Could Children’s Gender Predict Their Parents’ Housework Behavior?

This study’s objective was to examine whether parents show different patterns in housework behavior in relation to children’s gender. Previous studies have suggested that children’s gender might shape family processes through various paths. However, in married couples, housework behavior is mainly explained by the economic or structural differences between partners, but little is known about the potential relevance between children’s gender and parents’ housework behavior. Using nationally representative data from the Chinese General Social Survey, we examined the link between children’s gender and parents’ weekly housework hours. Two dichotomous indicators represented children’s gender: having at least one son (or not) and having at least one daughter (or not). An ordinary least squares regression model was employed for the analysis. The results indicated that parents who have at least one son, rather than those who have at least one daughter, showed significant differences in housework behavior, with mothers spend more time on housework whereas fathers spend less. The findings suggest that parents do behave differently in relation to children’s gender. By emphasizing its symbolic attribute, we argue that children’s gender constitutes part of the social gender structure related to family processes.

Could the gender of children predict their parents' housework behavior?This seems to be a new issue attracting only rare attention in the field of housework.Scholars in this field generally accept two explanatory patterns that may predict housework division.One is the gender gap pattern, which focuses on the household division of married couples; the other is the gender socialization pattern, concerning how parents shape their sons and daughters with respect to housework (Cordero-Coma & Esping-Andersen, 2018;Perry-Jenkins & Gerstel, 2020;Schulz, 2020).To examine the potential relevance between children's gender and parents' housework behavior, it is necessary to build a theoretical bridge between them.
Based on the understanding of identity derived from symbolic interaction theory, we no longer take gender role performance as an individual, one-way, passive, and static model; instead, it is an ongoing interaction process between parents and children (Stryker, 1968).Children's gender may serve as a specific symbol in a given situation to invoke a particular identity, such as a son's parent or a daughter's parent, and promote the position of this identity in the salience hierarchy.According to Stryker (1968), the higher an identity in a salience hierarchy, the higher the probability of role performance consistent with that identity.Therefore, this paper establishes a theoretical relationship between children's gender and their parents' housework behavior, which holds that parents with children of different genders may actively display various role performances in the situation of housework division.
Previous empirical studies on housework division have mainly focused on changes and its dynamic mechanism between husbands and wives, sons and daughters, and parents and their children, including sons and daughters (Hu, 2018;Perry-Jenkins & Gerstel, 2020;Sani, 2016;Schulz, 2020).However, children's gender has been found to be an important predictor of marital stability and marital satisfaction (Kaba´tek & Ribar, 2021;Raley & Bianchi, 2006;W. H. Yu & Kuo, 2018), which are two of the most important predictors of housework division between husbands and wives (Antonelli, 2018;Baxter et al., 2008;Carlson et al., 2020).These empirical studies revealed a potential link between children's gender and the parents' division of housework, which is consistent with the theoretical link established in the above discussion.
In the past 40 years, almost all relevant empirical researches were conducted in the United States and explored behaviors outside the household (such as labor market outcomes and social policy).By investigating the relationship between children's gender and parents' housework behavior in Chinese society-a patriarchal, patrimonial, and patrilineal system-this study aims to answer the following question: Do parents behave differently in relation to children's gender?Shaped by the traditions of the agricultural economy and Confucian culture over the past 2,000 years, Chinese families are still clearly characterized by the patriarchal clan system (zong fa zhi) (Hu & Scott, 2016;Thornton & Lin, 1994).As there are many differences in family structure and gender beliefs between China and the United States, this study helps to broaden the traditional boundaries of gender research and enriches understanding of the theoretical mechanism between children's gender and parents' behavior.We used data from the Chinese General Social Survey 2012 to analyze the question above and found that parents with at least one son showed significant differences in housework behavior.

Gender Division of Housework and Its Influencing Mechanisms
Housework refers to the unpaid-work of family members for the purpose of maintaining family life (Bianchi & Milkie, 2010;A. Y. Liu et al., 2015).Whether or not wives participate in the market labor, they typically perform a larger proportion of housework than do their husbands (Perry-Jenkins & Gerstel, 2020).Although some recent studies have found that the gender division of housework is becoming more egalitarian over time, the pace of change is still much slower than in other areas of social life (Bianchi et al., 2012;Carlson & Lynch, 2017;Yang, 2014).
There are four main theoretical models that explain the influence mechanism of gender division of housework.First, the relative resource model asserts that the gender division of housework depends on the personal resources that a married man and woman bring to the exchange (Brines, 1993;Carlson & Lynch, 2017;Mandel et al., 2021).Second, the structural constraint model suggests that, especially for women, the housework behavior of married couples depends on how much time they devote to other activities, such as labor force participation or bearing children (Bianchi & Milkie, 2010;Geist & Ruppanner, 2018;Stanczyk et al., 2017).Third, the gender ideology model argues that the housework behavior of married individuals is highly consistent with their gender ideology, with husbands traditionally spending less time on housework than their wives (An, 2020;Carriero & Todesco, 2018).Fourth, the life-course transformation model suggests that the change in personal identity as a couple become parents can significantly affect the gender division of intra-household labor (Dribe & Stanfors, 2009;Greenstein, 2000).

The Functional Mechanism on the Relationship Between Children's Gender and Parents' Housework Behavior
This paper's research hypotheses are mainly based on previous studies' empirical findings and the symbolic interaction theory.As a simple and powerful framework, symbolic interaction theory extensively explains family behaviors (Carter, 2014;Stryker, 1972).The theory emphasizes that symbolic meaning, identity salience, and role performance are logically congruent and continuous in a given situation.Discrete identities exist in a hierarchy of salience, and identity salience is defined as the probability of a given identity being invoked in a given situation.The higher an identity in a salience hierarchy, the higher the probability of role performance consistent with that identity, thus the higher the probability the individual will behave in accordance with the expectations of that identity (Stryker, 1968).Prior to the invoking process and role performance, an individual must define the situation in symbolic form (Stryker, 1959).This study regards children's gender as an important symbol that constitutes parents' definition of the situation.Parents with children of different genders may invoke different identities which consist with different gender role performances.

Having at Least One Son and Parents' Housework Behavior
Symbolic interaction theory emphasizes the situational constraints for the invoke of social identities.Long-term exposure to situations with traditional ideas and role models will lead to a gender role performance consistent with traditional gender role expectations (Davis & Greenstein, 2009).Empirical studies suggest that parents of sons are more likely and more consistently involved in patriarchal interactive situations.First, parents who have at least one son are more likely to experience long-term marriage, with a lower perceived risk of divorce (Kaba´tek & Ribar, 2021;Katzev et al., 1994;Takaku, 2018).Second, parents who have at least one son are more exposed to the daily interactive situations of the patriarchal system (Ding & Zhang, 2014;Gupta et al., 2003).In this context, the existence of son (sons) becomes an important symbol for parents to define the situation as patriarchal and invoke the identity of ''a son's parent.'' While the identity of a son's parent takes a higher position in the salience hierarchy, along with this is the importance of the father's involvement in the son's growth and socialization.Based on structural similarity theory, fathers and sons share obvious similarities in biological characteristics and individual expectations (e.g., gender conception and economic status) over their lifetime (Dahl & Moretti, 2008;Odenweller et al., 2018).Fathers are more ''productive'' for sons' growth, and the negative effects of the lack of fathers' role modeling are stronger on sons than on daughters (Autor et al., 2019;Lundberg, 2005).Through providing role models, fathers display and transmit the identity of breadwinner to their sons, all of which are key to sons' gender socialization in a patriarchal society.In this sense, for fathers who have at least one son, the identity of a son's parent is consistent with a breadwinner.
According to Lundberg's (2005) parenthood constraint model, the productive fathers, who are more important to the development of sons' emotional stability and interactive social skills, may not be beneficial for mothers, instead burdening them with the perceived need to invest more resources in preventing fathers from leaving their married family.Yang (2014) suggests that wives may sacrifice and compromise in the gender division of housework, increasing their tolerance for unequal labor to improve their husbands' satisfaction.By taking traditional gender roles, wives highlight the identity of breadwinner for their husbands and the identity of homemaker for themselves, through which the stability of the family is maintained (Perry-Jenkins & Gerstel, 2020;Qian & Sayer, 2016).In this sense, for mothers who have at least one son, the identity of a son's parents is consistent with homemaker.
Over the past two thousand years, patriarchy has remained the core feature of social structure in China.The family and the state are structurally consistent in the political field, and the political network is mainly composed of men (Hu & Scott, 2016;Thornton & Lin, 1994).In the cultural field, Confucianism is consistent in value with son preference, emphasizing the religious importance of men in carrying on the family name and bloodline (Sun & Lai, 2017).In family life, the gender division of labor remains traditional, men and women are regarded respectively as the breadwinner and homemaker (Qian & Li, 2020;Qian & Sayer, 2016).In this context, the identities of breadwinner and homemaker are consistent with the expectation of traditional gender roles.
Based on the above discussion, for parents who have at least one son, the identity of breadwinner is more salient for fathers and the identity of homemaker is more salient for mothers.Therefore, the role performance for parents who have at least one son will be more consistent with the breadwinner (for fathers) and homemaker (for mothers).Thus, fathers who have at least one son will be more likely to behave in accordance with the expectations of breadwinner, and mothers who have at least one son will be more probable to behave in accordance with the expectations of homemaker.In housework division, parents who have at least one son may behave consistent with the expectation of traditional gender roles, that is, fathers devote less in housework while mother devote more.Thus, we propose the following research hypothesis: H1: Compared to those without sons, fathers who have at least one son will spend less time in housework, mothers who have at least one son will spend more time in housework.

Having at Least One Daughter and Parents' Housework Behavior
In the socialist era, the Chinese government made gender equality a national policy, propagandized the policy by claiming ''Women could hold half the sky'' and enhanced women's social status by promoting female labor force participation (Ji et al., 2017;Ye & Zhao, 2018).In 1979, China implemented the one-child policy, which has had a profound and complex effect on the family process (Zhang, 2017).One of the most important consequences has been the rise in the status of daughters in Chinese families.Although a son preference still exists, the absence of brothers has meant that parental investment in economic and educational resources could focus on daughters (Fong, 2002;Mei & Jiang, 2021).As a result, the gender gap in educational attainment has reduced dramatically.Moreover, higher level of educational attainment further increases women's participation in labor market and social affairs, which can improve their social status (Jiang & Zhang, 2021;Y. Liu et al., 2020).Over the past 30 years, egalitarian gender ideology has been widely spread through social networks and mass media, and parents are more likely to be aware of the unequal treatment that daughters may experience in social life (Ji & Wu, 2018;Qian & Li, 2020).In this context, ''a daughter's parent'' becomes a salient identity in the situation of family interaction.
Glynn and Sen's ( 2015) study helps us understand the relationship between having at least one daughter and parents' housework behavior.They put forward three mechanisms: the learning mechanism, protectionism, and the lobbying mechanism.Their study found that, with the number of children controlled, judges who have at least one daughter vote in a more gender egalitarian fashion on gender issues compared to judges without daughters.Glynn and Sen's (2015) study builds a theoretical link between having at least one daughter and parents' behavior through the perception of gender inequality.
We apply this link to analyze the identity salience and role performance in the situation of family interaction through three mechanisms: protectionism, the reflection mechanism, and the sharing mechanism.
Protectionism suggests that in societies embracing egalitarianism in gender ideology, parents who have at least one daughter consciously or unconsciously exhibit certain egalitarian role models in daily life, aiming to protect their daughters from the psychological distress of traditional gender division of labor (Borrell-Porta et al., 2019;Oswald & Powdthavee, 2010).The reflective mechanism suggests that the children's interests are among the most important factors in planning the blueprint of family processes (Baunach, 2002).As parents who have at least one daughter may more strongly perceive structural and institutional discrimination in their daughters' growth, they are more likely to oppose the traditional gender roles and support the egalitarian role model (Greenlee et al., 2020;Prokos et al., 2010;Shafer & Malhotra, 2011;Sharrow et al., 2018).
Based on the protectionism and reflective mechanism, parents who have at least one daughter may invoke an identity that is more consistent with the egalitarian couple, which means the identity of the egalitarian couple is more salient.Therefore, their role performance will be more consistent with the identity of egalitarian couples.Thus, parents who have at least one daughter will be more likely to behave in accordance with the expectations of an egalitarian couple.In the situation of housework division, parents who have at least one daughter may behave in consistent with the expectation of egalitarian gender roles, that is, fathers devote more in housework while mothers devote less.
In addition, there may exist sharing mechanism for families with daughters.The sharing mechanism emphasizes the unintended consequences of gender socialization.Previous studies have found that daughters tend to take on housework usually borne by her mother, and it is easier for parents to assign this kind of housework to their daughters (Cordero-Coma & Esping-Andersen, 2018;Schulz, 2020).Thus, the mother may assign housework to her daughter in order to cultivate her ''femininity.''The unintended consequence is a reduction in the total amount of housework done by mothers who have at least one daughter.
Based on the above discussions, we propose the following research hypothesis: H2: Compared to those without daughters, fathers who have at least one daughter will spend more time in housework, mothers who have at least one daughter will spend less time in housework.

Data
In this paper, we use data from the Chinese General Social Survey (CGSS) 2012 to analyze the relationship between children's gender and parents' housework behavior.CGSS is one of the earliest national social surveys in China, and the questionnaire covers multiple layers, such as individual, family, community, and society.Our investigation includes two key variables, children's gender and parents' housework time, and explores multiple themes regarding housework time.The CGSS 2012 data cover all provincial administrative units of China, excluding Hong Kong, Macao, and Taiwan.The survey adopted a multi-stage stratified sampling design and a structured interview questionnaire; the response rate was 71.5%.
The questionnaire of CGSS2012 is divided into two parts (A and B), and respondents were randomly divided into each part, with only one spouse in each household being interviewed.Part B includes specific housework time, and the sample size is 5,915.We limited the age composition of the sample to between 20 (China's national legal age for marriage of Chinese residents) and 45 (age groups before middle age), thereby eliminating 3,656 respondents from the sample.We also limited the sample to married couples, thus eliminating 427 unmarried, cohabitating, or divorced respondents.We also eliminated 125 respondents who have no children.After finally removing respondents with missing values, the sample analyzed through our regression model comprised 1,573 respondents.

Measurement
Dependent Variables.The indicator of weekly housework hours is widely used for operationalizing housework time (Hu, 2018;Leopold et al., 2018;Skopek & Leopold, 2019).As the CGSS 2012 questionnaire directly asked respondents to report their housework time over the past week (excluding child-care and emotional concerns), we took its natural log as the dependent variable.
Independent Variables.The independent variable is children's gender.Drawing on previous theoretical and empirical studies (Lundberg, 2005;Wang, 2015), we operationalized children's gender using two indicators: (1) having at least one son (or not); and (2) having at least one daughter (or not).In the ''family'' section of the questionnaire, respondents were asked to report the number of their sons and daughters (including stepsons and stepdaughters).We encoded the numbers of sons and daughters as continuous variables, and then encoded them as two nominal variables: having at least one son is coded as 1, and having no sons is coded as 0; likewise, having at least one daughter is coded as 1, and having no daughters is coded as 0. Because prior studies have demonstrated a relationship between the number of children and parents' housework time (Bianchi & Milkie, 2010;Greenstein, 2000;Yang, 2014), we controlled for the number of children in each regression model.

Controls.
We included several controlled variables in the regression models.A respondent's labor market participation was coded as 1 if he/she currently has a job, and 0 otherwise.Respondents' and their spouses' annual income were recoded as continuous variables, and we included the natural logarithms of respondents' annual income as their absolute income in the model.Following and Greenstein (2000), we calculated respondents' economic dependence (ED) as follows: ED = (respondent's annual income 2 spouse's annual income)/(respondent's annual income + spouse's annual income).The value of ED ranges from 21 to 1. Education is considered a key human capital that couples bring to the bargaining process on housework division (Perry-Jenkins & Gerstel, 2020), so we encoded respondents' highest education level into four nominal variables: ''primary school or below,'' ''junior high school,'' ''high school,'' and ''college or above.''We used ''primary school or below'' as the reference for the role of education on housework.
A series of demographic variables were included in the regression model.Parents' gender was encoded as a dichotomous variable (''female'' = 1, ''male'' = 0), while age and age-squared were encoded as continuous variables.The number of children was also encoded as a continuous variable.Considering the potential impact on family processes of structural and developmental differences between Chinese urban and rural areas (Whyte, 2010), we encoded Hukou (which refers to a locationbased household registration and resource distribution system that categorizes Chinese citizens into urban and rural residents) as a dichotomous variable (''urban Hukou'' = 1, ''rural Hukou'' = 0).Ethnic groups were encoded as a dichotomous variable (''Han'' = 1, ''minority'' = 0).Having at least one child under 18 years old was encoded as a dichotomous variable (''yes'' = 1, ''no'' = 0).Finally, we encoded self-reported health condition (ranging from 1 = ''very bad'' to 5 = ''very good'') as a continuous variable.The variables are described in Table 1.
According to Table 1, 68% of Chinese couples in our sample have at least one son, and nearly 60% have at least one daughter.The average age of the sample is about 36.6 years.90% of the sample have Han nationality, and there is a slightly higher representation of rural Hukou (compared to urban Hukou).On the distribution of education, 25% of respondents had completed no higher than primary school, and 21% had at least a college degree.The majority of respondents are engaged in a paid job.

Analytic Strategies
A series of OLS regression models were used to investigate the relationship between children's gender and parents' housework time.First, we analyzed the housework time of Chinese married parents and compared our findings with the results from other relevant surveys.Second, we investigated the relationship between children's gender and parents' housework time with the number of children controlled.Third, by establishing the interaction term between parents' gender and children's gender, we investigated whether fathers and mothers behave differently in relation to children's gender.Finally, by rebuilding models with substituted samples and variables, we tested the robustness of our findings.

Descriptive Analysis
Descriptive statistics for the dependent variables and independent variables are presented in Table 1.The mean housework time of Chinese married parents is 13.20 hr/ week.In Table 2, we compare the values for mean housework time of husbands and wives against descriptive statistics on Chinese married couples' housework time from other surveys.
According to the CGSS 2012, among Chinese married couples aged between 20 and 45 with at least one child, wives' housework time (17.55 hr/week) is more than twice as much as that of husbands (8.08 hr/week), which is consistent with prior studies (Yang, 2014).The mean weekly difference of 9.47 hr is close to the results from the Survey of Women's Social Status in China 2010 (7.21 hr/week) and the Chinese Family Panel Survey 2010 (10.36 hr/week).As a proportion of couples' total housework time, husbands' housework time is about 30% across these surveys.

Having at Least One Son and Parents' Housework Time
Results from the regression analysis of the relationship between having at least one son and parents' housework time are reported in Table 3. Model 1 tests the influence of the demographic variables, socio-economic status variables, number of children under 18, and a series of control variables on the dependent variable.Based on Model 1, Model 2 adds the variable of having at least one son to examine its influence on the dependent variable.Model 3 further adds the interaction term of parents' gender and having at least one son into the regression model to examine whether there exists a moderating effect.
The results of Model 1 show that mothers' housework time is 127% (e 0.82 21, p \ .05)longer than that of fathers, which illustrates a clear ongoing gender gap between husbands and wives in housework.The results of Model 2 show that having at least one son for predicting parents' housework time is not significant at the 0.05 level.We then added the interaction term between  Liu et al., 2015).CFPS refers to the Chinese Family Panel Survey (Hu, 2015).
a Hours per week.
parent's gender and having at least one son in Model 3. The results show that having at least one son is significant (b = 20.18,p \ .05),and so is the interaction term (b = 0.34, p \.001), but their coefficients have opposite directions.Furthermore, the F test shows that the joined effect of having at least one son and its interaction term is significant at the .01level, which indicates that the joined effect does exist, with the influence on the dependent variable of having at least one son being moderated by parent's gender.Fathers with at least one son are found to spend 16.47% (e 20.18 -1) less time on housework compared to fathers without sons, whereas mothers with at least one son are found to spend 24.02% (e 20.18 -1 + e 0.34 -1) more time on housework compared to mothers without sons.As Figure 1 shows, parents' gender moderates between having at least one son and parents' housework time, suggesting that having at least one son widens the gap of housework time between fathers and mothers.The results are consistent with H1.

Having at Least One Daughter and Parents' Housework Time
Results from the regressions of the relationship between having at least one daughter and parents' housework time are reported in Table 4.We do not duplicate the results of the baseline model in Table 4 because completely consistent control variables were used.Model 4 adds having at least one daughter, and Model 5 further adds the interaction term of parents' gender (''female'' = 1) and having at least one daughter.
The results for Model 4 show that the influence of having at least one daughter on parents' housework time is not significant.In Model 5, the interaction term between having at least one daughter and parent's gender is not significant either, indicating that the effect on housework time of having on least one daughter does not significantly differ between fathers and mothers.Therefore, the empirical data do not support H2.

Robustness Tests
To test the robustness of our findings, we first limit the samples with to those with only one child.We found that having a son is negatively correlated with the dependent variable (b = 20.15,p = .06\ .1),while the relationship between the interaction term (between parent's gender and having a son) and the dependent variable is significantly positive (b = 0.25, p \ .05),indicating that fathers who have a son spend more time on housework and mother who have a son spend less time on housework.Second, we separated the samples into the fathers' sample and mothers' sample and reran the models.The results show that housework time significantly increases for mothers with sons (b = 0.14, SE = 0.06, p \ .05)but decreases for fathers with sons (b = 20.17,SE = 0.08, p ł 0.05).The impact of having at least one daughter on parents' housework time is not significant.
Third, we changed the core independent variables to the gender structure of children (having only sons, having only daughters, and having both son and daughter).The results show that having only sons and having both son and daughter are negatively correlated with the dependent variable.And the interaction terms (parent's gender and having only sons, and parent's gender and having both son and daughter) are positively correlated with the dependent variable (both are significant at the .01level).Furthermore, compared with those who have no sons (only daughters), fathers who have at least one son (only sons and both son and daughter) spend less time in housework, mothers who have at least one son (only sons and both son and daughter) spend more time in housework.
Fourth, we re-examined the effects in samples aged 20 to 60 (age of the elderly defined by law in China).Results show that having at least one son is negatively correlated with the dependent variable.However, the interaction term of parent's gender and having at least one son is significantly positively correlated with the dependent variable at the .05level.Again, the impact of having at least one daughter on parents' housework time is not significant.
In conclusion, resetting the samples and recoding the independent variables yield results that are highly consistent with our main findings.This indicates that our study's findings are highly robust.

Discussion
In the second decade of the 21st century, housework behavior remains a hot issue in sociological studies (Daminger, 2019;Perry-Jenkins & Gerstel, 2020).Prior research on housework has reached two general consensuses: first, there exists a gender gap in the division of housework between wives and husbands; second, there is  Bianchi, 2006;Schulz, 2020).Based on the symbolic interaction theory, this study contributes to the literature by exploring whether parents behave differently in relation to children's gender.This study finds that parents with at least one son showed significant differences in housework behavior, with fathers spend more time in housework and mothers spending less.By contrast, our empirical results do not support the hypothesis that parents may behave differently in relation to having at least one daughter.In the Chinese social context, where housework is widely considered a private domain, we thus find no evidence that parents who have daughters promote an egalitarian trend in gender role performances.It seems that Chinese parents have not widely accepted the gender equality of sons and daughters, under the strong tradition of son preference.Even if parents have a daughter, they will not choose to perform egalitarianism in housework behavior.
Our findings deepen the literature in at least two respects.Our study is the first to find that parents show different patterns of housework behavior in relation to children's gender, thus provides empirical evidence for the issue that children's gender can predict family processes.Second, this study confirms that children's gender is correlated with intra-household labor division in China, thus answering Lundberg's (2005) call for empirical research of how children's gender may shape parents' behavior in a non-American context.
Based on these findings, further discussion should focus on the symbolic meaning carried by children's gender.Feminist social theorists have proposed a new concept of gender structure, with gender analyzed as a paradigm similar to economy or class.This gender paradigm focuses on the process of gender construction embedded in daily interaction, which explains the production and reproduction of gender ideology and gendered behaviors (Risman & Davis, 2013;Scarborough & Risman, 2017).Children's gender becomes a specific symbol that embedded in parents' definition of situation, the invoke of identities, and the display of gender role performances.By emphasizing its symbolic attribute, we argue that children's gender should constitutes part of the social gender structure.Accordingly, children's gender arguably has theoretical power in explaining the development of and changes in family processes across the individual, group, and institutional levels.
Gender roles performances are displayed and transmitted in daily interactions, and the different role models of housework division that parents choose to perform lead to different understandings of gendered behaviors for children.We argue that such gender role performance by parents (perhaps inadvertently) constitutes a key component of the maintaining mechanism of gender inequality.Moreover, considering the importance of fathers' housework participation in shaping the future gendered behaviors of their children (Hu, 2018;Sani, 2016), it is necessary to promote the social expectation of Chinese fathers' time invested in housework.To eliminate the cultural context for ''females' strategy in China'' (Yang, 2014), propagandizing the equivalent role of husbands and wives in performing housework to maintain the marriage stability is also vitally important.
Children's gender and parents' housework behavior is only a small part of the research on children's gender and family processes.With the deepening of research on the intersection between family and gender, issues worth exploring include the relationship between children's gender and education investment, parents' labor market participation or outcome, and parents' mental health.Future sociological studies should focus more on explaining the theoretical mechanism between children's gender and family processes, so as to enhance understanding of the basic human behavior that occurs within and outside the basic unit of the family.

Limitation
The findings are conditioned on the following limitations that need to be addressed in future research.First, we only analyzed married couples, and so neglected cohabitating couples.Cohabitating couples have previously been found to be more egalitarian in the gender division of housework compared to married couples (Baxter, 2005;Pepin et al., 2018).Therefore, it needs to be verified whether marital status has a moderating effect on the relationship between children's gender and parents' housework behavior.Second, the CGSS data lack some critical variables that affect married couple's housework time in previous studies, including the number of adults in the household, children's age, and other family members' housework time (e.g., that of children and grandparents).In addition, the data lack the variables that could separate housework behaviors into core (include cooking, cleaning, cleaning, etc.) and un-core categories (include paying, driving, etc.), the current measurement is not enough to capture the full extent of housework.Third, as gender role performance is a long-term, dynamic, and ongoing process, panel data (tracking at least 10 years) is vital to analyze the relationship between children's gender and parents' behavior.Unfortunately, we did not have access to such panel data for families in China, and our analysis of cross-sectional data prevents firm conclusions on causal connections.Although we performed a variety of robustness tests, the study's results reveal correlation rather than causality.Therefore, it is important to provide additional measures and further investigations based on panel data in future researches.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Figure 1 .
Figure 1.The moderating effect of parents' gender on having at least one son and parents' housework time.

Table 1 .
Description of the Variables.

Table 2 .
Comparison of Different Surveys' Findings on Housework Time in Chinese Couples.
Note.SWSSC refers to the Survey of Women's Social Status in China (A.Y.

Table 3 .
Summary of OLS Regression Models of the Impact of Having at Least One Son on Parents' Housework Time (N = 1,573).

Table 4 .
Summary of OLS Regression Models of the Impact of Having at Least One 73 Daughter on Parents' Housework Time (N = 1,573).