member states to test the inhibiting effect of economic uncertainty on fertility, analyze the heterogeneity of the impact of economic uncertainty on fertility in different economic and cultural contexts, and verify the reinforcing effect of cultural and ideational transition on such an inhibiting effect. We also discuss the possible mechanisms of how economic uncertainty affects fertility, including the timing of entry into parenthood, marriage stability, and the percentage of childless women.

By doing so, we hope to develop a more inclusive analytical framework that links economic uncertainty with cultural and ideational transition, to explain fertility behaviors with macroscopic evidence. In comparison with previous studies, the contributions of this paper are as follows. Firstly, economic uncertainty and cultural and ideational transition are originally put under the same analysis framework, to reveal their interactions and cumulative effects on fertility, which is compensation for and an expansion of previous frameworks. Secondly, the entropy evaluation method and arithmetic mean method are combined to generate the composite index, which measures economic uncertainty and cultural and ideational transition, reflecting major economic, social and cultural changes in relation to low fertility in a conceptualized and concrete way. Thirdly, according to the heterogeneity analysis, individuals’ perceptions of economic uncertainty at the same level and their responding fertility behaviors may be different, depending on the economy, culture, and history of the society in which they live. Fourthly, new mechanisms regarding the way that economic uncertainty shapes the fertility behaviors of individuals are uncovered; that is, in addition to delayed parenthood, as already discussed by others, fertility is also inhibited by microscopic factors such as marriage substitution and fertility preference. Furthermore, we preliminarily analyzed and tested the moderating effect of cultural and ideational transition, which is valuable for understanding the motives and consequences of convergent fertility behaviors in more depth.

2. Theoretical Analysis and Research Hypotheses

This study is mainly related to two types of research in the literature. The first type explores the impact of economic uncertainty on fertility. An individual, in terms of population economics, is believed to be a rational, economic man who determines the timing of parenthood and the number of children by balancing cost against return. Economic uncertainty, which usually refers to the economic fluctuations that affect the employment and income of individuals [1], is the external constraint that significantly influences anticipated cost and returns. The relationship between economic uncertainty and the fertility rate has gained increasing attention in the academic community since the global financial crisis of 2007–2008.

Technically, macroeconomic uncertainty, which cannot be directly measured without a unified standard, is most often represented by the proxy variables. According to studies where labor productivity or economic growth rate is used as the proxy variable, the fertility rate is procyclical and is positively correlated to the economic growth rate since the 2008 financial crisis [2,3]. Studies measuring economic uncertainty using labor market conditions conclude that the increase in unemployment and non-standard employment (e.g., part-time jobs, short-term contract jobs, and seasonal jobs) significantly inhibits fertility desire and the fertility rate [4,5]. In addition, repeated spells of joblessness, along with the number of employment spells, were used to measure economic uncertainty; it was found that a high level of repeated spells of joblessness was linked to a low fertility desire in females [6,7]. Similar findings are also found in the literature, using indicators such as CCI and economic policy uncertainty (EPU) to measure economic uncertainty. For example, according to estimations by Fokkema et al. (2008), an increase of about 0.04% in TFR was observed for each 10% increase in CCI [8]. When examining the fertility data of 32 countries from 2000 to 2013, Comolli (2017) found that a rise of 1% in EPU was associated with a decline of about 0.04% in TFR in the following year [9].

There are no direct studies on the impact of economic uncertainty on the fertility rate in China, except for indirect literature on female employment, economic pressure, or fertility cost. Female employment has been proven to be positively related to fertility, as employment and financial security were deemed to be a prerequisite for childbearing by women, who became risk-averse in the face of social risks [10]. Likewise, sound financial conditions were found to be fundamental for childbearing [11].

All the studies mentioned above show that macro-level economic instability translates into individual-level financial uncertainty and that this discourages people from starting a family and having children. Such a mechanism may be aggravated if an “added worker effect” is in place and women try to compensate for their partners’ loss of income by remaining in the labor market [12].

However, individuals’ responses to economic uncertainty are not homogeneous, and individual factors and socioeconomic background will affect individuals’ perceptions of economic uncertainty and their ability to cope with it. Grouping analysis shows that the impact of economic uncertainty on fertility varies with age, gender, income in particular, immigration status, and country cluster [13].Institutional situations, including welfare and family policies [14], the time period examined [15], and the duration and intensity of economic recession [16] will also influence the research results. Considering the varying impact of economic uncertainty on fertility in different groups and contexts, heterogeneity analysis is essential.

The second type of literature concerns the impact of cultural and value shifts regarding fertility. Since the late 1960s, Western countries have seen a multifaceted revolution, also known as the “contraceptive revolution”, “sexual revolution”, and “gender revolution” [17], which has affected fertility rates in many ways. Northern and Western Europe and many other foreign countries saw a dropping fertility rate that had already fallen below the replacement level, in conjunction with significant changes in marriage and fertility patterns, family structure, and other aspects, including delayed marriage and increasing cohabitation. This is referred to as the “second demographic transition” (SDT) [18]. Since the beginning of the 21st century, signs of SDT have been observed in an increasing number of countries. It indicates that extremely low fertility, which is not individual or transient, represents another systematic transition of the marital and childbearing behaviors of human beings [19]. For example, by using data fromthe National Population Census, the 1% National Population Sample Survey, and China Family Panel Studies, Yu and Xie (2019) evaluated important indicators of the SDT in China and suggested that China and the Western countries have commonalities in relation to certain features of the SDT, including the timing of the first marriage, cohabitation rate, and divorce rate [20].

The theory of SDT originates from the hierarchy of needs developed by Maslow (1954); that is, the needs of individuals will transform from material possessions to higher-level, non-material needs as a result of economic and social development. In line with the theory that a change in individuals’ needs induces cultural, structural, and ideational changes, Lesthaeghe and Meekers (1986) demonstrated that family formation was highly conditioned by ideational changes, with individuals and couples making the transition to parenthood to satisfy their own private needs rather thanto satisfypublic needs [21]. As Liefbroer (2005) showed, having children emerged as something to be carefully planned, which might influence the partnership and the further economic well-being of the couple. Consequently, the transition into parenthood increasingly emerged as an extended expression of individualism and self-expression of the individual [22]. By using the data from the China Family Panel Studies, Chen and Hu (2020) found that this individualistic motivation, an important driver of fertility behaviors in modern China, would induce lower fertility desire and a delayed entry into marriage, thus reducing the number of children born [23].

The two types of literature are valuable for explaining fertility behaviors in the new era. They have both advantages and disadvantages. Literature focusing on economic uncertainty paid attention to key institutional or resource restrictions that lead to low fertility, while the reason why individuals with abundant resources have fewer children was not explained. Literature studying the cultural and ideational transition gave prominence to individuals’ preferences and motivations in a specific institutional and resource context and considered the diversified fertility motivations and their determinants, but why the actual fertility rate was lower than fertility desire would suggest was not clarified. In fact, whether an individual procreates depends not only on external circumstances but also on subjective preferences. Therefore, it is of theoretical value to analyze the interaction between economic uncertainty and cultural and ideational transition, and their impact on the fertility rate. Here, we try to compensate for these disadvantages.

Firstly, uncertainty is one of the fundamental attributes of modern economic society [24]. As a result of the structural changes caused by economic globalization, a growing number of individuals have been engaged in unstable, low-paid, and low-quality jobs [25], giving rise to the anticipation of employment instability and economic insecurity [26]. Job insecurity, which is related to other areas of their personal life, makes an individual more uncertain about the envisioned future and further affects the balance between work and family life. Such uncertainties are regarded as important reasons for delayed parenthood and decreased fertility rates in Europe [27].

Secondly, the studies above imply that individuals set personal expectations and perform fertility-related actions according to objective macroeconomic information. However, economic uncertainty involves the subjective perceptions of individuals, in addition to external environmental changes [28]. Objective indicators alone cannot fully describe the economic conditions perceived by individuals. Individuals facing the same objective environmental conditions may have diverse perceptions and fertility behaviors because they have different values and modes of thinking. Therefore, values are non-negligible in the study of the impact of economic uncertainty on the fertility rate.

Thirdly, according to the SDT theory developed on the basis of sociological studies, cultural and ideational transition is the main cause of low fertility, which is marked by a decline in fertility to a level below the sub-replacement fertility level, as a result of a weakened traditional, family-oriented system. Supporters of the SDT theory believe that this is a transition of cultural orientation from materialism to post-materialism and, fundamentally, a change in the development goals and needs of human beings, namely, a shifting of individual values from material possessions to autonomy and self-expression [29].

Finally, it is theoretically predictable that the needs of individuals in modern society will transform from material possessions to higher-level non-material needs as a result of economic and social development. It gives rise to cultural and value changes and further promotes an individualism that emphasizes the freedom and ability of individuals in shaping their lives (including marriage and fertility). Being able to control their fertility independently and freely on the basis of utility evaluation, individuals will raise their expectations for their personal lifestyle and living standards and will become more sensitive to changes in material possessions, making economic uncertainty more perceptible.

Based on the above analysis, we propose two hypotheses and test them successively.

Hypothesis 1 (H1).

An increase in economic uncertainty will result in the decline of the fertility rate.

Hypothesis 2 (H2).

The rise of the level of subjective needs and the associated cultural and ideational transitions make economic uncertainty more perceptible for individuals, and reinforce the inhibiting effect of economic uncertainty on the fertility rate.

3. Mechanism Analysis

3.1. The Effect of Delayed Parenthood

The first mechanism in the literature that we incorporated into our framework concerns financial strain or other resource constraints. Financial security in relation to employment and income are prerequisites for fertility behavior in males and females, and individuals and families are inclined to consider their income and budgetary restraints when making fertility decisions. An increase in economic uncertainty, which means a decrease in an individual’s current and future income, weakens that individual’s desire and ability to make long-term commitments. In consideration of the immediate expenses of marriage and childbearing, and the huge amount of time and financial resources that would be required in the more distant future, the decision to delay marriage and parenthood seems rational. Furthermore, according to the relative income hypothesis proposed by Easterlin (1976), the impact of relative income (the earning potential of a couple/the material aspirations of a couple) on the fertility rate outweighs that of absolute income [30]. Despite the existence of higher-level subjective needs, which represent greater material aspirations, there may be a gap between the actual and desired financial resources. This gap is widened when economic uncertainty increases, enhancing individuals’ perceptions of uncertainty. Such subtle psychological changes make individuals more inclined to delay parenthood in the context of increasing uncertainty (e.g., rising unemployment) until they feel financially secure. According to the existing studies, delayed parenthood is recognized as an important cause of the declining fertility rate in some European and East Asian countries [31], and the recovery of the fertility rate in Europe is mainly attributed to the ceasing of delayed entry into parenthood [32]. Therefore, economic uncertainty is presumed to result in a decline in the fertility rate by delaying parenthood, which is more prominent when individuals have higher-level needs and stronger material aspirations.

3.2. The Effect of the Substitution of Marriage

A second mechanism in the literature that we incorporated into our framework concerns uncertainty spillover. Employment uncertainty adds to the pressure on individuals, and they often become irritable or bad-tempered at home as a result of stress at work [33]. Some studies suggest that economic uncertainty significantly deteriorates the relationship between intimate partners and affects marital stability [34]. The relationship between couples is an important factor influencing the fertility rate. Couples without a stable relationship are less likely to have children [35]. Some Chinese scholars also suggest that a stable marriage has a significant positive influence on the fertility rate, while an increase in the divorce rate will result in a decline in the fertility rate [36].

The lifestyle of and relationships between intimate partners change with cultural and ideational transitions in relation to marriage and childbearing, including the decline in the marriage rate, the rise in the divorce rate, and the increase in cohabitation before marriage. The popularity of these new types of intimate relationships contributes to their substitution for traditional marriage or the nuclear family, to some extent, gradually breaking the link between marriage, sex, and fertility. The fertility rate decreases when the uncertainty of an intimate relationship increases. Thus, economic uncertainty results in a decline in the fertility rate by affecting intimate relationships and marital stability, and cultural and ideational transitions in relation to marriage and childbearing further weaken the relationship between intimate partners, increasing the frequency of substitutes for marriage.

3.3. The Effect of Fertility Preference

The third mechanism in the literature that we incorporated into our framework concerns the balance between work and family. Individuals’ control over their jobs (e.g., working time and working mode autonomy) declines when economic uncertainty (particularly labor market fluctuation) increases, adding to work–family conflicts [37]. According to studies investigating the influences of job control on fertility desire, on the basis of the European Social Survey (ESS)’s data of 23 countries from 2004 to 2005, mothers with one child are more likely to have a second child if they gain greater control over their jobs; the fertility desire of childless women is significantly weakened by long working hours [38]. The decline in fertility desire will significantly affect the fertility rate. For example, the decrease in the fertility rate in the US from 1970 to 2014 has been found to be related toan80% increase in the number of mothers with one child and a 50% increase in the number of childless women [39].

Cultural and ideational transition in relation to marriage and childbearing weakens the role of the family system and a family-oriented lifestyle. According to the preference theory of Hakim (2003), individuals make fertility decisions according to their preference for a specific lifestyle (which is relatively constant throughout their lives): either work-oriented, home-oriented, or one that combines paid work with family time [40]. The preference for different lifestyles is recognized as the main factor affecting individuals’ preference regarding the number of children. In a comparative study in Europe, Vitali et al. (2009) found that home-oriented women had more children than work-oriented women (some of whom were childless) [41]. Thus, increased economic uncertainty, a decline in job control, and difficulty in balancing work and family life weaken individuals’ fertility desire. As a result of the cultural and ideational transition, the family-oriented lifestyle is weakened and young people prefer to have a smaller family and use voluntary birth control, resulting in a decline in the desire for children. The mechanism of effect mentioned above is shown in Figure 1.

4. Variable Selection and Model Building

4.1. Variable Selection

Considering data availability and representativeness, this paper provides an empirical analysis of data from 34 OECD countries (the United States, the United Kingdom, France, Germany, Italy, Canada, Ireland, the Netherlands, Belgium, Luxembourg, Austria, Switzerland, Norway, Iceland, Denmark, Sweden, Spain, Portugal, Greece, Turkey, Japan, Finland, Australia, New Zealand, Mexico, Czech Republic, Hungary, Poland, South Korea, Slovakia, Chile, Slovenia, Estonia, and Israel)from 2000 to 2018.The variables involved include the dependent variable, explanatory variable, moderating variable, and control variable.

• Dependent variable. Total fertility rate (TFR) represents the average number of children borne by a woman. Given that the fertility rate remains constant at different ages, TFR is the sum of fertility rates at different ages; it can be used to compare the fertility rates of different regions and different groups. Thus, TFR is used in this paper to indicate the fertility level. Data originate from the OECD database.
• Explanatory variable. According to the analyses above, economic uncertainty, an undesirable factor that is expected to be eliminated or mitigated, may be construed as a macroeconomic change affecting individuals’ perceptions of financial security. Thus, economic uncertainty must be measured according to multiple dimensions. Firstly, considering that employment has been regarded as a factor influencing fertility desire in many previous studies, this paper mainly uses the unemployment rate, youth unemployment rate (the proportion of unemployed 15–24-year-olds to the total number of unemployed people aged 15 and over), the percentage of vulnerable youth employment (the proportion of 15–29-year-olds, who are neither in employment nor in education or training, to the total number of 15–29-year-olds), and the ratio of full-time employment to part-time employment, to measure the labor market. Secondly, in the context of the knowledge economy and globalization, the level of education of young people directly determines their potential earnings and their ability to cope with market fluctuations. Thus, the college enrollment rate is measured in this paper. Lastly, economic inequality and economic uncertainty are intertwined. In the globalized labor market, it is barely possible for the young generation, particularly young people without a favorable socio-economic status, to gain greater socio-economic benefits than their parents did [42]. Thus, the Gini coefficient is included in the model (See Table 1). The original data are orthogonalized and standardized, followed by the determination of the weights of indexes, based on the entropy weight method. After that, the weighted average is calculated, and the economic uncertainty index is re-measured and comprehensively evaluated, generating a composite index (see attachment for details of the calculation process).
• Moderating variable. In light of the analyses above, cultural and ideational transition, which enhances an individual’s perceptions of economic uncertainty, may show a moderating effect on the relationship between economic uncertainty and fertility rate. Social media has greatly facilitated the social interaction between people of the same age group and has created a tremendous opportunity for them to learn lessons and seek advice from their peers, which may affect some individuals’ fertility decision-making process [42]. In this regard, cultural and ideational transition, which influences fertility decisions, depends on the media of information flow and its dissemination in modern society, to some extent. Referring to the research by Chen [29], this paper uses global information flow and dissemination as the proxy variable of cultural and ideational transition. It is measured by mobile cellular subscriptions (per 100 people), Internet subscriptions (per 100 people), and households with internet subscriptions (%). The data originate from the CEI database.

Table 1. Descriptive statistics ofthedimensions of economic uncertainty.

Dimension	Sub-Dimension	Unit	Mean	Std. Dev	Min	Max	Types
Employment	Total unemployment rate	%	7.441	3.983	2.211	27.466	+
	Youth unemployment rate	%	27.024	7.637	10.579	54.658	+
	Percentage of vulnerable youth employment	%	12.317	6.02	3.27	43.6	+
	Ratio of full-time employment to part-time employment	-	3.763	2.237	0.548	9.951	-
Education	College enrollment rate	%	65.451	17.427	9.815	142.852	-
Income distribution	Gini coefficient	%	32.455	4.524	23.7	51.5	+

Table 1. Descriptive statistics ofthedimensions of economic uncertainty.

Dimension	Sub-Dimension	Unit	Mean	Std. Dev	Min	Max	Types
Employment	Total unemployment rate	%	7.441	3.983	2.211	27.466	+
	Youth unemployment rate	%	27.024	7.637	10.579	54.658	+
	Percentage of vulnerable youth employment	%	12.317	6.02	3.27	43.6	+
	Ratio of full-time employment to part-time employment	-	3.763	2.237	0.548	9.951	-
Education	College enrollment rate	%	65.451	17.427	9.815	142.852	-
Income distribution	Gini coefficient	%	32.455	4.524	23.7	51.5	+

4.

Other Control Variables.

(1) Infant mortality rate (Imr): this refers to the number of deaths of infants under one year of age per 1000 live births (data source: WDI database). (2) Economic growth (Eg): this refers to the nominal GDP per capita (data source: WDI database). (3) Average life expectancy (Ale): this refers to life expectancy at birth, which is defined as how long, on average, a newborn can expect to live, if the death rates at birth do not change (data source: WDI database); (4) Urbanization (Ur): As a structural change in society, urbanization is commonly recognized as an important cause of the decline in the fertility rate. The urbanization rate refers to the ratio of urban population to the total population (data source: WDI database). (5) Population migration(Pm) considers the influences of international migration on the fertility rate; population migration is included as a control variable. This refers to the inflow of asylum-seekers (data source: OECD database).

Based on the above, the data of the related variables are collected and interpolation is performed to fill in the missing values, according to the mean value or tendency. Please refer to

Table 2. Descriptive statistics of main variables.

Variable		Obs	Mean	Std. Dev	Min	Max
Dependent variable	Total fertility rate	646	1.683	0.373	0.98	3.11
Explanatory variable	Economic uncertainty index	646	0.345	0.125	0.039	0.766
Moderating variables	Internet subscriptions (per 100 people)	646	21.942	11.841	0.017	46.331
	Mobile cellular subscriptions (per 100 people)	646	102.318	27.24	14.234	172.122
	Households with internet subscriptions (%)	420	71.662	19.996	7.66	99.5
Control variables	Infant mortality rate (%)	646	4.534	2.878	0.7	29
	GDP per capita (USD)	646	10.257	0.7	8.053	11.685
	Average life expectancy (years)	646	79.404	2.736	70.9	84.2
	Urbanization rate (%)	646	76.2	12	50.8	98
	Log of international migration	646	8.372	2.1	1.099	13.49

for the descriptive statistics of the variables used in this paper. TFR measures the fertility rate, with a minimum value of 0.98, a maximum value of 3.11, a mean value of 1.683, and a standard deviation of 0.373. EU measures economic uncertainty, with a mean value of 0.345 and a standard deviation of 0.125. This indicates that fertility and economic uncertainty vary greatly among countries.

4.2. Model-Building

Below is the benchmarking model in this paper:

$$TF{R}_{it}={\alpha }_0+{\alpha }_{1}E{U}_{it}+{\alpha }_{2}Contro{l}_{it}+{Ƴ}_i+{\lambda }_t+{\epsilon }_{it}$$

(1)

where i and t refer to country and year, respectively, and $TF{R}_{it}$ refers to the total fertility rate, which is a proxy variable measuring fertility level. $E{U}_{it}$, which refers to the economic uncertainty index, is the core explanatory variable that measures the level of economic uncertainty. $Contro{l}_{it}$ refers to the control variable above, while ${Ƴ}_i$ refers to the country fixed effect, and ${\lambda }_t$ refers to the year fixed effect, controlling other variables and temporal trends that may affect the fertility rate, and ${\epsilon }_{i,t}$ is the error term of the model.

To test the moderating effect of cultural and ideational transition, the interaction term $E{U}_{it}\times C{T}_{it}$ is included in the following model. $C{T}_{it}$ refers to the level of cultural and ideational transition. The definitions of other variables are the same as those in Equation (1).

$$TF{R}_{it}={\beta }_0+{\beta }_{1}E{U}_{it}+{\beta }_{2}E{U}_{it}\times C{T}_{it}+{\beta }_{3}C{T}_{it}+{\beta }_{4}Contro{l}_{it}+{Ƴ}_i+{\lambda }_t+{\epsilon }_{it}$$

(2)

According to the theoretical hypotheses of this paper, both ${\alpha }_1$ and ${\beta }_2$ are estimated to be negative.

5. Impact of Economic Uncertainty on the Fertility Rate

5.1. Analysis of Benchmarking Regression Results

The panel data from 34 OECD countries, from 2000 to 2018, was used for the estimation of Equation (1). Firstly, the impact of economic uncertainty on the fertility rate is estimated, without the inclusion of the interaction term. As shown in Column (1) of

Table 3. Benchmarking regression.

Variable	Control Variables Excluded	Control Variables Included	RE Model
	(1)	(2)	(3)
Eu	−0.3146 *** (0.0620)	−0.2915 *** (0.0650)	−0.2547 *** (0.0662)
Imr		0.0028 (0.0030)	0.0038 (0.0030)
$\ln \mathrm{GDP}$		−0.0175 (0.0328)	0.0156 (0.0312)
Ale		0.0213 *** (0.0072)	0.0189 *** (0.0072)
Ur		−1.0518 *** (0.3061)	−0.3658 (0.2477)
$\ln \mathrm{Pm}$		−0.0263 *** (0.0043)	−0.0275 *** (0.0043)
Constant	1.7971 *** (0.0289)	1.2967 ** (0.6233)	0.6385 (0.5609)
Country fixed-effect	Yes	Yes	Yes
Year fixed-effect	Yes	Yes	Yes
Observations	646	646	646
R-squared	0.1907	0.2770

Note: Robust standard errors are shown in brackets. ***, ** represent 1%, 5% significance levels, respectively.

, where the control variables are excluded, an increase in the economic uncertainty index by 1 unit is associated with a decline in the TFR of 0.3146. The regression result of Column (2), where other control variables are included, shows that the coefficient of the economic uncertainty index is significantly negative; namely, TFR drops by 0.2915 when the economic uncertainty index increases by 1 unit. Thus, H1, as given above in Section 2, is supported. A random-effect (RE) model is used to check the robustness of the regression result. As shown in Column (3), the coefficients remain statistically significant. The F-value of the Hausman test is negative, rejecting the hypothesis of the RE model. Thus, the fixed-effect model is used for the analyses below.

The regression results of control variables show that the coefficient of the infant mortality rate is not statistically significant. The reason may lie in the fact that the infant mortality rate has been low as a result of a well-developed economy, resulting in the decline and even disappearance of the motive of having children for preventive purposes. The coefficient of GDP per capita is not statistically significant, indicating that the relationship between economic growth and fertility rate may not be linear. An extended average life expectancy helps increase the fertility rate. A possible reason for this is that an extended life expectancy and improved health and nutrition intake contribute to a longer childbearing period and help to increase the fertility rate of older women. The urbanization rate is negatively related to the fertility rate, while a positive relationship is found between the youth dependency ratio and the fertility rate, which is consistent with the hypothesis. The impact of migration on the fertility rate is negative and insignificant. Similar results are also found in the studies by Goldstein et al. (2009) [32]. In other words, the impact of migration on the fertility rate may be overestimated.

5.2. Discussion ofEndogeneity

Quantitative analysis in this paper may be associated with endogeneity, which is mainly caused by reverse causality and omitted variables. Firstly, reverse causality is possible; namely, economic uncertainty (and its dimensions) and the fertility rate may be jointly determined. For example, the continuous decline of the fertility rate results in population aging and negative population growth, which may give rise to the deterioration of the economy and employment and, thus, may increase economic uncertainty. Secondly, there may be numerous omitted variables, as the fertility rate is affected by complicated multi-dimensional factors, the data for which may not be completely available, giving rise to an estimation bias. Two approaches are used to control the possible endogeneity. On the one hand, the explanatory variable and all control variables are lagged by one period, reducing the influences of reverse causality by turning the economic uncertainty index and other control variables into predetermined variables. Secondly, the ratio of the central government debt to GDP is used as an instrumental variable of economic uncertainty, mainly on the grounds that: (1) the central government debt is an important variable affecting market fluctuations [13] and uncontrollable financial risks may arise if the ratio of central government debt to GDP is too high, resulting in increased economic uncertainty; (2) the central government debt is unrelated to residuals (dependent variable), namely, the central government debt cannot affect the fertility rate by any means other than economic uncertainty. Considering that there is no evidence proving the existence of a direct relationship between the central government debt and the fertility rate, the use of this instrumental variable is believed to be reasonable.

Please refer to

Table 4. Endogeneity test.

Variable	With the Lagged Variable Included	2SLS
		First Stage	Second Stage
	(1)	(2)	(3)
Eu	−0.4824 *** (0.0657)		−1.2993 *** (0.3265)
Iv		0.1051 *** (0.0180)
Anerson LM statistic		33.46 [0.0000]
C-D Wald F statistic		34.05 {16.38}
Controls	Yes	Yes	Yes
Country fixed effect	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes
Observations	612	608	608

Note: The p-value of the statistic is indicated in square brackets ([ ]), and critical values of the Stock-Yogo test atthe10% level are shown in braces ({ }). The regression results of control variables and constant terms are omitted due to limited space. *** represent 1% significance levels.

for the regression results of Column (1), where the lagged variable is included, along with the 2SLS regression results of Column (2) and Column (3) of the instrumental variable. According to the result of the regression model, where the lagged variable is included, economic uncertainty is negatively related to the fertility rate, which is statistically significant ata1% level of significance. The result of the first-stage regression shows that the ratio of central government debt to GDP is positively related to economic uncertainty, which is statistically significant at a 1% level of significance. According to the result of the second-stage regression, after the instrumental variable is included in the regression model, the regression coefficient of economic uncertainty on the fertility rate is −1.2993, which is statistically significant at a 1% level of significance. This indicates that the actual effect of economic uncertainty on fertility may be underestimated due to endogeneity (the coefficient shown in the benchmarking regression is −0.2915). The result of the LM test by Anderson is against the null hypothesis that the instrumental variable is associated with under-identification. A higher Cragg-Donald Wald F-statistic than the critical value of the Stock–Yogo weak-identification test (at 10%) indicates a rejection of the null hypothesis that the instrument variable is weak. Thus, the instrumental variable selected in this paper is reasonable and effective.

This result is consistent with the previous research, which showed that a 1-percentage-point increase in state-level unemployment was associated with a reduction of 0.67 percentage points in the general fertility rate (GFR) in the period from 2001 to 2012 [43]. However, given that the paper focused on a single measure of unemployment, it is possible to underestimate the impact of economic uncertainty on fertility.

5.3. Heterogeneity Analysis

The economic uncertainty index implies that individuals’ perceptions and responses to economic uncertainty are homogeneous. In reality, individuals may have diversified perceptions and fertility-related responses to economic uncertainty at the same level, due to the difference in material possessions, cultural background, and economic environment. To figure out whether the impact of economic uncertainty on the fertility rate differs among regions with diversified economic levels and cultural backgrounds, and whether the estimation results in this paper are affected by the 2008 financial crisis, the tests below were performed.

Firstly, a high-income country is defined as a dummy variable, with a value of 1 if the median of its GDP per capita is higher than the sample average and is 0 otherwise. The interaction term between a high-income country (dummy variable) and the economic uncertainty index is included in the benchmarking model. According to the regression result of Column (1) in

Table 5. Heterogeneity analysis.

Variable	TFR	TFR	TFR
	(1)	(2)	(3)
Eu	−0.3997 *** (0.0719)	−0.2724 *** (0.0652)	−0.1791 ** (0.0750)
Dum_High income × Eu	0.2800 *** (0.0824)
Dum_Confucianism × Eu		−0.5726 ** (0.2414)
Dum_Financial crisis × Eu			−0.2110 *** (0.0717)
Observations	646	646	646
R-squared	0.2909	0.2838	0.2875

Note: The control variables, country fixed-effect and year fixed-effect are controlled. ***, ** represent 1%, 5% significance levels, respectively.

, the interaction coefficient is positive and is statistically significant at a 1% level of significance, indicating that economic uncertainty has a stronger inhibiting effect in low-income countries than in high-income countries. To some extent, this shows that a higher level of economic development is necessary for resisting economic risks and achieving fertility recovery. This is consistent with the findings of Luci-Greulich and Thevenon (2013) [44], which showed that governments in higher-income countries have a stronger will and ability to increase support for family-related expenditures, making households more capable of coping with financial risks and having as many children as desired. Similar studies in China showed that China’s fertility rate might rebound, as long as better economic and social development are achieved [45].

Secondly, both the Republic of Korea and Japan are strongly influenced by Confucianism, while other OECD countries are deeply influenced by Western culture. Confucianism attaches importance to collectivism and ethics in terms of social norms, while Western culture values individualism, competition, and efficiency [46]. Cultural values, which affect the way that one perceives issues, may influence the way that individuals perceive and adapt themselves to economic uncertainty. Thus, Confucianism is defined as a dummy variable (with a value of 1 if the country considered is Japan or the Republic of Korea, and 0 otherwise). The interaction term between Confucianism (dummy variable) and the economic uncertainty index is included in the benchmarking model. According to the regression result of Column (2) in Table 5, the interaction coefficient is significantly negative, indicating that economic uncertainty has a stronger inhibiting effect on those countries strongly influenced by Confucianism. A possible reason is that cultural values influence individuals’ fertility decisions by affecting their attitudes toward risks. Confucianism is more conservative than Western culture, making individuals more inclined to avoid childbearing or to have fewer children for the sake of security, when the level of economic uncertainty is too high.

This is consistent with other previous findings. For example, Du and Zhan (2019) found that people upholding risk-averse Confucianism were more prudent when making financial and investment decisions and were less likely to be involved in venture capital [47]. Miao and Pang (2019) suggested that Confucianism in East Asia attached great importance to balancing “return against cost”. In the context of the decreasing return on childbearing, Confucianism inhibits the fertility level [48].

Thirdly, the financial crisis greatly increases labor market fluctuations and adds to uncertainties in fertility decision-making. The influences are continuous. A financial crisis is defined as a dummy variable (with the value of 1 if the year considered is later than 2008, and 0 otherwise), and the interaction term between a financial crisis and the economic uncertainty index is included in the benchmarking model. According to the regression result of Column (3) in Table 5, the interaction coefficient is significantly negative, indicating that the inhibiting effect of economic uncertainty on the fertility rate became stronger after the 2008 financial crisis. This is also consistent with previous research, which showed that the negative impacts of economic conditions on fertility were more pronounced during the economic recession, compared with the pre-recession period in Europe in 2002–2014 [15].

This shows that the fertility rate depends on individuals’ expectations about the future, in addition to the current reality. The insecurity brought by economic uncertainty makes young people hesitant about marriage and childbearing. In particular, economic uncertainties such as unemployment in perceived economic distress are barriers to childbearing, while individuals may feel safer and become more capable of withstanding economic uncertainty in a perceived economic boom.

5.4. Moderating the Effect of Cultural and Ideational Transition

5.4.1. Reinforcing the Effect of Cultural and Ideational Transition on the Negative Influences of Economic Uncertainty

The reinforcing effect of cultural and ideational transition on the negative influences of economic uncertainty is investigated, according to Equation (2). According to the regression results in

Table 6. Test of the moderating effect of cultural and ideational transition.

Variable	TFR	TFR	TFR	TFR
	(1)	(2)	(3)	(4)
Eu × Proxy variable 1	−0.1110 **
	(0.0440)
Eu × Proxy variable 2		−0.0055 ***
		(0.0013)
Eu × Proxy variable 3			−0.0058 **
			(0.0023)
Eu × Ct				−0.4854 ***
				(0.1286)
Observations	646	646	420	420
R-squared	0.2854	0.3111	0.2925	0.2995

Note: Control variables, country fixed-effect, and year fixed-effect are controlled. Only the interaction coefficients are shown in the table. due to limited space. ***, ** represent 1%, 5% significance levels, respectively.

, the interaction coefficients between the economic uncertainty index and proxy variable 1 (Internet subscriptions (per 100 people)), proxy variable 2 (mobile cellular subscriptions (per 100 people)), and proxy variable 3 (households with internet subscription (%)) in terms of cultural and ideational transition are significantly negative. Three proxy variables are added together after being standardized (formula of standardization: ${\mathrm{X}}_{\mathrm{t}}{}^{\mathrm{standardized}}=\left({\mathrm{X}}_{\mathrm{t}}-{\mathrm{X}}_{\min }\right)/\left({\mathrm{X}}_{\max }-{\mathrm{X}}_{\min }\right)$), generating the cultural and ideational transition index (Ct). The interaction coefficient between the cultural and ideational transition index and the economic uncertainty index is also significantly negative, suggesting that cultural and ideational transition reinforces the inhibiting effect of economic uncertainty on the fertility rate. Thus, H2 is supported.

As shown in Column (4) in Table 6, given a mean level (0.655) of the cultural and ideational transition index, the negative impact of economic uncertainty on fertility will be increased by 0.32 (0.4854 × 0.655). This suggests that cultural and ideational transition reinforces the inhibiting effect of economic uncertainty on the fertility rate. To our knowledge, we are the first to empirically link fertility to economic uncertainty and cultural and ideational transition, providing new insights into low fertility rates.

5.4.2. Mechanism Analysis

Based on the above, it is believed that economic uncertainty may affect the fertility rate by means of the effect of delayed parenthood, the effect of the substitution of marriage with cohabitation relationships and the effect of fertility preference, and that cultural and ideational transition acts as a moderator. Childbearing age, measured by the mother’s mean age at her first birth, marriage preference, represented by the divorce rate, and fertility preference, measured by the percentage of childless women aged 40–44, are included in the test as instrumental variables. The data originate from the OECD Family Database. The higher a mother’s mean age at first birth, the stronger the effect of delayed parenthood. A higher divorce rate indicates a stronger impact on the traditional marriage system. An increase in the percentage of childless women indicates a decline in fertility desire. In light of the above, the increase in mothers’ mean ages at first birth, the divorce rate, and the percentage of childless women are believed to be adverse to the increase in the fertility rate. It is, therefore, necessary to test: (1) whether the increase in economic uncertainty results in the increase in the mothers’ mean age at first birth, the divorce rate and the percentage of childless women; and (2) whether cultural and ideational transition reinforces the positive relationship between economic uncertainty and the mother’s mean age at first birth, the divorce rate, and the percentage of childless women. Below are econometric Equations (3) and (4).

$$B{M}_{it}={\gamma }_0+{\gamma }_{1}U{C}_{it}+{\gamma }_{2}Contro{l}_{it}+{Ƴ}_i+{\lambda }_t+{\epsilon }_{it}$$

(3)

$$B{M}_{it}={\eta }_0+{\eta }_{1}U{C}_{it}+{\eta }_{2}U{C}_{it}\times C{T}_{it}+{\eta }_{3}C{T}_{it}+{\eta }_{4}Contro{l}_{it}+{Ƴ}_i+{\lambda }_t+{\epsilon }_{it}$$

(4)

where $BM$ refers to marriage and childbearing patterns and is measured by the mother’s mean age at the first birth, divorce rate, and the percentage of childless women. The definitions of other variables are the same as those specified above. The cultural and ideational transition index above is defined as the proxy value of the level of cultural and ideational transition. It is estimated that both ${\gamma }_1$ and${\eta }_2$ are positive.

According to the estimation results of Column (1) and Column (2) in

Table 7. Mechanism test.

Variable	Effect of Delayed Parenthood		Effect of the Substitution of Marriage		Effect of Fertility Preference
	Childbearing Age	Childbearing Age	Divorce Rate	Divorce Rate	Percentage of Childless Women	Percentage of Childless Women
	(1)	(2)	(3)	(4)	(5)	(6)
Eu	1.0897 ***	−0.3222	0.3022 *	0.0994	0.0236 *	0.0285 *
	(0.1954)	(0.2183)	(0.1814)	(0.1973)	(0.0122)	(0.0157)
Eu × Ct		0.7550 **		1.1582 ***		0.0149 *
		(0.3408)		(0.3081)		(0.0082)
Observations	646	420	646	420	646	420
R-squared	0.8583	0.8497	0.1368	0.2778	0.4310	0.3747

Note: Control variables, country fixed-effect and year fixed-effect are controlled, and the results of other variables are omitted due to limited space. ***, **, * represent 1%, 5% and 10% significance levels, respectively.

, in relation to the effect of delayed parenthood, the coefficient of the economic uncertainty index is significantly positive ata 1% level of significance, and the interaction coefficient is significantly positive ata 5% level of significance. For every 1-unit increase in the economic uncertainty index, the childbearing age is delayed by 1.0897 years. Given the mean level of cultural and ideational transition (0.655), economic uncertainty increases the delay of childbearing age from 1.0897 years to 1.58 years (1.0897 + 0.755 × 0.655).

Table 7 indicates that economic uncertainty gives rise to an increase in childbearing age, and the cultural and ideational transition has a significant positive moderating effect on the relationship therein, which is consistent with the theoretical hypothesis.

This is similar to the previous research, which showed that economic strain suppresses childbearing expectations as it increases uncertainty about expecting to be a parent and increases the likelihood of expecting to make the transition to parenthood later in life [49].

Compared with the previous studies, we reveal that increased economic uncertainty leads to financial strain or resource restriction and that the cultural and ideational transition encourages individuals to spend more than they make, exacerbating the delay of entry into parenthood.

As shown in the estimation results of Column (3) and Column (4) in Table 7, in relation to the effect of the substitution of marriage, the interaction coefficient and the coefficient of economic uncertainty index are significantly positive. The coefficient shows that, for every 1-unit increase in the economic uncertainty index, the divorce rate increases by 32.22 percent. Given the mean level of the cultural and ideational index (0.655), the positive influence of economic uncertainty on the divorce rate increased from 32.22% to 108.08% (32.22% + 115.82 × 0.655). This suggests that the increase in economic uncertainty causes the rise in the divorce rate, while the cultural and ideational transition has a significantly positive moderating effect on the relationship between them, which is consistent with the theoretical hypotheses. It is also consistent with the previous research, which showed that high levels of economic strain, such as short-term debt, have been linked to depressive symptoms [50] and were found to interfere with family stability and functioning [51].

Compared with previous studies, we found a spillover effect of the increase in economic uncertainty, which intensifies the conflict between intimate partners. We also prove that, as a result of the transition of values concerning marriage and childbearing, the marriage relationship between partners is more vulnerable to economic fluctuations, as they attach more importance to emotional needs in a relationship and become less tolerant of conflicts in marriage and more desirous of a high-quality marriage.

According to the estimation results of Column (5) and Column (6) in Table 7 in relation to the effect of fertility preference, the interaction coefficient and the coefficient of economic uncertainty index are significantly positive. The coefficient shows that for every 1-unit increase in the economic uncertainty index, the childlessness rate increases by 2.36 percentage points. The positive influence of economic uncertainty on the rate of childlessness increased from 2.36% to 3.34% (2.36% + 1.49% × 0.655) at the mean level of the culture and values change index (mean 0.655).

This suggests that an increase in economic uncertainty results in an increase in the percentage of childless women, while the cultural and ideational transition has a significantly positive moderating effect on the relationship between them, which is consistent with the theoretical hypotheses.

This is also consistent with the previous research, which showed that the effect of the Great Recession on the probability of being childless for women aged 37–39 in the US was positive, although moderate, with a point estimate of +1.8/1.9%,dependingon whether there was a control for state fixed-effects or not [3].In comparison with previous findings, we find that increased economic uncertainty makes individuals inclined to have fewer children or to be childless because they often find it difficult to balance work and family life. We further find that a weakened traditional marriage and family system and the shift to greater individualism add to the negative influences of economic uncertainty on childbearing.

6. Conclusions and Insights

In the context of economic globalization and the SDT, increased economic uncertainty and cultural and ideational transition are recognized as two structural dimensions affecting fertility in the modern society. In the meantime, more regions are seeing a continuously declining fertility rate. Moreover, realistic evidence shows increasing commonalities between China, a developing country, and developed countries in terms of the transformation of social structure and the change tendency of fertility rate.

On the one hand, in the context of worldwide instability and economic uncertainty (particularly the effect ofCOVID-19, which significantly influenced global economic recovery), there is no way for China’s economy to remain unaffected. Faced with slowing economic growth, a widening income gap and increased job insecurity, more and more young people are choosing to delay their entry into parenthood or to have fewer children.

On the other hand, as a result of the changes in values concerning marriage and the family, the marriage and fertility behaviors of young people in China are taking on the characteristics of SDT. Such changes in the external environment and subjective preferences have generated a huge negative impact on the reproduction rate of China’s population. According to the seventh census, China’s TFR was 1.3 in 2020, and the number of newborns hit a record low of 10.62 million in 2021. In July 2021, China’s central government announced the introduction of its “third-child” policy, which is an effective measure to cope with low fertility rates in the country. However, complaints about the high costs of bearing and raising a child and comments expressing people’s unwillingness regarding childbearing are prevalent online, indicating the existence of objective restrictions and the change in subjective attitudes.

By using the panel data of 34 OECD member countries to empirically investigate the inhibiting effect of economic uncertainty on fertility, and whether cultural and ideational transition reinforces such an effect, our study is believed to be meaningful for the recovery of fertility rates in China.

According to the findings, economic uncertainty has a significant negative effect on the fertility rate, which is more noticeable after the 2008 financial crisis and among low-income countries and countries that are strongly influenced by Confucianism. After the lagged variable and instrumental variable are used to control endogeneity, it was found that the negative impact may have been underestimated. In addition, the cultural and ideational transition reinforces the negative influence of economic uncertainty. Further tests reveal that the fertility rate is inhibited by delayed parenthood, the substitution of marriage with cohabitation, and fertility preference. Specifically, resource restrictions or financial strain as a result of increased economic uncertainty leads to a delayed entry into parenthood, indirectly damages marriage stability, and increases conflicts between work and family life, thus discouraging individuals from having more children. In addition, cultural and ideational transition is a catalyst reinforcing such impacts.

The findings suggest that the low fertility rate is the result of objective and subjective inhibiting factors and that the inhibiting effect is self-reinforcing. Economic uncertainty has a greater impact on the fertility rate of low-income countries and countries strongly influenced by Confucianism, and individuals’ expectations of the future economy affect their fertility desire. The key to increasing the fertility rate lies in the creation of a “favorable” physical and psychological environment, to enhance individuals’ expectations for and confidence in childbearing.

To this end, supportive policies are invaluable. Firstly, childbearing-related policies should be integrated with social and economic policies in such a way that job security, fair education, and income increase are achieved, in a bid to minimize economic uncertainty and enhance the perception of individuals and households of economic and social safety, stability, and predictability. Secondly, the media should play its role in properly guiding public opinion so that the correct ideas and concepts are communicated and spread to create a healthy, encouraging, and enabling environment favorable for childbearing. An efficient early-warning and evaluation system should be established to prevent the media from spreading or amplifying the economic risks and inducing or encouraging the preference for a small family.

There are limitations to this study to be noted and addressed in the future. Firstly, only macro and objective economic uncertainties have been considered, while insufficient attention has been paid to micro-level subjective economic uncertainties. Secondly, due to limited data availability, economic uncertainty indicators have been restricted to employment, education, and income distribution, while the cultural and ideational transition is merely measured by several proxy variables. Finally, the dominating factor of economic uncertainty, a composite index, is unknown and ise xpected to be further discussed.