Men’s and Women’s Gender-Role Attitudes across the Transition to Parenthood: Accounting for Child’s Gender

Men’s and Women’s Gender-Role Attitudes across the Transition to Parenthood: Accounting for... ABSTRACT Gender-role attitudes capture individuals’ degree of support for traditional divisions of paid and domestic work and have been linked to the production and reproduction of gender inequality in different social spheres. Previous research has established that life-course transitions are related to within-individual over-time change in gender-role attitudes. Most importantly, becoming a parent is associated with shifts toward more traditional viewpoints. Theories of attitude change suggest that the gender of children should influence the pattern of gender-attitude shifts that accompany parenthood, but very few studies have investigated this. We add to this literature using Australian panel data from the Household, Income, and Labour Dynamics in Australia Survey (n = 29,918 observations) stretching over 15 years and fixed-effects panel regression models. We find that men’s and women’s gender-role attitudes become more traditional when they become parents, with evidence that this process is more pronounced among men, parents of daughters and, most of all, male parents of daughters. Introduction Gender-role attitudes capture individuals’ degree of support for traditional divisions of paid and domestic work and have been linked to the production and reproduction of gender inequality in different social spheres. This is because such attitudes influence the organization of domestic work and childcare responsibilities within households, and shape employment pathways and career aspirations in gendered ways (see Davis and Greenstein [2009] for a review). It is therefore important that we understand the factors associated with variations in individuals’ support for gender-egalitarian attitudes. Research on changes in gender-role attitudes has chiefly examined long-term trends in societal levels of gender egalitarianism, differences across cohorts, and the relative contributions of cohort-replacement and intra-cohort aging in producing attitude change at the aggregate level (Brewster and Padavic 2000; Danigelis, Hardy, and Cutler 2007). A more recent and smaller pool of studies has begun to shift attention to whether and how gender-role attitudes change within individuals over their life courses. These studies have provided compelling evidence that key life-course transitions (e.g., attaining educational qualifications, relationship entry and breakdown, and parenthood) are associated with within-individual change in gender-role attitudes (Cunningham et al. 2005; Evertsson 2013; Kroska and Elman 2009; Schober and Scott 2012). The transition to parenthood has been the subject of a great deal of attention in this literature (Baxter et al. 2015), yet few studies have paid attention to whether the child’s gender moderates parenthood effects on gender-role attitudes. This possibility has nevertheless been more thoroughly tested in relation to other types of attitudes and behaviors (see Raley and Bianchi [2006] for a review). As Lee and Conley (2016, 1104) suggest, it may be that “children socialize their parents (rather than the other way around).” As will be discussed, the notion of child’s gender being a factor influencing parental gender-role attitudes is in fact embedded in theories of gender-attitude change, including exposure-based and interest-based theories (Bolzendahl and Myers 2004; Conley and Rauscher 2013; Kroska and Elman 2009; Lee and Conley 2016), perspectives based on gendered societal expectations (Bianchi and Milkie 2010; Connell and Messerschmidt 2005; Deaux 1985; Lips 2001; Lorber 1995; Steiner 2007), and backfire effect theories (Nyhan and Reifler 2010; Nyhan, Reifler, and Ubel 2013). The existing empirical evidence is nevertheless limited and mixed with, to our knowledge, only four North American studies having examined this issue (Conley and Rauscher 2013; Downey, Jackson, and Powell 1994; Shafer and Malhotra 2011; Warner 1991). Of these, only one leverages longitudinal data (Shafer and Malhotra 2011). In this paper, we examine whether and how the traditionalizing effect of parenthood on the gender-role attitudes of men and women varies with the gender of firstborn children, considering all permutations of parents’ and child’s gender. Unlike most previous cross-sectional studies, we use panel data from the Household, Income, and Labour Dynamics in Australia (HILDA) Survey stretching over 15 years and fixed-effects panel regression models. Background Existing empirical evidence A growing literature spanning across the social sciences is concerned with the associations between the gender of children and parental and family outcomes (Raley and Bianchi 2006). For example, there are reported links between children’s gender and marital stability (Morgan, Lye, and Condran 1988), parenting practices (Lytton and Romney 1991), the allocation of household labor (Pollmann-Schult hias. 2015. Sons, Daughters, and the Parental Division of Paid Work and Housework. Journal of Family Issues 38(1):100–23." 2015), educational investments in children (Freese and Powell 1999), and parental employment patterns (Lundberg and Rose 2002). Studies have also revealed associations between children’s gender and individual partisanship (Conley and Rauscher 2013), CEO’s wage policies (Dahl, Dezso, and Ross 2012), approval of military interventions (Urbatsch 2009), and support for gender-equity policies (Warner and Steel 1999) and the conservative party (Oswald and Powdthavee 2010). Additionally, judges and legislators with daughters are more likely to vote in favor of women’s rights legislation than those with sons (Glynn and Sen 2015; Washington 2008). Specific studies on the relationship between the gender of children and parental gender-role attitudes are, however, sparse. Warner (1991) used cross-sectional data from individuals in Detroit and Toronto (n = 1,808) and found that men and women with firstborn daughters were more supportive of gender-egalitarian attitudes than men and women with firstborn sons. This association was apparent for Canadian but not American men. Similarly, Downey, Jackson, and Powell (1994) used cross-sectional data from mothers in Indiana (n = 228) and found that those with firstborn sons were more likely to support traditional gender roles than those with firstborn daughters. These studies relied on non-probability, non-nationally representative, and relatively small samples, and so their findings are not generalizable to the broader population. Conley and Rauscher (2013) were the first to use representative data from the 1994 US General Social Survey (n = 1,051) and found no evidence that having a firstborn daughter relative to a firstborn son was associated with parental gender-role attitudes. However, this and the previous studies relied on cross-sectional data to document a process (attitude change) that is inherently longitudinal, which limited their ability to assess over-time change. The data they used are now also quite old. A more recent study using US panel data from the National Longitudinal Survey of Youth 1979 (n = 3,145 individuals) was undertaken by Shafer and Malhotra (2011). This found that having a firstborn daughter (relative to having a firstborn son) slightly reduces men’s support for traditional gender roles, but has no effect on women’s support for such roles. Aims and contributions Our paper adds to the existing literature in several ways. First, while previous studies have examined the relationships between parenthood, child’s gender, and gender-role attitudes, none of them invoked the four complementary perspectives on life-course gender-attitude change that we use here (interest-based, exposure-based, gendered societal expectation, and backfire effect theories). Second, we examine the effect of child’s gender on gender-role attitudes within individuals over time using nationally representative Australian panel data. This enables us to compare the same individuals before and after the transition to parenthood, generalize our findings to the Australian population, and test the generalizability of the available North American evidence in a different socio-cultural environment. Third, we examine gender-attitude trajectories over time since entry to parenthood. This allows us to provide a more granular picture of the ways in which attitudes change over and beyond the transition to parenthood, and whether or not individuals revert to their pre-parenthood gender-role attitudes. Interest-based theories of life-course attitude change Interest-based theories of gender-attitude change rest on the assumption that individuals’ interest structures (i.e., the goals they strive for) are the driving force behind their gender beliefs (Bolzendahl and Myers 2004; Kroska and Elman 2009). It follows that, if individuals’ interest structures change, their gender-role attitudes should change in response. Importantly, the notion of “interest” in this context can be extended beyond the self to encompass significant others. For instance, if a man’s wife enters the workforce, he might benefit more from gender equality (e.g., his household income would be higher in the absence of gender pay gaps) and change his attitudes toward more gender-egalitarian beliefs as a result (Cha and Thébaud 2009). Interest-based explanations for gender-attitude change can be used to make predictions about how child’s gender may affect gender-role attitudes across the transition to parenthood. Men and women who become parents of a girl should benefit more from a gender-egalitarian society in which their daughters are treated fairly and permitted to enjoy the full range of opportunities. For example, it would be in the best interest of parents of daughters to live in a society in which intimate partner violence against women is not tolerated, or in which there are no gender pay gaps. For parents of sons, however, there may be fewer perceived advantages associated with societal gender egalitarianism. The perpetuation of the current status quo, in which girls and women remain disadvantaged in a range of life domains, may in fact result in a comparative advantage for their male sons. Hence, the interest structures of parents of girls should become more closely aligned with the goal of gender equality than the interest structures of parents of boys and, as a result, their gender-role attitudes should become comparatively more egalitarian. It is also possible that parental gender moderates how interest structures operate in this context. On the one hand, out of their own personal interest, women’s gender-role attitudes prior to the transition to parenthood may already reflect that women benefit more than men from a gender-egalitarian society. Hence, the arrival of a firstborn daughter may be associated with a stronger shift toward egalitarianism in gender-role attitudes among men, for whom their presence would constitute a more significant addition to their interest structures (Davis and Greenstein 2009). On the other hand, psychological studies on parent-child attachment have reported stronger bonds between same-gender parent-child dyads (or same-gender filial preferences), whereby fathers have a predilection for their sons and women for their daughters (McHale, Crouter, and Whiteman 2003; Raley and Bianchi 2006; Rossi and Rossi 1990). Thus, firstborn daughters may have a greater potential to shift mothers’ than fathers’ interest structures. Hence, becoming a parent of a firstborn daughter may be associated with a stronger shift toward more gender-egalitarian attitudes among women. Exposure-based theories of life-course attitude change Exposure-based theories of gender-attitude change argue that gender beliefs are rooted in ill-founded, stereotypical assumptions about women’s (and men’s) capabilities and the nature of femininity (Bolzendahl and Myers 2004). Gender-role attitudes can thus change if individuals become exposed to circumstances, situations, and experiences that challenge such assumptions (Davis and Greenstein 2009). For example, men may change their perceptions about women being ill suited to undertake certain jobs if they meet successful women at the workplace (Bolzendahl and Myers 2004). Based on exposure-based theories, it can be argued that individuals who become parents of girls will likely face situations that expose them to unfair, discriminatory behavior toward females. For example, parents may witness their daughters being tracked into gender-typical play groups and educational pathways (e.g., home economics lessons), denied access to clubs and societies (e.g., sporting clubs), or being the subject of the “male gaze” and inappropriate stereotypical or sexual comments (Kane 2012). These experiences and circumstances should make parents of girls more aware of structural inequalities unfavorable to women that emerge due to traditional gender ideologies, and should in turn lead them to question and reassess their own gender-role attitudes toward more egalitarian standpoints (see, e.g., Weitzman 2015). Parents of sons, on the other hand, should be exposed to few (if any) structural factors disadvantaging their male children, given a societal status quo that clearly favors men and masculinity. Instead, parents of sons may be more likely to encounter situations in which (hegemonic) masculinity (Connell and Messerschmidt 2005) is exalted and reinforced, such as participation in and attendance at sporting activities or consumption of male-typed entertainment and media products. Hence, exposure-based theories also lead to the prediction that men and women with firstborn daughters should experience less traditionalization in their gender-role attitudes after the transition to parenthood than men and women with firstborn sons. Using exposure-based perspectives, it is also possible to anticipate parental gender to have a moderating role. On the one hand, women may be more knowledgeable about gender-based discrimination than men due to their own experiences prior to becoming mothers, and so the addition of their daughters to their lives may entail less exposure to new situations than for men (Lee and Conley 2016; Shafer and Malhotra 2011). In these circumstances, one would expect a stronger shift toward gender egalitarianism among men. On the other hand, parents spend more time with children of their same gender (McHale, Crouter, and Whiteman 2003; Raley and Bianchi 2006; Rossi and Rossi 1990). Hence, women with firstborn girls may be more likely than men with firstborn girls to witness acts of discrimination against their daughters that prompt them to reconsider their gender-role attitudes (Bolzendahl and Myers 2004). Becoming a parent of a firstborn daughter may therefore be associated with a comparatively stronger shift toward egalitarianism in gender-role attitudes among women. Altogether, both interest- and exposure-based theories lead us to predict that: Hypothesis 1: Men and women with firstborn daughters will experience less traditionalization in their gender-role attitudes after the transition to parenthood than men and women with firstborn sons. In addition, both theories suggest that parents with firstborn girls should become more aware of gender-based discrimination and develop stronger interests in gender equality as their daughters grow older and face a greater variety of social contexts and circumstances (Shafer and Malhotra 2011). In our analyses, we will test this premise empirically by estimating models that account for time since the birth of the first child (details below). Gendered societal expectations During the early years, new parents may mainly think of their sons and daughters as dependents and receivers of care, which has implications for how the child’s interest is defined. In this context, parents may shift their worldviews to place more value on a system in which an adult is ever-present and fully committed to providing care and emotional support to the child (Rose and Elicker 2010). From the parental side, this process has a well-established and strong gender component: normative, institutionally enforced gender scripts dictate that it should be the child’s mother who adopts the main caregiver role (Bianchi and Milkie 2010; Steiner 2007). As we explain below, this process may also be gendered on the child’s side. Parents draw upon normative expectations when adapting to the requirements of and changes brought about by parenthood. This is particularly applicable to first-time parents, as they lack personal experiences on which to draw. Contemporary societal discourses around parenthood are often deeply gendered, as exemplified by well-established normative beliefs that mothers are better equipped and more capable than fathers to care for young children. Additionally, social pressures operate to make parents conform to these normative expectations, with new parents being “bombarded” with advice about parenthood and parenting by family members, friends, acquaintances, health professionals, and even strangers, as well as media channels (Moseley, Freed, and Goold 2010). In addition, there are also deeply ingrained societal discourses about the nature of boyhood and girlhood. Consistent with the social construction of femininity and masculinity in Western societies, a common theme in these discourses is the portrayal of girls as weak, fragile, passive, and dependent, and of boys as strong, able, active, and independent (Connell and Messerschmidt 2005; Deaux 1985; Lips 2001; Lorber 1995). It follows that the arrival of a firstborn daughter may elicit stronger protective and intensive parenting feelings among first-time parents than the arrival of a firstborn son. These feelings may involve more acute perceptions that the child is a delicate entity that requires parental attention, care, and protection, and that it is not appropriate for young children to attend out-of-home childcare on a full-time basis. This resonates with psychological research evidence that parents treat their daughters different than their sons in ways that reproduce gender stereotypes. For example, parents of daughters are more likely than parents of sons to discourage aggression or to display warmth toward the child, with gender-biased parental treatment being more prevalent among fathers than mothers (Raley and Bianchi 2006). The argument is also consistent with findings from criminology research that parents of girls experience fear of crime more often than parents of boys (Vozmediano et al. 2017). Due to the gendered nature of household divisions of labor and of government support to parents in countries such as Australia, for most parents the only realistic or conceivable option to engage in intensive parenting is for the mother to assume the associated responsibilities (Buchler, Perales, and Baxter 2017). In these circumstances, changes toward stronger beliefs in protective or intensive parenting across the transition to parenthood actually equate to changes toward stronger beliefs in traditional gender divisions (Rose and Elicker 2010). Therefore, one could expect shifts toward more traditional parental gender-role attitudes with the arrival of a firstborn daughter, compared to a firstborn son. A corollary is that the predicted shift toward more traditional gender-role attitudes with the birth of a firstborn girl may be more pronounced among men, for whom “traditionalizing” is less costly—it involves changing their views but not their behaviors. In fact, it would be in men’s personal benefit to traditionalize and adopt viewpoints that depict a status quo in which they are not responsible for activities that are typically not highly valued—such as routine childcare tasks. In contrast, for most women, traditionalizing involves not only reassessing their attitudes, but also reconsidering how these fit with their new roles and behaviors as mothers, which may lead to cognitive dissonance—a misalignment between one’s attitudes and behaviors that produces psychological strain (Baxter et al. 2015; Buchler, Perales, and Baxter 2017). Hence, it is less costly for men than women to “indulge” social expectations and adopt views of girls requiring more intensive parenting. This suggests that shifts toward more traditional gender-role attitudes across the transition to parenthood should be stronger among men than women with firstborn daughters. Based on this, we develop a second set of hypotheses: Hypothesis 2: Men and women with firstborn daughters will experience more traditionalization in their gender-role attitudes after the transition to parenthood than men and women with firstborn sons. Hypothesis 3: Men with firstborn daughters will experience comparatively more traditionalization in their gender-role attitudes after the transition to parenthood than women with firstborn daughters. Backfire effect theories Hypotheses 2 and 3 are consistent with predictions from backfire effect theories. These argue that, when people’s personal attitudes are based on unfounded convictions, encountering new situations or information challenging their views may actually result in people holding more strongly to their beliefs (Nyhan and Reifler 2010; Nyhan, Reifler, and Ubel 2013). Evidence of backfire effects has been found for attitudes toward the Iraq War, tax cuts, stem cell research, health care expenditure, or global warming (see e.g., Nyhan and Reifler 2010; Nyhan, Reifler, and Ubel 2013). Backfire effect theories suggest that, if parents of firstborn daughters become disproportionately exposed to situations that challenge their gender beliefs, such exposure would lead these individuals to hold on to traditional gender beliefs. Additionally, these perspectives suggest that those people who hold the most conservative gender-role attitudes prior to parenthood would more strongly hold on to them post-parenthood. This suggests that, with parenthood, the attitudes of men with firstborn daughters should become comparatively more traditional than those of women with firstborn daughters. Table A1 provides a summary of the expectations of each of the theories discussed. While our four theoretical perspectives are presented separately, it must be noted that in practice there are significant overlaps between them. For example, consistent with exposure-based theories, interest-based theories assume awareness and exposure to discriminatory practices against girls, with interest being structured around avoidance of such practices and their potential consequences on female daughters. Likewise, backfire effect theories also assume the existence of exposure to such situations, with the difference being that under this framework they are expected to elicit different psychological reactions in parents. Also, the different perspectives bear diverging temporal implications: interest-based, exposure-based, and backfire effect theories assume a progressive shift toward more traditional parental gender ideologies as children age and parents encounter new situations that challenge their gender attitudes, while gendered societal expectations theory assumes more immediate and perhaps more fleeting effects of childbirth on parental attitudes. The Australian context The available research on child’s gender and parental gender-role attitudes has, to our knowledge, exclusively relied on data from the United States and one Canadian city. An innovative aspect of our paper is our focus on a different country: Australia. Expanding the evidence base beyond the United States is important to ascertain the generalizability of the available findings, and to begin to tease out how institutional contexts may matter. Doing so, however, poses questions about whether or not the theoretical mechanisms outlined before operate similarly or differently across countries, and specifically between the United States (where most research on this topic has been conducted) and Australia (where our data come from). Interest- and exposure-based theories should operate more strongly in societal contexts in which being female is associated with a deeper degree of disadvantage. In this regard, the Australian and US contexts are very similar. For instance, Australia and the United States rank 45th and 46th (out of 144 countries) in the 2016 Global Gender Gap Index (World Economic Forum 2016). Therefore, according to interest- and exposure-based theories, we should expect similar relationships between child’s gender and parental gender-role attitudes in Australia and the United States. The gendered societal expectations argument should operate more strongly in societies in which intensive parenting ideologies prevail, and in which gender stereotypes remain deeply ingrained. While, to our knowledge, there is no cross-national comparative evidence of (dis)similarities in gender stereotyping in Australia and the United States, there is evidence to suggest that intensive parenting ideologies are more widespread in Australia than the United States. For example, both working and non-working adults in Australia spend a greater share of their time on care activities than their US counterparts (OECD 2016a). In addition, Australia features a more generous, mandated paid parental leave scheme: women are entitled to 18 weeks of paid maternity leave paid at a 42 percent payment rate (equivalent to 7.6 weeks of full-time pay), while—since 2013—men are entitled to two weeks of paid paternity leave at the same rate (equivalent to 0.8 weeks of full-time pay) (Buchler, Perales, and Baxter 2017). This imbalance in leave entitlements between Australian men and women, together with higher rates in maternal time out of employment and part-time work rates (Baxter et al. 2015), exemplifies deep-rooted normative expectations in Australia of mothers as caregivers. In the United States, however, neither mothers nor fathers are entitled to paid parental leave (OECD 2016b). The US Family and Medical Leave Act provides eligible employees with 12 weeks of annual leave for family/medical reasons, but this is not mandated to be paid leave. Taken together, these aspects suggest that in Australia, more than in the United States, normative practices and features of the institutional environment place a larger premium on parental childcare, with a strong expectation that the bulk of it will be undertaken by mothers rather than fathers. It follows that, if the gendered societal expectations hypothesis is correct, we should expect parental gender-role attitudes (and especially paternal gender-role attitudes) to move toward comparatively more traditional standpoints in Australia than in the United States. Methodology Dataset and sample selection We examine whether the gender of firstborn children affects the rate of change in parental gender-role attitudes across the transition to parenthood. To test this, we use data from the Household, Income, and Labour Dynamics in Australia (HILDA) Survey, a household panel survey that tracks individuals living in the same households in Australia for the 2001 to 2015 period and is largely representative of the Australian population in 2001 (Summerfield et al. 2016). This is one of the largest and most respected panel surveys at a global scale, and is part of the Cross-National Equivalent File of household panels. Most of the HILDA Survey data are collected via computer-assisted face-to-face interviews taking place at the respondents’ households, with some information (including that on gender attitudes) being completed privately by respondents via a self-complete instrument. This is most often handed over to the interviewer prior to her/his departure, although some respondents opt to mail it afterward. Attrition rates in the HILDA Survey are remarkably low for international standards. For instance, only around 5 percent of previous-wave respondents left the survey between its two most recent sweeps (14 and 15). Unlike in most cohort studies, individuals can enter household panels after the initial sweep. In the HILDA Survey, new individuals can join the panel if they live in participating households and become 15 years of age, or if they begin sharing a residence with a sample member. If a new panel member marries or has a child with an existing sample member, that panel member would also be followed over time. Hence, by design, not all individuals in the HILDA Survey are observed the same number of times for reasons other than panel attrition. We use data from the five HILDA Survey waves containing information on gender-role attitudes: wave one (2001), wave five (2005), wave eight (2008), wave 11 (2011), and wave 15 (2015). We consider only person-year observations in these waves in which respondents were aged between 18 and 50 years, inclusive (to focus on prime childbearing and childrearing ages) (n = 43,388). We then exclude 6,379 person-year observations from respondents who had missing data on model variables. Of these, a vast majority (over 99 percent) were dropped due to missing information on gender-role attitudes, as the self-complete questionnaire in which this is collected incurs higher non-response. We exclude also 7,096 person-year observations from respondents who were only observed once after applying the previous exclusion criteria, because we fit fixed-effect panel regression models that require at least two observations per individual (see details below). We refrain from imputing missing information at the item level because most information is missing on the outcome variable, and because of the absence of widely accepted methods to do so in a panel environment. Our final analytical sample comprises 29,918 person-year observations from 9,583 individuals. Within this sample, 3,980 individuals were observed twice over the life of the panel, 2,045 individuals were observed three times, 1,967 individuals were observed four times, and 1,590 individuals were observed all five times. This, however, does not constitute an issue for our estimation, as our fixed-effects models can handle unbalanced data. Of note, we do not exclude individuals who were parents when the HILDA Survey commenced (2001) or who were parents and entered the study later on (e.g., by joining a participating household). Because we fit fixed-effects models that only use within-individual over-time changes in the explanatory variables to estimate their model coefficients (details below), these individuals do not contribute to estimation of the model parameters on the transition to parenthood (they are always observed in the category of “parents”). Similarly, childless individuals who enter the study and are never observed to have a child also do not contribute to the estimation of the parenthood effects in our fixed-effects models (they are always observed in the category of “non-parents”). However, retaining these two types of individuals in the sample is preferable to excluding them, as they contribute to estimation of other model variables (e.g., age or education) on which they do experience change over time. This approach is customary in studies using fixed-effects models; see for example Baxter et al. (2015). Outcome variable: Gender-role attitudes Following Baxter et al. (2015), we operationalize gender-role attitudes using a time-varying variable summarizing respondents’ degree of agreement with seven separate statements (see table A2). Response options range from (1) “strongly agree” to (7) “strongly disagree,” and where necessary they were recoded so that high values always represent more traditional views about gender roles. We construct an additive scale by summing the scores of the seven items. For ease of interpretation, we rescale the resulting variable so that it ranges from 0 (most egalitarian gender attitudes) to 100 (most traditional gender attitudes). While the Cronbach’s alpha score for this scale is only moderate (0.6), factor analyses reveal that only one factor had an Eigenvalue over 1 (1.4), all items loaded positively on this factor, and the second-highest Eigenvalue among factors was very low (0.4). We take this as evidence of unidimensionality. Key explanatory variables: The transition to parenthood In our HILDA Survey sample, 1,430 men and 1,615 women become parents for the first time. Of these, 691 men and 790 women have a firstborn son, and 739 men and 825 women have a firstborn daughter. As is common practice in studies of the effects of children’s gender on parental and family outcomes, we focus exclusively on first births. This minimizes selection bias due to “endogenous stopping rules” arising from differential fertility choices and preferences for children of either gender (Dahl and Moretti 2008). In addition, a large body of research has identified first births as a distinct and critical life-course transition (Baxter et al. 2015). We do not consider cases in which first births were twins (n = 136 pairs). To reassess the relationship between the transition to parenthood and gender-role attitudes using our unique dataset, we first derive a “base” parenthood indicator. This is a time-varying dummy variable taking the value one if the respondent has been observed to have a firstborn child during the life of the panel, and the value zero otherwise. To distinguish by parental gender, we subsequently split this “base” parenthood variable into two variables: “father” is a dummy variable taking the value one if the respondent (i) is male, and (ii) has been observed to have his firstborn child during the life of the panel, and the value zero otherwise; and “mother” is a dummy variable taking the value one if the respondent (i) is female, and (ii) has been observed to have her firstborn child during the life of the panel, and the value zero otherwise. To distinguish by child’s gender, we then split the “base” parenthood variable into two new variables: “daughter” is a dummy variable taking the value one if the respondent (i) has been observed to have a firstborn child during the life of the panel, and (ii) has a firstborn girl, and the value zero otherwise; and “son” is a dummy variable taking the value one if the respondent (i) has been observed to have a firstborn child during the life of the panel, and (ii) has a firstborn boy, and the value zero otherwise. The independent variables in our main model combine information from the previous parenthood variables on the birth of the first child, the gender of the first child, and the gender of the parents into four parent-child gender variables: “male parent of daughter” is a dummy variable taking the value one if the respondent (i) is male, (ii) has been observed to have his firstborn child during the life of the panel, and (iii) has a firstborn girl; and “female parent of daughter” is a dummy variable taking the value one if the respondent (i) is female, (ii) has been observed to have her firstborn child during the life of the panel, and (iii) has a firstborn girl. In addition, “male parent of son” is a dummy variable taking the value one if the respondent (i) is male, (ii) has been observed to have his firstborn child during the life of the panel, and (iii) has a firstborn boy; and “female parent of son,” is a dummy variable taking the value one if the respondent (i) is female, (ii) has been observed to have her firstborn child during the life of the panel, and (iii) has a firstborn boy. Note that an individual who does not satisfy the criteria for a given parenthood variable would be assigned a score of zero in all survey waves. Since we use fixed-effects models estimated using change over time (see details below), such individuals would not contribute to the estimation of the regression coefficients on that variable. Control variables In our multivariate panel regression models described below, we control for a range of variables that may confound the associations between parenthood, child’s gender, and gender-role attitudes. These variables are time-varying, measured at the individual level, and are based on those used in previous studies in this field. We control for a set of dummy variables capturing marital status [married/in a de facto relationship/divorced, separated, or widowed/single (never married)], as marital status has been shown to be associated with gender-role attitudes—with married individuals displaying comparatively traditional views (Brewster and Padavic 2000). Education is associated with individuals holding more egalitarian gender attitudes (Cunningham et al. 2005), so we control for a set of dummy variables capturing respondents’ highest education qualification (university degree or higher/vocational education certificate or equivalent/secondary education/lower than secondary education). We include a control variable for religiosity, as this has recurrently been linked to more traditional gender attitudes (see, e.g., Mikołajczak and Pietrzak 2014). Our measure of religiosity is based on individuals’ responses to a question asking “On a scale from 0 to 10, how important is religion in your life?.” Finally, we control for individuals’ age at last birthday (expressed in years) and its square, as substantial literature documents trends in gender-role attitudes over the life course (see e.g., Scott, Alwin, and Braun 1996). We explicitly refrain from adjusting for factors that may themselves be a consequence of shifts in gender-role attitudes across the transition to parenthood, such as changes in employment status and subsequent births. This is because we are interested in documenting the “base” parenthood effect, rather than identifying the intervening mechanisms that produce it. In addition, while second children may strengthen traditional gender attitudes, they are naturally posterior to parenthood on the causal chain, can only be experienced by parents, and may be correlated with the gender of the first child. Hence, adding variables capturing employment status and higher-order births to the models can be seen as an instance of “over controlling” and would result in artificially downward-biased estimates of the parenthood effects. Nevertheless, results including variables capturing these aspects are available from the authors upon request. Since we fit fixed-effects regression models, we need not (and cannot) adjust for time-constant factors such as socio-economic, ethnic, or migrant background. Table 1 shows means and standard deviations for model variables. Table 1. Means and Standard Deviations for Model Variables, HILDA Survey (Australia) Women Men Mean/% SD Mean/% SD Gender-role attitudes (0–100) 39.20 15.33 43.74 14.29 Parenthood  Daughter 9% 9%  Son 8% 8% Control variables Marital status  Married 48% 46%  De facto relationship 19% 19%  Divorced, separated, widowed 8% 5%  Single 25% 30% Highest education qualification  University degree or higher 30% 24%  Vocational education certificate or equivalent 26% 36%  Secondary education 21% 20%  Lower than secondary education 23% 20% Religiosity (0–10) 3.20 3.48 2.43 3.21 Religion-missing flag 14% 16% Age in years 34.33 9.51 34.34 9.56 Years since first birth* 4.27 3.64 4.28 3.64 Women Men Mean/% SD Mean/% SD Gender-role attitudes (0–100) 39.20 15.33 43.74 14.29 Parenthood  Daughter 9% 9%  Son 8% 8% Control variables Marital status  Married 48% 46%  De facto relationship 19% 19%  Divorced, separated, widowed 8% 5%  Single 25% 30% Highest education qualification  University degree or higher 30% 24%  Vocational education certificate or equivalent 26% 36%  Secondary education 21% 20%  Lower than secondary education 23% 20% Religiosity (0–10) 3.20 3.48 2.43 3.21 Religion-missing flag 14% 16% Age in years 34.33 9.51 34.34 9.56 Years since first birth* 4.27 3.64 4.28 3.64 Note: HILDA Survey data. n (observations) = 29,918; n (individuals) = 9,583. * Only for individuals who experienced a first birth over the life of the panel. Table 1. Means and Standard Deviations for Model Variables, HILDA Survey (Australia) Women Men Mean/% SD Mean/% SD Gender-role attitudes (0–100) 39.20 15.33 43.74 14.29 Parenthood  Daughter 9% 9%  Son 8% 8% Control variables Marital status  Married 48% 46%  De facto relationship 19% 19%  Divorced, separated, widowed 8% 5%  Single 25% 30% Highest education qualification  University degree or higher 30% 24%  Vocational education certificate or equivalent 26% 36%  Secondary education 21% 20%  Lower than secondary education 23% 20% Religiosity (0–10) 3.20 3.48 2.43 3.21 Religion-missing flag 14% 16% Age in years 34.33 9.51 34.34 9.56 Years since first birth* 4.27 3.64 4.28 3.64 Women Men Mean/% SD Mean/% SD Gender-role attitudes (0–100) 39.20 15.33 43.74 14.29 Parenthood  Daughter 9% 9%  Son 8% 8% Control variables Marital status  Married 48% 46%  De facto relationship 19% 19%  Divorced, separated, widowed 8% 5%  Single 25% 30% Highest education qualification  University degree or higher 30% 24%  Vocational education certificate or equivalent 26% 36%  Secondary education 21% 20%  Lower than secondary education 23% 20% Religiosity (0–10) 3.20 3.48 2.43 3.21 Religion-missing flag 14% 16% Age in years 34.33 9.51 34.34 9.56 Years since first birth* 4.27 3.64 4.28 3.64 Note: HILDA Survey data. n (observations) = 29,918; n (individuals) = 9,583. * Only for individuals who experienced a first birth over the life of the panel. Modelling strategy: Fixed-effects panel regression models We model the relationships of interest using fixed-effects models. These are estimated by regressing deviations from individuals’ person-means in the dependent variable on deviations from their person-means in the independent variables (Allison 2009). The main model examines the effect of parenthood on gender-role attitudes for each of our “parent-child gender” variables: (GRAit−GRA̅i)=(PCGit−PCG̅i)β+(Xit−X̅i)γ+(εit−ε̅i) (1) where i and t denote individual and time, GRA stands for gender-role attitudes, PCG is a set of variables representing the four different parent-child gender combinations, X is a vector of time-varying control variables, β and γ are model coefficients, and ε is a random error term. It is important to note that, while fixed-effects models cannot incorporate time-constant variables (such as gender), they can incorporate interactions between time-constant variables and time-varying variables (see, e.g., Allison 2009, 37–38). Our variables capturing the interactions between parenthood (time-varying), parent’s gender (time-constant), and child’s gender (time-constant) fall under this banner. An extension to the previous model allows us to examine gender-role attitude trajectories for each of the different parent-child gender variables as firstborn children age. This is useful to examine whether or not, as hinted by interest- and (especially) exposure-based theories, parents become more aware of structural discrimination against girls as their daughters grow older, and consequently change their gender attitudes at a faster rate as time since birth elapses. To do so, we interact each of the parent-child gender variables with a variable capturing the number of years since the first birth (YSB): (GRAit−GRA̅i)=(PCGit−PCG̅i)β+(PCGit−PCG̅i)∗(YSBit−YSB̅i)θ+(Xit−X̅i)γ+(εit−ε̅i) (2) One could think of the θ coefficients in this model as a version of the growth parameter(s) estimated using “growth models,” as they measure trajectories in gender-role attitudes since parenthood. However, unlike those from traditional growth models, our coefficients are estimated in a fixed-effects rather than a random-effects framework, and so they are more robust to omitted-variable bias due to time-constant unobserved heterogeneity (Allison 2009). In sensitivity analyses, we tested for non-linear effects in the years since first birth variable, but found no evidence of these. We use Wald tests to examine whether parenthood impacts gender-role attitudes differently for individuals falling into the different categories of the variables capturing parental gender (“father,” “mother”), child’s gender (“daughter,” “son”), and parent-child gender dyads (“male parent of son,” “male parent of daughter,” “female parent of son,” and “female parent of daughter”). Results Table 2 presents the results of our fixed-effects models of gender-role attitudes. Positive coefficients on the independent variables indicate that the variables are associated with more traditional gender-role attitudes, whereas negative coefficients on the explanatory variables indicate that the variables are associated with more egalitarian gender-role attitudes. Across all models, the coefficients on the parenthood variables give the average difference in support for traditional gender-role attitudes across all observations before and all observations after the experience of parenthood for those individuals who become parents over the life of the panel, all else being equal. Table 2. Predictors of Gender-Role Attitudes, Fixed-Effects Models, HILDA Survey (Australia) Model 1 2 3 4 5 Effect of the transition to parenthood for…  All parents 1.91***  Male parent 2.50***  Female parent 1.39***  Parent of daughter 2.41***  Parent of son 1.39***  Female parent of daughter 1.70** 1.93***  Male parent of daughter 3.20*** 3.60***  Female parent of son 1.06* 1.68**  Male parent of son 1.75** 1.74** Effect of years since parenthood for …  Female parent of daughter −0.10  Male parent of daughter −0.15  Female parent of son −0.21**  Male parent of son −0.03 Marital status (ref. married)  De facto relationship −0.53 −0.52 −0.53 −0.53 −0.50  Divorced, separated, widowed −0.59 −0.58 −0.59 −0.57 −0.54  Single −0.45 −0.42 −0.44 −0.42 −0.28 Highest educ. qualification (ref. secondary educ.)  University degree or higher −1.93*** −1.93*** −1.93*** −1.93*** −2.04***  Vocational educ. certificate or equivalent −0.75 −0.76 −0.76 −0.77 −0.82  Lower than secondary education −0.66 −0.67 −0.66 −0.68 −0.67 Religiosity 0.18*** 0.18*** 0.18*** 0.18*** 0.19*** Age −0.54*** −0.53*** −0.54*** −0.53*** −0.49*** Age2 0.00 0.00 0.00 0.00 0.00 Constant 57.82*** 57.67*** 57.82*** 57.65*** 56.55*** R2 (within) 0.04 0.04 0.04 0.04 0.04 N (observations) 29,918 29,918 29,918 29,918 29,918 N (individuals) 9,583 9,583 9,583 9,583 9,583 Wald tests (p values)  βMother = βFather 0.03  βDaughter = βSon 0.05 βMother of daughter = βFather of daughter 0.04 0.04  βMother of son = βFather of son 0.35 0.94  βMother of daughter = βMother of son 0.36 0.76  βFather of daughter = βFather of son 0.05 0.03  βMother of daughter, trajectory = βFather of daughter, trajectory 0.69  βMother of son, trajectory = βFather of son, trajectory 0.13  βMother of daughter, trajectory = βMother of son, trajectory 0.33  βMother of daughter, trajectory = βFather of son, trajectory 0.35 Model 1 2 3 4 5 Effect of the transition to parenthood for…  All parents 1.91***  Male parent 2.50***  Female parent 1.39***  Parent of daughter 2.41***  Parent of son 1.39***  Female parent of daughter 1.70** 1.93***  Male parent of daughter 3.20*** 3.60***  Female parent of son 1.06* 1.68**  Male parent of son 1.75** 1.74** Effect of years since parenthood for …  Female parent of daughter −0.10  Male parent of daughter −0.15  Female parent of son −0.21**  Male parent of son −0.03 Marital status (ref. married)  De facto relationship −0.53 −0.52 −0.53 −0.53 −0.50  Divorced, separated, widowed −0.59 −0.58 −0.59 −0.57 −0.54  Single −0.45 −0.42 −0.44 −0.42 −0.28 Highest educ. qualification (ref. secondary educ.)  University degree or higher −1.93*** −1.93*** −1.93*** −1.93*** −2.04***  Vocational educ. certificate or equivalent −0.75 −0.76 −0.76 −0.77 −0.82  Lower than secondary education −0.66 −0.67 −0.66 −0.68 −0.67 Religiosity 0.18*** 0.18*** 0.18*** 0.18*** 0.19*** Age −0.54*** −0.53*** −0.54*** −0.53*** −0.49*** Age2 0.00 0.00 0.00 0.00 0.00 Constant 57.82*** 57.67*** 57.82*** 57.65*** 56.55*** R2 (within) 0.04 0.04 0.04 0.04 0.04 N (observations) 29,918 29,918 29,918 29,918 29,918 N (individuals) 9,583 9,583 9,583 9,583 9,583 Wald tests (p values)  βMother = βFather 0.03  βDaughter = βSon 0.05 βMother of daughter = βFather of daughter 0.04 0.04  βMother of son = βFather of son 0.35 0.94  βMother of daughter = βMother of son 0.36 0.76  βFather of daughter = βFather of son 0.05 0.03  βMother of daughter, trajectory = βFather of daughter, trajectory 0.69  βMother of son, trajectory = βFather of son, trajectory 0.13  βMother of daughter, trajectory = βMother of son, trajectory 0.33  βMother of daughter, trajectory = βFather of son, trajectory 0.35 Note: Fixed-effects models using HILDA Survey data. Higher values represent more traditional gender-role attitudes. Models also control for missing information on religion. Significance levels: ***p < 0.001 **p < 0.01 *p < 0.05 Table 2. Predictors of Gender-Role Attitudes, Fixed-Effects Models, HILDA Survey (Australia) Model 1 2 3 4 5 Effect of the transition to parenthood for…  All parents 1.91***  Male parent 2.50***  Female parent 1.39***  Parent of daughter 2.41***  Parent of son 1.39***  Female parent of daughter 1.70** 1.93***  Male parent of daughter 3.20*** 3.60***  Female parent of son 1.06* 1.68**  Male parent of son 1.75** 1.74** Effect of years since parenthood for …  Female parent of daughter −0.10  Male parent of daughter −0.15  Female parent of son −0.21**  Male parent of son −0.03 Marital status (ref. married)  De facto relationship −0.53 −0.52 −0.53 −0.53 −0.50  Divorced, separated, widowed −0.59 −0.58 −0.59 −0.57 −0.54  Single −0.45 −0.42 −0.44 −0.42 −0.28 Highest educ. qualification (ref. secondary educ.)  University degree or higher −1.93*** −1.93*** −1.93*** −1.93*** −2.04***  Vocational educ. certificate or equivalent −0.75 −0.76 −0.76 −0.77 −0.82  Lower than secondary education −0.66 −0.67 −0.66 −0.68 −0.67 Religiosity 0.18*** 0.18*** 0.18*** 0.18*** 0.19*** Age −0.54*** −0.53*** −0.54*** −0.53*** −0.49*** Age2 0.00 0.00 0.00 0.00 0.00 Constant 57.82*** 57.67*** 57.82*** 57.65*** 56.55*** R2 (within) 0.04 0.04 0.04 0.04 0.04 N (observations) 29,918 29,918 29,918 29,918 29,918 N (individuals) 9,583 9,583 9,583 9,583 9,583 Wald tests (p values)  βMother = βFather 0.03  βDaughter = βSon 0.05 βMother of daughter = βFather of daughter 0.04 0.04  βMother of son = βFather of son 0.35 0.94  βMother of daughter = βMother of son 0.36 0.76  βFather of daughter = βFather of son 0.05 0.03  βMother of daughter, trajectory = βFather of daughter, trajectory 0.69  βMother of son, trajectory = βFather of son, trajectory 0.13  βMother of daughter, trajectory = βMother of son, trajectory 0.33  βMother of daughter, trajectory = βFather of son, trajectory 0.35 Model 1 2 3 4 5 Effect of the transition to parenthood for…  All parents 1.91***  Male parent 2.50***  Female parent 1.39***  Parent of daughter 2.41***  Parent of son 1.39***  Female parent of daughter 1.70** 1.93***  Male parent of daughter 3.20*** 3.60***  Female parent of son 1.06* 1.68**  Male parent of son 1.75** 1.74** Effect of years since parenthood for …  Female parent of daughter −0.10  Male parent of daughter −0.15  Female parent of son −0.21**  Male parent of son −0.03 Marital status (ref. married)  De facto relationship −0.53 −0.52 −0.53 −0.53 −0.50  Divorced, separated, widowed −0.59 −0.58 −0.59 −0.57 −0.54  Single −0.45 −0.42 −0.44 −0.42 −0.28 Highest educ. qualification (ref. secondary educ.)  University degree or higher −1.93*** −1.93*** −1.93*** −1.93*** −2.04***  Vocational educ. certificate or equivalent −0.75 −0.76 −0.76 −0.77 −0.82  Lower than secondary education −0.66 −0.67 −0.66 −0.68 −0.67 Religiosity 0.18*** 0.18*** 0.18*** 0.18*** 0.19*** Age −0.54*** −0.53*** −0.54*** −0.53*** −0.49*** Age2 0.00 0.00 0.00 0.00 0.00 Constant 57.82*** 57.67*** 57.82*** 57.65*** 56.55*** R2 (within) 0.04 0.04 0.04 0.04 0.04 N (observations) 29,918 29,918 29,918 29,918 29,918 N (individuals) 9,583 9,583 9,583 9,583 9,583 Wald tests (p values)  βMother = βFather 0.03  βDaughter = βSon 0.05 βMother of daughter = βFather of daughter 0.04 0.04  βMother of son = βFather of son 0.35 0.94  βMother of daughter = βMother of son 0.36 0.76  βFather of daughter = βFather of son 0.05 0.03  βMother of daughter, trajectory = βFather of daughter, trajectory 0.69  βMother of son, trajectory = βFather of son, trajectory 0.13  βMother of daughter, trajectory = βMother of son, trajectory 0.33  βMother of daughter, trajectory = βFather of son, trajectory 0.35 Note: Fixed-effects models using HILDA Survey data. Higher values represent more traditional gender-role attitudes. Models also control for missing information on religion. Significance levels: ***p < 0.001 **p < 0.01 *p < 0.05 Model 1 estimates the effect on gender-role attitudes of the “base” parenthood measure. On a scale from 0 to 100, the transition to parenthood is associated with an increase of 1.91 units (p < 0.001) in the degree of traditionalism in individuals’ gender-role attitudes. Model 2 provides evidence on whether and how this parenthood effect is moderated by parental gender. Becoming a parent leads to an increase in support for traditional gender-role attitudes of 2.5 units (p < 0.001) among men and 1.39 units (p < 0.001) among women. This gender difference, whereby men traditionalize more than women upon the experience of parenthood, is statistically significant in Wald tests (p < 0.05). These results are consistent with previous studies of changes in gender-role attitudes across the transition to parenthood (see, e.g., Baxter et al. 2015; Evertsson 2013; Kroska and Elman 2009; Schober and Scott 2012). Model 3 presents the results of an initial model examining moderation in the effect of parenthood on gender-role attitudes by child’s gender. In this model, parental gender is not (yet) taken into account. Results indicate that the gender-role attitudes of individuals who become parents of firstborn daughters become more traditional across the transition to parenthood (β = 2.41, p < 0.001), and so do the gender-role attitudes of individuals who become parents of firstborn sons (β = 1.39, p < 0.001). Therefore, having a firstborn daughter seems to be more strongly associated with a traditionalization of gender-role attitudes than having a firstborn son, with this difference being statistically significant in a Wald test (p = 0.05). This pattern of results is inconsistent with Hypothesis 1 (based on interest-/exposure-based theories), but consistent with Hypothesis 2 (based on gendered societal expectations and backfire effect theories). Model 4 further considers how different permutations of parental and child’s gender affect gender-role attitudes. The coefficients on all of the parent-child gender variables are positive and statistically significant, which suggests shifts toward more traditional gender-role attitudes irrespective of parental and child’s gender. Having a firstborn daughter is associated with an increase in support for traditional gender-role attitudes of 3.2 units (p < 0.001) among men and 1.7 units (p < 0.01) among women. This gender difference, whereby men become more traditional than women upon the arrival of a firstborn girl, is statistically significant (p < 0.05). This is consistent with Hypothesis 3. Having a firstborn son is associated with an increase in support for traditional gender-role attitudes of 1.7 units (p < 0.01) among men and 1.06 units (p < 0.05) among women, with the gender difference not being statistically significant (p > 0.05). Of particular interest is whether having a daughter relative to having a son has a differential effect on attitude shifts across the transition to parenthood. For men, we find evidence that having a firstborn daughter (β = 3.2) is associated with a significantly larger increase in support for traditional gender-role attitudes than having a firstborn son (β = 1.75) (p = 0.05). For women, there are no statistically significant differences (p > 0.05) in the effects of having a firstborn daughter (β = 1.7) and having a firstborn son (β = 1.06). Model 5 examines men’s and women’s gender-role attitude trajectories after the birth of their first child. In discussing the results of this model, we focus on the estimated coefficients on the interactions between each of the parent-child gender variables and the variable capturing the number of years since the birth of the first child. These coefficients give the expected change in parental gender-role attitudes associated with an additional year since the transition to parenthood, ceteris paribus. The effects are estimated separately for individuals in each of our four parent-child gender groups to allow for heterogeneity in trends. For female and male parents of firstborn daughters as well as for male parents of firstborn sons, the number of years after the birth of the firstborn child is not statistically related to gender-role attitudes (p > 0.05). For female parents of firstborn sons, however, the model suggests a trend toward less traditional gender attitudes over time (β = −0.21, p < 0.01). When this is considered in conjunction with the estimated effect of the transition to parenthood for this subgroup (β = 1.68, p < 0.001), this suggests that, on average, mothers of firstborn sons return to their pre-parenthood gender-role attitudes when their children turn eight years of age (1.68/0.21 = 8). Nevertheless, results from Wald tests reveal that the gender-attitude trajectories for the different child-parent gender groups are not statistically different from each other. Altogether, these analyses yield virtually no evidence of trends toward more or less egalitarian gender attitudes after parenthood. Discussion & Conclusion Our base results resemble those from previous studies: becoming a parent shifts individuals’ gender-role attitudes toward more traditional views, with the effect being larger among men than women (Baxter et al. 2015; Evertsson 2013; Kroska and Elman 2009; Schober and Scott 2012). This finding is consistent with Australian and international scholarship demonstrating that children are often a turning point in how couples distribute and rationalize household labor: the arrival of children (and particularly the first child) is associated with the emergence of more traditional gender divisions (Baxter, Hewitt, and Haynes 2008; Perales, Baxter, and Tai 2015; Pollmann-Schult 2015). The traditionalizing effect of parenthood on gender-role attitudes is apparent, to different degrees, across all four permutations of parental and child’s gender. This finding extends the previous evidence base by demonstrating that parenthood is associated with a shift toward more traditional gender beliefs irrespective of the gender of firstborn children. When considered collectively, the magnitude of the parenthood effects is moderate to large, equivalent to 7 to 25 percent of the standard deviation in the gender-role attitude scale. Such effects are also seemingly larger than those of other variables, including education level and partnership status. Using interest-based and exposure-based theories of within-individual attitude change, we hypothesized that individuals with firstborn daughters would experience less traditionalization in their gender-role attitudes across the transition to parenthood than individuals with firstborn sons (Hypothesis 1). Our results provided no support for this proposition. Instead, they yielded evidence in favor of the predictions of gendered societal expectations and backfire effect theories: individuals with firstborn daughters experienced more traditionalization in their gender-role attitudes after the transition to parenthood than those with firstborn sons (Hypothesis 2). This pattern of results was apparent when considering male and female parents separately, although the pre-/post-parenthood difference in gender-role attitudes was only statistically significant among men. The latter constitutes evidence in favor of Hypothesis 3. The pattern of results in our Australian national sample is therefore inconsistent with that reported by Shafer and Malhotra (2011), the only other longitudinal study available on this topic. Their results indicated that, in one US cohort, having a daughter reduced men’s support for traditional gender roles slightly, and did not affect women’s support for such roles. However, the effect on men reported by Shafer and Malhotra was small in magnitude (about 11 percent of the variable’s standard deviation) and only marginally statistically significant (p = 0.09). While it is not possible to identify the reasons behind these differences in results, contextual and design differences across studies may have contributed to these. For example, the sample in Shafer and Malhotra (2011) represents a single cohort from the United States born between 1958 and 1965 and interviewed between 1979 and 2004, whereas our sample comprises multiple cohorts from Australia born between 1955 and 1993 and interviewed between 2001 and 2015. It is thus possible that their results apply to a bygone time in which prevailing societal-level gender attitudes and arrangements in the United States were comparatively more traditional, and the arrival of a firstborn girl would trigger new lived experiences among first-time fathers. Our finding that daughters shift parents’ (and particularly fathers’) gender-role attitudes toward less egalitarian standpoints aligns instead with predictions based on gendered societal expectations, which poses the question of why these may operate comparatively strongly in contemporary Australia. As previously hinted, this pattern of results is highly consistent with the idea that the Australian institutional context, more than that in the United States, places importance on parental (and particularly maternal) childcare. Similar to Shafer and Malhotra (2011) for the United States, our analyses of post-parenthood gender-attitude trajectories using Australian data were not highly patterned. All but one of the estimated trajectories for the different parent-child gender combinations and all of the differences in trajectories across subgroups were statistically insignificant. For mothers of firstborn sons, their attitudes were found to revert to pre-parenthood levels after their children turned eight years of age. For the remaining parent-child gender subgroups, the finding of no trends in parental gender-role attitudes after birth suggests that attitude shifts accompanying the transition to parenthood are long-lasting: individuals’ gender-role attitudes generally do not revert back to their pre-birth levels as time elapses, at least to the extent that we can observe. If parents become progressively more exposed to circumstances that challenge their gender attitudes as children grow older, these results suggest that such exposure is unlikely to be a driver of parental shifts in gender attitudes. Our analyses are, however, not without shortcomings. First, the alpha score denoting internal consistency for our attitude scale is only moderate (0.60), which may have introduced some statistical noise and effect attenuation in our estimates. However, our parameters of interest remain moderate to large in magnitude when evaluated against the standard deviation of the outcome variable (they account for 7 to 25 percent of it) and precisely estimated, which adds confidence to our findings. Second, the gender-attitude items in the HILDA Survey spread “only” over 15 years (2001–2015). Hence, individuals who become parents over the life of the survey can only be subsequently tracked for one to 14 years. This means that we can only observe changes in their post-parenthood attitudes for a limited amount of time, and that individuals who become parents early in the observation window are observed for longer spells of time. In addition, it is likely that the influence of sons and daughters in shifting parental gender-role attitudes becomes more pronounced when children are older than 14 years (McHale, Crouter, and Whiteman 2003). For example, concerns about equal access to educational opportunities, gender pay gaps, sex life, and domestic violence may not influence parents’ beliefs until their sons and daughters are old enough to have encountered these (Weitzman 2015). It follows that future studies in this area should aim to leverage longer panel datasets that track parents’ attitudes as their children move into adolescence and young adulthood. Such data may not currently exist, and so pursuing these methodological refinements may require the collection and maturation of new fit-for-purpose datasets. Another important avenue for further research consists of identifying whether the estimated effects of parent-child gender dyads on gender-role attitudes operate consistently across social strata (Lee and Conley 2016). For example, it is possible that the moderating effect of child’s gender is stronger among lowly than highly educated parents, or individuals who held comparatively less traditional gender attitudes prior to parenthood. Third, the differences in the results we find for Australia and those previously reported for the United States suggest that cross-national differences in cultural and institutional regimes may shape the processes under consideration. Hence, future studies should also examine the moderating effect of child’s gender on parental gender-role attitudes in other country contexts. More broadly, our findings illustrate the need for further studies that follow a life-course approach to the study of gender-attitude change. The principle of “linked lives” hints at the need to move away from a focus on how personal circumstances and dynamics shape individual gender-role attitudes, and into the role played by social contexts (in our application, family context). We show that there are spillover effects across family members, whereby a personal trait of children (their gender) influences the gender attitudes of their parents. Taking a broader view, it is likely that parents and children act as mutual co-influencers on each other’s worldviews as their lives unfold. Future research applying a life-course perspective to the analysis of gender-attitude change should therefore pay attention to how the attitudes of parents and children evolve in response not only to their own life events and transitions, but also in response to the life experiences of one another. Altogether, our findings indicate that in Australia a child’s gender makes a difference to how parents experience and react to parenthood, with daughters being raised in more traditionalizing households. This process may be problematic if it means that Australian girls are raised in family environments in which parents are less likely to appreciate and invest in their talents, for example, by tracking them into gender-typical educational pathways. In this scenario, the comparatively higher rates of gender-role traditionalization observed for parents of firstborn girls would result in their daughters encountering obstacles that limit their life chances not only outside but also within the family home beginning early in their life course, even if their parents are well intentioned. Such a situation may constitute an important factor hampering much needed progress toward gender equality in Australia. Appendix Table A1. Summary of Theoretical Predictions Theory/Predictions Main effect of having a daughter vs. having a son Differences by parental gender Interest-based theories The interest structures of parents of daughters shift so that they more strongly support a gender-egalitarian society: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women already benefit from gender equality prior to having a daughter: Stronger effect among men Women feel more attached to their daughters: Stronger effect among women Exposure-based theories Exposure to discriminatory behavior against daughters will make parents reassess their gender attitudes: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women are already knowledgeable about gender-based discrimination prior to parenthood: Stronger effect among men Women spend more time with their daughters and are more likely to witness gender discrimination: Stronger effect among women Gendered societal expectations The social construction of femininity leads parents to more strongly support intensive parenting of daughters: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) It is less costly for men to adhere to normative gender scripts about intensive parenting of daughters: Stronger effect among men (Hypothesis 3) Backfire effect theories People hold more feverously on to their attitudes when presented with information challenging them: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) Men’s gender-role attitudes are more conservative and thus more prone to “backfire effects”: Stronger effect among men (Hypothesis 3) Theory/Predictions Main effect of having a daughter vs. having a son Differences by parental gender Interest-based theories The interest structures of parents of daughters shift so that they more strongly support a gender-egalitarian society: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women already benefit from gender equality prior to having a daughter: Stronger effect among men Women feel more attached to their daughters: Stronger effect among women Exposure-based theories Exposure to discriminatory behavior against daughters will make parents reassess their gender attitudes: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women are already knowledgeable about gender-based discrimination prior to parenthood: Stronger effect among men Women spend more time with their daughters and are more likely to witness gender discrimination: Stronger effect among women Gendered societal expectations The social construction of femininity leads parents to more strongly support intensive parenting of daughters: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) It is less costly for men to adhere to normative gender scripts about intensive parenting of daughters: Stronger effect among men (Hypothesis 3) Backfire effect theories People hold more feverously on to their attitudes when presented with information challenging them: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) Men’s gender-role attitudes are more conservative and thus more prone to “backfire effects”: Stronger effect among men (Hypothesis 3) Table A1. Summary of Theoretical Predictions Theory/Predictions Main effect of having a daughter vs. having a son Differences by parental gender Interest-based theories The interest structures of parents of daughters shift so that they more strongly support a gender-egalitarian society: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women already benefit from gender equality prior to having a daughter: Stronger effect among men Women feel more attached to their daughters: Stronger effect among women Exposure-based theories Exposure to discriminatory behavior against daughters will make parents reassess their gender attitudes: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women are already knowledgeable about gender-based discrimination prior to parenthood: Stronger effect among men Women spend more time with their daughters and are more likely to witness gender discrimination: Stronger effect among women Gendered societal expectations The social construction of femininity leads parents to more strongly support intensive parenting of daughters: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) It is less costly for men to adhere to normative gender scripts about intensive parenting of daughters: Stronger effect among men (Hypothesis 3) Backfire effect theories People hold more feverously on to their attitudes when presented with information challenging them: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) Men’s gender-role attitudes are more conservative and thus more prone to “backfire effects”: Stronger effect among men (Hypothesis 3) Theory/Predictions Main effect of having a daughter vs. having a son Differences by parental gender Interest-based theories The interest structures of parents of daughters shift so that they more strongly support a gender-egalitarian society: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women already benefit from gender equality prior to having a daughter: Stronger effect among men Women feel more attached to their daughters: Stronger effect among women Exposure-based theories Exposure to discriminatory behavior against daughters will make parents reassess their gender attitudes: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women are already knowledgeable about gender-based discrimination prior to parenthood: Stronger effect among men Women spend more time with their daughters and are more likely to witness gender discrimination: Stronger effect among women Gendered societal expectations The social construction of femininity leads parents to more strongly support intensive parenting of daughters: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) It is less costly for men to adhere to normative gender scripts about intensive parenting of daughters: Stronger effect among men (Hypothesis 3) Backfire effect theories People hold more feverously on to their attitudes when presented with information challenging them: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) Men’s gender-role attitudes are more conservative and thus more prone to “backfire effects”: Stronger effect among men (Hypothesis 3) Table A2. Individual Items Used to Measure Gender-Role Attitudes No. Statement Reverse-coded 1 “Many working mothers seem to care more about being successful at work than meeting the needs of their children” No 2 “If both partners in a couple work, they should share equally in the housework and care of children” Yes 3 “Whatever career a woman may have, her most important role in life is still that of being a mother” No 4 “Mothers who don’t really need the money shouldn’t work” No 5 “Children do just as well if the mother earns the money and the father cares for the home and the children” Yes 6 “As long as the care is good, it is fine for children under 3 years of age to be placed in childcare all day for 5 days a week” Yes 7 “A working mother can establish just as good a relationship with her children as a mother who does not work for pay” Yes No. Statement Reverse-coded 1 “Many working mothers seem to care more about being successful at work than meeting the needs of their children” No 2 “If both partners in a couple work, they should share equally in the housework and care of children” Yes 3 “Whatever career a woman may have, her most important role in life is still that of being a mother” No 4 “Mothers who don’t really need the money shouldn’t work” No 5 “Children do just as well if the mother earns the money and the father cares for the home and the children” Yes 6 “As long as the care is good, it is fine for children under 3 years of age to be placed in childcare all day for 5 days a week” Yes 7 “A working mother can establish just as good a relationship with her children as a mother who does not work for pay” Yes Note: HILDA Survey. Table A2. Individual Items Used to Measure Gender-Role Attitudes No. Statement Reverse-coded 1 “Many working mothers seem to care more about being successful at work than meeting the needs of their children” No 2 “If both partners in a couple work, they should share equally in the housework and care of children” Yes 3 “Whatever career a woman may have, her most important role in life is still that of being a mother” No 4 “Mothers who don’t really need the money shouldn’t work” No 5 “Children do just as well if the mother earns the money and the father cares for the home and the children” Yes 6 “As long as the care is good, it is fine for children under 3 years of age to be placed in childcare all day for 5 days a week” Yes 7 “A working mother can establish just as good a relationship with her children as a mother who does not work for pay” Yes No. Statement Reverse-coded 1 “Many working mothers seem to care more about being successful at work than meeting the needs of their children” No 2 “If both partners in a couple work, they should share equally in the housework and care of children” Yes 3 “Whatever career a woman may have, her most important role in life is still that of being a mother” No 4 “Mothers who don’t really need the money shouldn’t work” No 5 “Children do just as well if the mother earns the money and the father cares for the home and the children” Yes 6 “As long as the care is good, it is fine for children under 3 years of age to be placed in childcare all day for 5 days a week” Yes 7 “A working mother can establish just as good a relationship with her children as a mother who does not work for pay” Yes Note: HILDA Survey. About the Authors Francisco (Paco) Perales is Senior Research Fellow and ARC DECRA Fellow at the Life Course Centre, Institute for Social Science Research (University of Queensland). His research focuses on understanding socio-economic inequalities by gender and sexual identity and relies on longitudinal and life-course approaches. His recent work has been published in outlets such as Social Forces, Journal of Marriage and Family, Sex Roles, European Sociological Review, Work, Employment & Society, and Social Science Research. Yara Jarallah is Postdoctoral Research Fellow at the Life Course Centre, Institute for Social Science Research (University of Queensland). Her research investigates changes in the family, including fertility and union formation, in light of political conflict and structural forces of control in the Arab world. She also works on other research fields, including gender-role attitudes, union dissolution and childbearing, adolescent health and well-being, women’s health and well-being, domestic labor, and forced migration. Janeen Baxter is Professor of Sociology and Director of the Life Course Centre, Institute for Social Science Research (University of Queensland). She has research interests in gender inequality, unpaid work, social disadvantage, well-being, and life-course and longitudinal research. She has published widely in these areas, including Negotiating the Life Course: Stability and Change in Life Pathways (Springer, 2013). References Allison , Paul . 2009 . Fixed Effects Regression Models . London : Sage . Google Scholar CrossRef Search ADS Baxter , Janeen , Sandra Buchler , Francisco Perales , and Mark Western . 2015 . “ A Life-Changing Event: First Births and Men’s and Women’s Attitudes to Mothering and Gender Divisions of Labor .” Social Forces 93 ( 3 ): 989 – 1014 . Google Scholar CrossRef Search ADS Baxter , Janeen , Belinda Hewitt , and Michele Haynes . 2008 . “ Life Course Transitions and Housework: Marriage, Parenthood, and Time on Housework .” Journal of Marriage and Family 70 ( 2 ): 259 – 72 . Google Scholar CrossRef Search ADS Bianchi , Suzanne , and Melissa Milkie . 2010 . “ Work and Family Research in the First Decade of the 21st Century .” Journal of Marriage and Family 72 ( 3 ): 705 – 25 . Google Scholar CrossRef Search ADS Bolzendahl , Catherine , and Daniel Myers . 2004 . “ Feminist Attitudes and Support for Gender Equality: Opinion Change in Women and Men, 1974–1998 .” Social Forces 83 ( 2 ): 759 – 89 . Google Scholar CrossRef Search ADS Brewster , Karin , and Irene Padavic . 2000 . “ Change in Gender-Ideology, 1977–1996: The Contributions of Intracohort Change and Population Turnover .” Journal of Marriage and Family 62 ( 2 ): 477 – 87 . Google Scholar CrossRef Search ADS Buchler , Sandra , Francisco Perales , and Janeen Baxter . 2017 . “ Does Parenthood Change Attitudes to Fathering? Evidence from Australia and Britain .” Sex Roles 77 ( 9–10 ): 663 – 675 . Google Scholar CrossRef Search ADS Cha , Youngjoo , and Sarah Thébaud . 2009 . “ Labor Markets, Breadwinning, and Beliefs: How Economic Context Shapes Men’s Gender Ideology .” Gender and Society 23 ( 2 ): 215 – 43 . Google Scholar CrossRef Search ADS Conley , Dalton , and Emily Rauscher . 2013 . “ The Effect of Daughters on Partisanship and Social Attitudes toward Women .” Sociological Forum 28 ( 4 ): 700 – 718 . Google Scholar CrossRef Search ADS Connell , R.W. , and James Messerschmidt . 2005 . “ Hegemonic Masculinity: Rethinking the Concept .” Gender and Society 19 ( 6 ): 829 – 59 . Google Scholar CrossRef Search ADS Cunningham , Mick , Ann Beutel , Jennifer Barber , and Arland Thornton . 2005 . “ Reciprocal Relationship between Attitudes about Gender and Social Contexts during Young Adulthood .” Social Science Research 34 ( 4 ): 862 – 92 . Google Scholar CrossRef Search ADS Dahl , Michael , Cristian Dezso , and David Ross . 2012 . “ Fatherhood and Managerial Style: How a Male CEO’s Children Affect the Wages of His Employees .” Administrative Science Quarterly 57 ( 4 ): 1 – 25 . Google Scholar CrossRef Search ADS Dahl , Gordon , and Enrico Moretti . 2008 . “ The Demand for Sons .” Review of Economic Studies 75 : 1085 – 1120 . Google Scholar CrossRef Search ADS Danigelis , Nicholas , Melissa Hardy , and Stephen Cutler . 2007 . “ Population Aging, Intracohort Aging, and Socio-Political Attitudes .” American Sociological Review 72 ( 5 ): 812 – 30 . Google Scholar CrossRef Search ADS Davis , Shannon , and Theodore Greenstein . 2009 . “ Gender Ideology: Components, Predictors, and Consequences .” Annual Review of Sociology 35 : 87 – 105 . Google Scholar CrossRef Search ADS Deaux , Kay . 1985 . “ Sex and Gender .” Annual Review of Psychology 36 : 49 – 81 . Google Scholar CrossRef Search ADS Downey , Douglas , Pamela Jackson , and Brian Powell . 1994 . “ Sons versus Daughters: Sex Composition of Children and Maternal Views on Socialization .” Sociological Quarterly 35 ( 1 ): 33 – 50 . Google Scholar CrossRef Search ADS Evertsson , Marie . 2013 . “ The Importance of Work: Changing Work Commitment Following the Transition to Motherhood .” Acta Sociologica 56 ( 2 ): 139 – 53 . Google Scholar CrossRef Search ADS Freese , Jeremy , and Brian Powell . 1999 . “ Sociobiology, Status, and Parental Investment in Sons and Daughters: Testing the Trivers-Willard Hypothesis .” American Journal of Sociology 104 ( 6 ): 1704 – 43 . Google Scholar CrossRef Search ADS Glynn , Adam , and Maya Sen . 2015 . “ Identifying Judicial Empathy: Does Having Daughters Cause Judges to Rule for Women’s Issues? ” American Journal of Political Science 59 ( 1 ): 37 – 54 . Google Scholar CrossRef Search ADS Kane , Emily . 2012 . The Gender Trap: Parents and the Pitfalls of Raising Boys and Girls . New York : New York University Press . Kroska , Amy , and Cheryl Elman . 2009 . “ Change in Attitudes about Employed Mothers: Exposure, Interests, and Gender Ideology Discrepancies .” Social Science Research 38 ( 2 ): 366 – 82 . Google Scholar CrossRef Search ADS PubMed Lee , Byungkyu , and Dalton Conley . 2016 . “ Does the Gender of Offspring Affect Parental Political Orientation? ” Social Forces 94 ( 3 ): 1103 – 27 . Google Scholar CrossRef Search ADS Lips , Hilary . 2001 . Sex and Gender: An Introduction , 4th ed . Mountain View : Mayfield Publishing . Lorber , Judith . 1995 . Paradoxes of Gender . New Haven, CT : Yale University Press . Lundberg , Shelly , and Elaina Rose . 2002 . “ The Effects of Sons and Daughters on Men’s Labor Supply and Wages .” Review of Economics and Statistics 84 ( 2 ): 251 – 68 . Google Scholar CrossRef Search ADS Lytton , Hugh , and David Romney . 1991 . “ Parents’ Differential Socialization of Boys and Girls: A Meta-Analysis .” Pyschological Bulletin 109 ( 2 ): 267 – 96 . Google Scholar CrossRef Search ADS McHale , Susan , Ann Crouter , and Shawn Whiteman . 2003 . “ The Family Contexts of Gender Development in Childhood and Adolescence .” Social Development 12 ( 1 ): 125 – 48 . Google Scholar CrossRef Search ADS Mikołajczak , Malgorzata , and Janina Pietrzak . 2014 . “ Ambivalent Sexism and Religion: Connected through Values .” Sex Roles 70 ( 9–10 ): 387 – 99 . Google Scholar CrossRef Search ADS PubMed Morgan , S. Phillip , Diane Lye , and Gretchen Condran . 1988 . “ Sons, Daughters, and the Risk of Marital Disruption .” American Journal of Sociology 94 ( 1 ): 110 – 29 . Google Scholar CrossRef Search ADS Moseley , Kathryn , Gary Freed , and Susan Goold . 2010 . “ Which Sources of Child Health Advice Do Parents Follow? ” Clinical Pediatrics 50 ( 1 ): 50 – 56 . Google Scholar CrossRef Search ADS PubMed Nyhan , Brendan , and Jason Reifler . 2010 . “ When Corrections Fail: The Persistence of Political Misperceptions .” Political Behavior 32 ( 2 ): 303 – 30 . Google Scholar CrossRef Search ADS Nyhan , Brendan , Jason Reifler , and P. A. Ubel . 2013 . “ The Hazards of Correcting Myths about Health Care Reform .” Medical Care 51 ( 2 ): 127 – 32 . Google Scholar CrossRef Search ADS PubMed OECD . 2016 a. Family Database: Time Used for Work, Care and Daily Household Chores. Retrieved from http://www.oecd.org/els/family/database.htm. OECD . 2016 b. Family Database: Key Characteristics of Parental Leave Systems. Retrieved from http://www.oecd.org/social/family/database.htm. Oswald , Andrew , and Nattavudh Powdthavee . 2010 . “ Daughters and Left-Wing Voting .” Review of Economics and Statistics 92 ( 2 ): 213 – 27 . Google Scholar CrossRef Search ADS Perales , Francisco , Janeen Baxter , and Tsui-o Tai . 2015 . “ Gender, Justice and Work: A Distributive Approach to Perceptions of Housework Fairness .” Social Science Research 51 : 51 – 63 . Google Scholar CrossRef Search ADS PubMed Pollmann-Schult , Matthias . 2015 . “ Sons, Daughters, and the Parental Division of Paid Work and Housework .” Journal of Family Issues 38 ( 1 ): 100 – 123 . Google Scholar CrossRef Search ADS Raley , Sara , and Suzanne Bianchi . 2006 . “ Sons, Daughters, and Family Processes: Does Gender of Children Matter? ” Annual Review of Sociology 32 : 401 – 21 . Google Scholar CrossRef Search ADS Rose , Katherine , and James Elicker . 2010 . “ Maternal Child Care Preferences for Infants, Toddlers, and Preschoolers: The Disconnect Between Policy and Preference in the USA .” Community, Work and Family 13 ( 2 ): 205 – 29 . Google Scholar CrossRef Search ADS Rossi , Alice , and Peter Rossi . 1990 . Of Human Bonding: Parent-Child Relations across the Life Course . New York : De Gruyter . Schober , Pia , and Jacqueline Scott . 2012 . “ Maternal Employment and Gender Role Attitudes: Dissonance among British Men and Women in the Transition to Parenthood .” Work, Employment and Society 26 ( 3 ): 514 – 30 . Google Scholar CrossRef Search ADS Scott , Jacqueline , Duane Alwin , and Michael Braun . 1996 . “ Generational Changes in Gender-Role Attitudes: Britain in a Cross-National Perspective .” Sociology 30 ( 3 ): 471 – 92 . Google Scholar CrossRef Search ADS Shafer , Emily , and Neil Malhotra . 2011 . “ The Effect of a Child’s Sex on Support for Traditional Gender Roles .” Social Forces 90 ( 1 ): 209 – 22 . Google Scholar CrossRef Search ADS Steiner , Leslie , ed. 2007 . Mommy Wars: Stay-at-Home and Career Moms Face Off on Their Choices, Their Lives, Their Families . New York : Random House . Summerfield , Michelle , Simon Freidin , Markus Hahn , Nathan La , Ning Li , Ninette Macalalad , Mark O’Shea , Nicole Watson , Roger Wilkins , and Mark Wooden . 2016 . “HILDA User Manual—Release 15.” Melbourne Institute of Applied Economic and Social Research, University of Melbourne. Urbatsch , R. 2009 . “ Interdependent Preferences, Militarism, and Child Gender .” International Studies Quarterly 53 ( 1 ): 1 – 21 . Google Scholar CrossRef Search ADS Vozmediano , Laura , Cesar San-Juan , Ana I. Vergara , and Natalia Alonso-Alberca . 2017 . “ Watch out, Sweetie: The Impact of Gender and Offence Type on Parents’ Altruistic Fear of Crime .” Sex Roles 77 ( 9–10 ): 676 – 86 . doi:10.1007/s11199-017-0758-7 . Google Scholar CrossRef Search ADS Warner , Rebecca L. 1991 . “ Does the Sex of Your Children Matter? Support for Feminism among Women and Men in the United States and Canada .” Journal of Marriage and the Family 53 : 1051 – 56 . Google Scholar CrossRef Search ADS Warner , Rebecca L. , and Brent Steel . 1999 . “ Child Rearing as a Mechanism for Social Change: The Relationship of Child Gender to Parent’s Commitment to Gender Equity .” Gender and Society 13 : 503 – 17 . Google Scholar CrossRef Search ADS Washington , Ebonya . 2008 . “ Female Socialization: How Daughters Affect Their Legislator Fathers’ Voting on Women’s Issues .” American Economic Review 98 ( 1 ): 311 – 32 . Google Scholar CrossRef Search ADS Weitzman , Abigail . 2015 . “ Do Fathers’ Sexual Behaviors Vary with the Sex of Firstborns? Evidence from 37 Countries .” Journal of Health and Social Behavior 56 ( 4 ): 495 – 513 . Google Scholar CrossRef Search ADS PubMed World Economic Forum . 2016 . “Rankings.” Retrieved from http://reports.weforum.org/global-gender-gap-report-2016/rankings/. © The Author(s) 2018. Published by Oxford University Press on behalf of the University of North Carolina at Chapel Hill. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Social Forces Oxford University Press

Men’s and Women’s Gender-Role Attitudes across the Transition to Parenthood: Accounting for Child’s Gender

Loading next page...
 
/lp/ou_press/men-s-and-women-s-gender-role-attitudes-across-the-transition-to-nyLCgT0XFA
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the University of North Carolina at Chapel Hill. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
0037-7732
eISSN
1534-7605
D.O.I.
10.1093/sf/soy015
Publisher site
See Article on Publisher Site

Abstract

ABSTRACT Gender-role attitudes capture individuals’ degree of support for traditional divisions of paid and domestic work and have been linked to the production and reproduction of gender inequality in different social spheres. Previous research has established that life-course transitions are related to within-individual over-time change in gender-role attitudes. Most importantly, becoming a parent is associated with shifts toward more traditional viewpoints. Theories of attitude change suggest that the gender of children should influence the pattern of gender-attitude shifts that accompany parenthood, but very few studies have investigated this. We add to this literature using Australian panel data from the Household, Income, and Labour Dynamics in Australia Survey (n = 29,918 observations) stretching over 15 years and fixed-effects panel regression models. We find that men’s and women’s gender-role attitudes become more traditional when they become parents, with evidence that this process is more pronounced among men, parents of daughters and, most of all, male parents of daughters. Introduction Gender-role attitudes capture individuals’ degree of support for traditional divisions of paid and domestic work and have been linked to the production and reproduction of gender inequality in different social spheres. This is because such attitudes influence the organization of domestic work and childcare responsibilities within households, and shape employment pathways and career aspirations in gendered ways (see Davis and Greenstein [2009] for a review). It is therefore important that we understand the factors associated with variations in individuals’ support for gender-egalitarian attitudes. Research on changes in gender-role attitudes has chiefly examined long-term trends in societal levels of gender egalitarianism, differences across cohorts, and the relative contributions of cohort-replacement and intra-cohort aging in producing attitude change at the aggregate level (Brewster and Padavic 2000; Danigelis, Hardy, and Cutler 2007). A more recent and smaller pool of studies has begun to shift attention to whether and how gender-role attitudes change within individuals over their life courses. These studies have provided compelling evidence that key life-course transitions (e.g., attaining educational qualifications, relationship entry and breakdown, and parenthood) are associated with within-individual change in gender-role attitudes (Cunningham et al. 2005; Evertsson 2013; Kroska and Elman 2009; Schober and Scott 2012). The transition to parenthood has been the subject of a great deal of attention in this literature (Baxter et al. 2015), yet few studies have paid attention to whether the child’s gender moderates parenthood effects on gender-role attitudes. This possibility has nevertheless been more thoroughly tested in relation to other types of attitudes and behaviors (see Raley and Bianchi [2006] for a review). As Lee and Conley (2016, 1104) suggest, it may be that “children socialize their parents (rather than the other way around).” As will be discussed, the notion of child’s gender being a factor influencing parental gender-role attitudes is in fact embedded in theories of gender-attitude change, including exposure-based and interest-based theories (Bolzendahl and Myers 2004; Conley and Rauscher 2013; Kroska and Elman 2009; Lee and Conley 2016), perspectives based on gendered societal expectations (Bianchi and Milkie 2010; Connell and Messerschmidt 2005; Deaux 1985; Lips 2001; Lorber 1995; Steiner 2007), and backfire effect theories (Nyhan and Reifler 2010; Nyhan, Reifler, and Ubel 2013). The existing empirical evidence is nevertheless limited and mixed with, to our knowledge, only four North American studies having examined this issue (Conley and Rauscher 2013; Downey, Jackson, and Powell 1994; Shafer and Malhotra 2011; Warner 1991). Of these, only one leverages longitudinal data (Shafer and Malhotra 2011). In this paper, we examine whether and how the traditionalizing effect of parenthood on the gender-role attitudes of men and women varies with the gender of firstborn children, considering all permutations of parents’ and child’s gender. Unlike most previous cross-sectional studies, we use panel data from the Household, Income, and Labour Dynamics in Australia (HILDA) Survey stretching over 15 years and fixed-effects panel regression models. Background Existing empirical evidence A growing literature spanning across the social sciences is concerned with the associations between the gender of children and parental and family outcomes (Raley and Bianchi 2006). For example, there are reported links between children’s gender and marital stability (Morgan, Lye, and Condran 1988), parenting practices (Lytton and Romney 1991), the allocation of household labor (Pollmann-Schult hias. 2015. Sons, Daughters, and the Parental Division of Paid Work and Housework. Journal of Family Issues 38(1):100–23." 2015), educational investments in children (Freese and Powell 1999), and parental employment patterns (Lundberg and Rose 2002). Studies have also revealed associations between children’s gender and individual partisanship (Conley and Rauscher 2013), CEO’s wage policies (Dahl, Dezso, and Ross 2012), approval of military interventions (Urbatsch 2009), and support for gender-equity policies (Warner and Steel 1999) and the conservative party (Oswald and Powdthavee 2010). Additionally, judges and legislators with daughters are more likely to vote in favor of women’s rights legislation than those with sons (Glynn and Sen 2015; Washington 2008). Specific studies on the relationship between the gender of children and parental gender-role attitudes are, however, sparse. Warner (1991) used cross-sectional data from individuals in Detroit and Toronto (n = 1,808) and found that men and women with firstborn daughters were more supportive of gender-egalitarian attitudes than men and women with firstborn sons. This association was apparent for Canadian but not American men. Similarly, Downey, Jackson, and Powell (1994) used cross-sectional data from mothers in Indiana (n = 228) and found that those with firstborn sons were more likely to support traditional gender roles than those with firstborn daughters. These studies relied on non-probability, non-nationally representative, and relatively small samples, and so their findings are not generalizable to the broader population. Conley and Rauscher (2013) were the first to use representative data from the 1994 US General Social Survey (n = 1,051) and found no evidence that having a firstborn daughter relative to a firstborn son was associated with parental gender-role attitudes. However, this and the previous studies relied on cross-sectional data to document a process (attitude change) that is inherently longitudinal, which limited their ability to assess over-time change. The data they used are now also quite old. A more recent study using US panel data from the National Longitudinal Survey of Youth 1979 (n = 3,145 individuals) was undertaken by Shafer and Malhotra (2011). This found that having a firstborn daughter (relative to having a firstborn son) slightly reduces men’s support for traditional gender roles, but has no effect on women’s support for such roles. Aims and contributions Our paper adds to the existing literature in several ways. First, while previous studies have examined the relationships between parenthood, child’s gender, and gender-role attitudes, none of them invoked the four complementary perspectives on life-course gender-attitude change that we use here (interest-based, exposure-based, gendered societal expectation, and backfire effect theories). Second, we examine the effect of child’s gender on gender-role attitudes within individuals over time using nationally representative Australian panel data. This enables us to compare the same individuals before and after the transition to parenthood, generalize our findings to the Australian population, and test the generalizability of the available North American evidence in a different socio-cultural environment. Third, we examine gender-attitude trajectories over time since entry to parenthood. This allows us to provide a more granular picture of the ways in which attitudes change over and beyond the transition to parenthood, and whether or not individuals revert to their pre-parenthood gender-role attitudes. Interest-based theories of life-course attitude change Interest-based theories of gender-attitude change rest on the assumption that individuals’ interest structures (i.e., the goals they strive for) are the driving force behind their gender beliefs (Bolzendahl and Myers 2004; Kroska and Elman 2009). It follows that, if individuals’ interest structures change, their gender-role attitudes should change in response. Importantly, the notion of “interest” in this context can be extended beyond the self to encompass significant others. For instance, if a man’s wife enters the workforce, he might benefit more from gender equality (e.g., his household income would be higher in the absence of gender pay gaps) and change his attitudes toward more gender-egalitarian beliefs as a result (Cha and Thébaud 2009). Interest-based explanations for gender-attitude change can be used to make predictions about how child’s gender may affect gender-role attitudes across the transition to parenthood. Men and women who become parents of a girl should benefit more from a gender-egalitarian society in which their daughters are treated fairly and permitted to enjoy the full range of opportunities. For example, it would be in the best interest of parents of daughters to live in a society in which intimate partner violence against women is not tolerated, or in which there are no gender pay gaps. For parents of sons, however, there may be fewer perceived advantages associated with societal gender egalitarianism. The perpetuation of the current status quo, in which girls and women remain disadvantaged in a range of life domains, may in fact result in a comparative advantage for their male sons. Hence, the interest structures of parents of girls should become more closely aligned with the goal of gender equality than the interest structures of parents of boys and, as a result, their gender-role attitudes should become comparatively more egalitarian. It is also possible that parental gender moderates how interest structures operate in this context. On the one hand, out of their own personal interest, women’s gender-role attitudes prior to the transition to parenthood may already reflect that women benefit more than men from a gender-egalitarian society. Hence, the arrival of a firstborn daughter may be associated with a stronger shift toward egalitarianism in gender-role attitudes among men, for whom their presence would constitute a more significant addition to their interest structures (Davis and Greenstein 2009). On the other hand, psychological studies on parent-child attachment have reported stronger bonds between same-gender parent-child dyads (or same-gender filial preferences), whereby fathers have a predilection for their sons and women for their daughters (McHale, Crouter, and Whiteman 2003; Raley and Bianchi 2006; Rossi and Rossi 1990). Thus, firstborn daughters may have a greater potential to shift mothers’ than fathers’ interest structures. Hence, becoming a parent of a firstborn daughter may be associated with a stronger shift toward more gender-egalitarian attitudes among women. Exposure-based theories of life-course attitude change Exposure-based theories of gender-attitude change argue that gender beliefs are rooted in ill-founded, stereotypical assumptions about women’s (and men’s) capabilities and the nature of femininity (Bolzendahl and Myers 2004). Gender-role attitudes can thus change if individuals become exposed to circumstances, situations, and experiences that challenge such assumptions (Davis and Greenstein 2009). For example, men may change their perceptions about women being ill suited to undertake certain jobs if they meet successful women at the workplace (Bolzendahl and Myers 2004). Based on exposure-based theories, it can be argued that individuals who become parents of girls will likely face situations that expose them to unfair, discriminatory behavior toward females. For example, parents may witness their daughters being tracked into gender-typical play groups and educational pathways (e.g., home economics lessons), denied access to clubs and societies (e.g., sporting clubs), or being the subject of the “male gaze” and inappropriate stereotypical or sexual comments (Kane 2012). These experiences and circumstances should make parents of girls more aware of structural inequalities unfavorable to women that emerge due to traditional gender ideologies, and should in turn lead them to question and reassess their own gender-role attitudes toward more egalitarian standpoints (see, e.g., Weitzman 2015). Parents of sons, on the other hand, should be exposed to few (if any) structural factors disadvantaging their male children, given a societal status quo that clearly favors men and masculinity. Instead, parents of sons may be more likely to encounter situations in which (hegemonic) masculinity (Connell and Messerschmidt 2005) is exalted and reinforced, such as participation in and attendance at sporting activities or consumption of male-typed entertainment and media products. Hence, exposure-based theories also lead to the prediction that men and women with firstborn daughters should experience less traditionalization in their gender-role attitudes after the transition to parenthood than men and women with firstborn sons. Using exposure-based perspectives, it is also possible to anticipate parental gender to have a moderating role. On the one hand, women may be more knowledgeable about gender-based discrimination than men due to their own experiences prior to becoming mothers, and so the addition of their daughters to their lives may entail less exposure to new situations than for men (Lee and Conley 2016; Shafer and Malhotra 2011). In these circumstances, one would expect a stronger shift toward gender egalitarianism among men. On the other hand, parents spend more time with children of their same gender (McHale, Crouter, and Whiteman 2003; Raley and Bianchi 2006; Rossi and Rossi 1990). Hence, women with firstborn girls may be more likely than men with firstborn girls to witness acts of discrimination against their daughters that prompt them to reconsider their gender-role attitudes (Bolzendahl and Myers 2004). Becoming a parent of a firstborn daughter may therefore be associated with a comparatively stronger shift toward egalitarianism in gender-role attitudes among women. Altogether, both interest- and exposure-based theories lead us to predict that: Hypothesis 1: Men and women with firstborn daughters will experience less traditionalization in their gender-role attitudes after the transition to parenthood than men and women with firstborn sons. In addition, both theories suggest that parents with firstborn girls should become more aware of gender-based discrimination and develop stronger interests in gender equality as their daughters grow older and face a greater variety of social contexts and circumstances (Shafer and Malhotra 2011). In our analyses, we will test this premise empirically by estimating models that account for time since the birth of the first child (details below). Gendered societal expectations During the early years, new parents may mainly think of their sons and daughters as dependents and receivers of care, which has implications for how the child’s interest is defined. In this context, parents may shift their worldviews to place more value on a system in which an adult is ever-present and fully committed to providing care and emotional support to the child (Rose and Elicker 2010). From the parental side, this process has a well-established and strong gender component: normative, institutionally enforced gender scripts dictate that it should be the child’s mother who adopts the main caregiver role (Bianchi and Milkie 2010; Steiner 2007). As we explain below, this process may also be gendered on the child’s side. Parents draw upon normative expectations when adapting to the requirements of and changes brought about by parenthood. This is particularly applicable to first-time parents, as they lack personal experiences on which to draw. Contemporary societal discourses around parenthood are often deeply gendered, as exemplified by well-established normative beliefs that mothers are better equipped and more capable than fathers to care for young children. Additionally, social pressures operate to make parents conform to these normative expectations, with new parents being “bombarded” with advice about parenthood and parenting by family members, friends, acquaintances, health professionals, and even strangers, as well as media channels (Moseley, Freed, and Goold 2010). In addition, there are also deeply ingrained societal discourses about the nature of boyhood and girlhood. Consistent with the social construction of femininity and masculinity in Western societies, a common theme in these discourses is the portrayal of girls as weak, fragile, passive, and dependent, and of boys as strong, able, active, and independent (Connell and Messerschmidt 2005; Deaux 1985; Lips 2001; Lorber 1995). It follows that the arrival of a firstborn daughter may elicit stronger protective and intensive parenting feelings among first-time parents than the arrival of a firstborn son. These feelings may involve more acute perceptions that the child is a delicate entity that requires parental attention, care, and protection, and that it is not appropriate for young children to attend out-of-home childcare on a full-time basis. This resonates with psychological research evidence that parents treat their daughters different than their sons in ways that reproduce gender stereotypes. For example, parents of daughters are more likely than parents of sons to discourage aggression or to display warmth toward the child, with gender-biased parental treatment being more prevalent among fathers than mothers (Raley and Bianchi 2006). The argument is also consistent with findings from criminology research that parents of girls experience fear of crime more often than parents of boys (Vozmediano et al. 2017). Due to the gendered nature of household divisions of labor and of government support to parents in countries such as Australia, for most parents the only realistic or conceivable option to engage in intensive parenting is for the mother to assume the associated responsibilities (Buchler, Perales, and Baxter 2017). In these circumstances, changes toward stronger beliefs in protective or intensive parenting across the transition to parenthood actually equate to changes toward stronger beliefs in traditional gender divisions (Rose and Elicker 2010). Therefore, one could expect shifts toward more traditional parental gender-role attitudes with the arrival of a firstborn daughter, compared to a firstborn son. A corollary is that the predicted shift toward more traditional gender-role attitudes with the birth of a firstborn girl may be more pronounced among men, for whom “traditionalizing” is less costly—it involves changing their views but not their behaviors. In fact, it would be in men’s personal benefit to traditionalize and adopt viewpoints that depict a status quo in which they are not responsible for activities that are typically not highly valued—such as routine childcare tasks. In contrast, for most women, traditionalizing involves not only reassessing their attitudes, but also reconsidering how these fit with their new roles and behaviors as mothers, which may lead to cognitive dissonance—a misalignment between one’s attitudes and behaviors that produces psychological strain (Baxter et al. 2015; Buchler, Perales, and Baxter 2017). Hence, it is less costly for men than women to “indulge” social expectations and adopt views of girls requiring more intensive parenting. This suggests that shifts toward more traditional gender-role attitudes across the transition to parenthood should be stronger among men than women with firstborn daughters. Based on this, we develop a second set of hypotheses: Hypothesis 2: Men and women with firstborn daughters will experience more traditionalization in their gender-role attitudes after the transition to parenthood than men and women with firstborn sons. Hypothesis 3: Men with firstborn daughters will experience comparatively more traditionalization in their gender-role attitudes after the transition to parenthood than women with firstborn daughters. Backfire effect theories Hypotheses 2 and 3 are consistent with predictions from backfire effect theories. These argue that, when people’s personal attitudes are based on unfounded convictions, encountering new situations or information challenging their views may actually result in people holding more strongly to their beliefs (Nyhan and Reifler 2010; Nyhan, Reifler, and Ubel 2013). Evidence of backfire effects has been found for attitudes toward the Iraq War, tax cuts, stem cell research, health care expenditure, or global warming (see e.g., Nyhan and Reifler 2010; Nyhan, Reifler, and Ubel 2013). Backfire effect theories suggest that, if parents of firstborn daughters become disproportionately exposed to situations that challenge their gender beliefs, such exposure would lead these individuals to hold on to traditional gender beliefs. Additionally, these perspectives suggest that those people who hold the most conservative gender-role attitudes prior to parenthood would more strongly hold on to them post-parenthood. This suggests that, with parenthood, the attitudes of men with firstborn daughters should become comparatively more traditional than those of women with firstborn daughters. Table A1 provides a summary of the expectations of each of the theories discussed. While our four theoretical perspectives are presented separately, it must be noted that in practice there are significant overlaps between them. For example, consistent with exposure-based theories, interest-based theories assume awareness and exposure to discriminatory practices against girls, with interest being structured around avoidance of such practices and their potential consequences on female daughters. Likewise, backfire effect theories also assume the existence of exposure to such situations, with the difference being that under this framework they are expected to elicit different psychological reactions in parents. Also, the different perspectives bear diverging temporal implications: interest-based, exposure-based, and backfire effect theories assume a progressive shift toward more traditional parental gender ideologies as children age and parents encounter new situations that challenge their gender attitudes, while gendered societal expectations theory assumes more immediate and perhaps more fleeting effects of childbirth on parental attitudes. The Australian context The available research on child’s gender and parental gender-role attitudes has, to our knowledge, exclusively relied on data from the United States and one Canadian city. An innovative aspect of our paper is our focus on a different country: Australia. Expanding the evidence base beyond the United States is important to ascertain the generalizability of the available findings, and to begin to tease out how institutional contexts may matter. Doing so, however, poses questions about whether or not the theoretical mechanisms outlined before operate similarly or differently across countries, and specifically between the United States (where most research on this topic has been conducted) and Australia (where our data come from). Interest- and exposure-based theories should operate more strongly in societal contexts in which being female is associated with a deeper degree of disadvantage. In this regard, the Australian and US contexts are very similar. For instance, Australia and the United States rank 45th and 46th (out of 144 countries) in the 2016 Global Gender Gap Index (World Economic Forum 2016). Therefore, according to interest- and exposure-based theories, we should expect similar relationships between child’s gender and parental gender-role attitudes in Australia and the United States. The gendered societal expectations argument should operate more strongly in societies in which intensive parenting ideologies prevail, and in which gender stereotypes remain deeply ingrained. While, to our knowledge, there is no cross-national comparative evidence of (dis)similarities in gender stereotyping in Australia and the United States, there is evidence to suggest that intensive parenting ideologies are more widespread in Australia than the United States. For example, both working and non-working adults in Australia spend a greater share of their time on care activities than their US counterparts (OECD 2016a). In addition, Australia features a more generous, mandated paid parental leave scheme: women are entitled to 18 weeks of paid maternity leave paid at a 42 percent payment rate (equivalent to 7.6 weeks of full-time pay), while—since 2013—men are entitled to two weeks of paid paternity leave at the same rate (equivalent to 0.8 weeks of full-time pay) (Buchler, Perales, and Baxter 2017). This imbalance in leave entitlements between Australian men and women, together with higher rates in maternal time out of employment and part-time work rates (Baxter et al. 2015), exemplifies deep-rooted normative expectations in Australia of mothers as caregivers. In the United States, however, neither mothers nor fathers are entitled to paid parental leave (OECD 2016b). The US Family and Medical Leave Act provides eligible employees with 12 weeks of annual leave for family/medical reasons, but this is not mandated to be paid leave. Taken together, these aspects suggest that in Australia, more than in the United States, normative practices and features of the institutional environment place a larger premium on parental childcare, with a strong expectation that the bulk of it will be undertaken by mothers rather than fathers. It follows that, if the gendered societal expectations hypothesis is correct, we should expect parental gender-role attitudes (and especially paternal gender-role attitudes) to move toward comparatively more traditional standpoints in Australia than in the United States. Methodology Dataset and sample selection We examine whether the gender of firstborn children affects the rate of change in parental gender-role attitudes across the transition to parenthood. To test this, we use data from the Household, Income, and Labour Dynamics in Australia (HILDA) Survey, a household panel survey that tracks individuals living in the same households in Australia for the 2001 to 2015 period and is largely representative of the Australian population in 2001 (Summerfield et al. 2016). This is one of the largest and most respected panel surveys at a global scale, and is part of the Cross-National Equivalent File of household panels. Most of the HILDA Survey data are collected via computer-assisted face-to-face interviews taking place at the respondents’ households, with some information (including that on gender attitudes) being completed privately by respondents via a self-complete instrument. This is most often handed over to the interviewer prior to her/his departure, although some respondents opt to mail it afterward. Attrition rates in the HILDA Survey are remarkably low for international standards. For instance, only around 5 percent of previous-wave respondents left the survey between its two most recent sweeps (14 and 15). Unlike in most cohort studies, individuals can enter household panels after the initial sweep. In the HILDA Survey, new individuals can join the panel if they live in participating households and become 15 years of age, or if they begin sharing a residence with a sample member. If a new panel member marries or has a child with an existing sample member, that panel member would also be followed over time. Hence, by design, not all individuals in the HILDA Survey are observed the same number of times for reasons other than panel attrition. We use data from the five HILDA Survey waves containing information on gender-role attitudes: wave one (2001), wave five (2005), wave eight (2008), wave 11 (2011), and wave 15 (2015). We consider only person-year observations in these waves in which respondents were aged between 18 and 50 years, inclusive (to focus on prime childbearing and childrearing ages) (n = 43,388). We then exclude 6,379 person-year observations from respondents who had missing data on model variables. Of these, a vast majority (over 99 percent) were dropped due to missing information on gender-role attitudes, as the self-complete questionnaire in which this is collected incurs higher non-response. We exclude also 7,096 person-year observations from respondents who were only observed once after applying the previous exclusion criteria, because we fit fixed-effect panel regression models that require at least two observations per individual (see details below). We refrain from imputing missing information at the item level because most information is missing on the outcome variable, and because of the absence of widely accepted methods to do so in a panel environment. Our final analytical sample comprises 29,918 person-year observations from 9,583 individuals. Within this sample, 3,980 individuals were observed twice over the life of the panel, 2,045 individuals were observed three times, 1,967 individuals were observed four times, and 1,590 individuals were observed all five times. This, however, does not constitute an issue for our estimation, as our fixed-effects models can handle unbalanced data. Of note, we do not exclude individuals who were parents when the HILDA Survey commenced (2001) or who were parents and entered the study later on (e.g., by joining a participating household). Because we fit fixed-effects models that only use within-individual over-time changes in the explanatory variables to estimate their model coefficients (details below), these individuals do not contribute to estimation of the model parameters on the transition to parenthood (they are always observed in the category of “parents”). Similarly, childless individuals who enter the study and are never observed to have a child also do not contribute to the estimation of the parenthood effects in our fixed-effects models (they are always observed in the category of “non-parents”). However, retaining these two types of individuals in the sample is preferable to excluding them, as they contribute to estimation of other model variables (e.g., age or education) on which they do experience change over time. This approach is customary in studies using fixed-effects models; see for example Baxter et al. (2015). Outcome variable: Gender-role attitudes Following Baxter et al. (2015), we operationalize gender-role attitudes using a time-varying variable summarizing respondents’ degree of agreement with seven separate statements (see table A2). Response options range from (1) “strongly agree” to (7) “strongly disagree,” and where necessary they were recoded so that high values always represent more traditional views about gender roles. We construct an additive scale by summing the scores of the seven items. For ease of interpretation, we rescale the resulting variable so that it ranges from 0 (most egalitarian gender attitudes) to 100 (most traditional gender attitudes). While the Cronbach’s alpha score for this scale is only moderate (0.6), factor analyses reveal that only one factor had an Eigenvalue over 1 (1.4), all items loaded positively on this factor, and the second-highest Eigenvalue among factors was very low (0.4). We take this as evidence of unidimensionality. Key explanatory variables: The transition to parenthood In our HILDA Survey sample, 1,430 men and 1,615 women become parents for the first time. Of these, 691 men and 790 women have a firstborn son, and 739 men and 825 women have a firstborn daughter. As is common practice in studies of the effects of children’s gender on parental and family outcomes, we focus exclusively on first births. This minimizes selection bias due to “endogenous stopping rules” arising from differential fertility choices and preferences for children of either gender (Dahl and Moretti 2008). In addition, a large body of research has identified first births as a distinct and critical life-course transition (Baxter et al. 2015). We do not consider cases in which first births were twins (n = 136 pairs). To reassess the relationship between the transition to parenthood and gender-role attitudes using our unique dataset, we first derive a “base” parenthood indicator. This is a time-varying dummy variable taking the value one if the respondent has been observed to have a firstborn child during the life of the panel, and the value zero otherwise. To distinguish by parental gender, we subsequently split this “base” parenthood variable into two variables: “father” is a dummy variable taking the value one if the respondent (i) is male, and (ii) has been observed to have his firstborn child during the life of the panel, and the value zero otherwise; and “mother” is a dummy variable taking the value one if the respondent (i) is female, and (ii) has been observed to have her firstborn child during the life of the panel, and the value zero otherwise. To distinguish by child’s gender, we then split the “base” parenthood variable into two new variables: “daughter” is a dummy variable taking the value one if the respondent (i) has been observed to have a firstborn child during the life of the panel, and (ii) has a firstborn girl, and the value zero otherwise; and “son” is a dummy variable taking the value one if the respondent (i) has been observed to have a firstborn child during the life of the panel, and (ii) has a firstborn boy, and the value zero otherwise. The independent variables in our main model combine information from the previous parenthood variables on the birth of the first child, the gender of the first child, and the gender of the parents into four parent-child gender variables: “male parent of daughter” is a dummy variable taking the value one if the respondent (i) is male, (ii) has been observed to have his firstborn child during the life of the panel, and (iii) has a firstborn girl; and “female parent of daughter” is a dummy variable taking the value one if the respondent (i) is female, (ii) has been observed to have her firstborn child during the life of the panel, and (iii) has a firstborn girl. In addition, “male parent of son” is a dummy variable taking the value one if the respondent (i) is male, (ii) has been observed to have his firstborn child during the life of the panel, and (iii) has a firstborn boy; and “female parent of son,” is a dummy variable taking the value one if the respondent (i) is female, (ii) has been observed to have her firstborn child during the life of the panel, and (iii) has a firstborn boy. Note that an individual who does not satisfy the criteria for a given parenthood variable would be assigned a score of zero in all survey waves. Since we use fixed-effects models estimated using change over time (see details below), such individuals would not contribute to the estimation of the regression coefficients on that variable. Control variables In our multivariate panel regression models described below, we control for a range of variables that may confound the associations between parenthood, child’s gender, and gender-role attitudes. These variables are time-varying, measured at the individual level, and are based on those used in previous studies in this field. We control for a set of dummy variables capturing marital status [married/in a de facto relationship/divorced, separated, or widowed/single (never married)], as marital status has been shown to be associated with gender-role attitudes—with married individuals displaying comparatively traditional views (Brewster and Padavic 2000). Education is associated with individuals holding more egalitarian gender attitudes (Cunningham et al. 2005), so we control for a set of dummy variables capturing respondents’ highest education qualification (university degree or higher/vocational education certificate or equivalent/secondary education/lower than secondary education). We include a control variable for religiosity, as this has recurrently been linked to more traditional gender attitudes (see, e.g., Mikołajczak and Pietrzak 2014). Our measure of religiosity is based on individuals’ responses to a question asking “On a scale from 0 to 10, how important is religion in your life?.” Finally, we control for individuals’ age at last birthday (expressed in years) and its square, as substantial literature documents trends in gender-role attitudes over the life course (see e.g., Scott, Alwin, and Braun 1996). We explicitly refrain from adjusting for factors that may themselves be a consequence of shifts in gender-role attitudes across the transition to parenthood, such as changes in employment status and subsequent births. This is because we are interested in documenting the “base” parenthood effect, rather than identifying the intervening mechanisms that produce it. In addition, while second children may strengthen traditional gender attitudes, they are naturally posterior to parenthood on the causal chain, can only be experienced by parents, and may be correlated with the gender of the first child. Hence, adding variables capturing employment status and higher-order births to the models can be seen as an instance of “over controlling” and would result in artificially downward-biased estimates of the parenthood effects. Nevertheless, results including variables capturing these aspects are available from the authors upon request. Since we fit fixed-effects regression models, we need not (and cannot) adjust for time-constant factors such as socio-economic, ethnic, or migrant background. Table 1 shows means and standard deviations for model variables. Table 1. Means and Standard Deviations for Model Variables, HILDA Survey (Australia) Women Men Mean/% SD Mean/% SD Gender-role attitudes (0–100) 39.20 15.33 43.74 14.29 Parenthood  Daughter 9% 9%  Son 8% 8% Control variables Marital status  Married 48% 46%  De facto relationship 19% 19%  Divorced, separated, widowed 8% 5%  Single 25% 30% Highest education qualification  University degree or higher 30% 24%  Vocational education certificate or equivalent 26% 36%  Secondary education 21% 20%  Lower than secondary education 23% 20% Religiosity (0–10) 3.20 3.48 2.43 3.21 Religion-missing flag 14% 16% Age in years 34.33 9.51 34.34 9.56 Years since first birth* 4.27 3.64 4.28 3.64 Women Men Mean/% SD Mean/% SD Gender-role attitudes (0–100) 39.20 15.33 43.74 14.29 Parenthood  Daughter 9% 9%  Son 8% 8% Control variables Marital status  Married 48% 46%  De facto relationship 19% 19%  Divorced, separated, widowed 8% 5%  Single 25% 30% Highest education qualification  University degree or higher 30% 24%  Vocational education certificate or equivalent 26% 36%  Secondary education 21% 20%  Lower than secondary education 23% 20% Religiosity (0–10) 3.20 3.48 2.43 3.21 Religion-missing flag 14% 16% Age in years 34.33 9.51 34.34 9.56 Years since first birth* 4.27 3.64 4.28 3.64 Note: HILDA Survey data. n (observations) = 29,918; n (individuals) = 9,583. * Only for individuals who experienced a first birth over the life of the panel. Table 1. Means and Standard Deviations for Model Variables, HILDA Survey (Australia) Women Men Mean/% SD Mean/% SD Gender-role attitudes (0–100) 39.20 15.33 43.74 14.29 Parenthood  Daughter 9% 9%  Son 8% 8% Control variables Marital status  Married 48% 46%  De facto relationship 19% 19%  Divorced, separated, widowed 8% 5%  Single 25% 30% Highest education qualification  University degree or higher 30% 24%  Vocational education certificate or equivalent 26% 36%  Secondary education 21% 20%  Lower than secondary education 23% 20% Religiosity (0–10) 3.20 3.48 2.43 3.21 Religion-missing flag 14% 16% Age in years 34.33 9.51 34.34 9.56 Years since first birth* 4.27 3.64 4.28 3.64 Women Men Mean/% SD Mean/% SD Gender-role attitudes (0–100) 39.20 15.33 43.74 14.29 Parenthood  Daughter 9% 9%  Son 8% 8% Control variables Marital status  Married 48% 46%  De facto relationship 19% 19%  Divorced, separated, widowed 8% 5%  Single 25% 30% Highest education qualification  University degree or higher 30% 24%  Vocational education certificate or equivalent 26% 36%  Secondary education 21% 20%  Lower than secondary education 23% 20% Religiosity (0–10) 3.20 3.48 2.43 3.21 Religion-missing flag 14% 16% Age in years 34.33 9.51 34.34 9.56 Years since first birth* 4.27 3.64 4.28 3.64 Note: HILDA Survey data. n (observations) = 29,918; n (individuals) = 9,583. * Only for individuals who experienced a first birth over the life of the panel. Modelling strategy: Fixed-effects panel regression models We model the relationships of interest using fixed-effects models. These are estimated by regressing deviations from individuals’ person-means in the dependent variable on deviations from their person-means in the independent variables (Allison 2009). The main model examines the effect of parenthood on gender-role attitudes for each of our “parent-child gender” variables: (GRAit−GRA̅i)=(PCGit−PCG̅i)β+(Xit−X̅i)γ+(εit−ε̅i) (1) where i and t denote individual and time, GRA stands for gender-role attitudes, PCG is a set of variables representing the four different parent-child gender combinations, X is a vector of time-varying control variables, β and γ are model coefficients, and ε is a random error term. It is important to note that, while fixed-effects models cannot incorporate time-constant variables (such as gender), they can incorporate interactions between time-constant variables and time-varying variables (see, e.g., Allison 2009, 37–38). Our variables capturing the interactions between parenthood (time-varying), parent’s gender (time-constant), and child’s gender (time-constant) fall under this banner. An extension to the previous model allows us to examine gender-role attitude trajectories for each of the different parent-child gender variables as firstborn children age. This is useful to examine whether or not, as hinted by interest- and (especially) exposure-based theories, parents become more aware of structural discrimination against girls as their daughters grow older, and consequently change their gender attitudes at a faster rate as time since birth elapses. To do so, we interact each of the parent-child gender variables with a variable capturing the number of years since the first birth (YSB): (GRAit−GRA̅i)=(PCGit−PCG̅i)β+(PCGit−PCG̅i)∗(YSBit−YSB̅i)θ+(Xit−X̅i)γ+(εit−ε̅i) (2) One could think of the θ coefficients in this model as a version of the growth parameter(s) estimated using “growth models,” as they measure trajectories in gender-role attitudes since parenthood. However, unlike those from traditional growth models, our coefficients are estimated in a fixed-effects rather than a random-effects framework, and so they are more robust to omitted-variable bias due to time-constant unobserved heterogeneity (Allison 2009). In sensitivity analyses, we tested for non-linear effects in the years since first birth variable, but found no evidence of these. We use Wald tests to examine whether parenthood impacts gender-role attitudes differently for individuals falling into the different categories of the variables capturing parental gender (“father,” “mother”), child’s gender (“daughter,” “son”), and parent-child gender dyads (“male parent of son,” “male parent of daughter,” “female parent of son,” and “female parent of daughter”). Results Table 2 presents the results of our fixed-effects models of gender-role attitudes. Positive coefficients on the independent variables indicate that the variables are associated with more traditional gender-role attitudes, whereas negative coefficients on the explanatory variables indicate that the variables are associated with more egalitarian gender-role attitudes. Across all models, the coefficients on the parenthood variables give the average difference in support for traditional gender-role attitudes across all observations before and all observations after the experience of parenthood for those individuals who become parents over the life of the panel, all else being equal. Table 2. Predictors of Gender-Role Attitudes, Fixed-Effects Models, HILDA Survey (Australia) Model 1 2 3 4 5 Effect of the transition to parenthood for…  All parents 1.91***  Male parent 2.50***  Female parent 1.39***  Parent of daughter 2.41***  Parent of son 1.39***  Female parent of daughter 1.70** 1.93***  Male parent of daughter 3.20*** 3.60***  Female parent of son 1.06* 1.68**  Male parent of son 1.75** 1.74** Effect of years since parenthood for …  Female parent of daughter −0.10  Male parent of daughter −0.15  Female parent of son −0.21**  Male parent of son −0.03 Marital status (ref. married)  De facto relationship −0.53 −0.52 −0.53 −0.53 −0.50  Divorced, separated, widowed −0.59 −0.58 −0.59 −0.57 −0.54  Single −0.45 −0.42 −0.44 −0.42 −0.28 Highest educ. qualification (ref. secondary educ.)  University degree or higher −1.93*** −1.93*** −1.93*** −1.93*** −2.04***  Vocational educ. certificate or equivalent −0.75 −0.76 −0.76 −0.77 −0.82  Lower than secondary education −0.66 −0.67 −0.66 −0.68 −0.67 Religiosity 0.18*** 0.18*** 0.18*** 0.18*** 0.19*** Age −0.54*** −0.53*** −0.54*** −0.53*** −0.49*** Age2 0.00 0.00 0.00 0.00 0.00 Constant 57.82*** 57.67*** 57.82*** 57.65*** 56.55*** R2 (within) 0.04 0.04 0.04 0.04 0.04 N (observations) 29,918 29,918 29,918 29,918 29,918 N (individuals) 9,583 9,583 9,583 9,583 9,583 Wald tests (p values)  βMother = βFather 0.03  βDaughter = βSon 0.05 βMother of daughter = βFather of daughter 0.04 0.04  βMother of son = βFather of son 0.35 0.94  βMother of daughter = βMother of son 0.36 0.76  βFather of daughter = βFather of son 0.05 0.03  βMother of daughter, trajectory = βFather of daughter, trajectory 0.69  βMother of son, trajectory = βFather of son, trajectory 0.13  βMother of daughter, trajectory = βMother of son, trajectory 0.33  βMother of daughter, trajectory = βFather of son, trajectory 0.35 Model 1 2 3 4 5 Effect of the transition to parenthood for…  All parents 1.91***  Male parent 2.50***  Female parent 1.39***  Parent of daughter 2.41***  Parent of son 1.39***  Female parent of daughter 1.70** 1.93***  Male parent of daughter 3.20*** 3.60***  Female parent of son 1.06* 1.68**  Male parent of son 1.75** 1.74** Effect of years since parenthood for …  Female parent of daughter −0.10  Male parent of daughter −0.15  Female parent of son −0.21**  Male parent of son −0.03 Marital status (ref. married)  De facto relationship −0.53 −0.52 −0.53 −0.53 −0.50  Divorced, separated, widowed −0.59 −0.58 −0.59 −0.57 −0.54  Single −0.45 −0.42 −0.44 −0.42 −0.28 Highest educ. qualification (ref. secondary educ.)  University degree or higher −1.93*** −1.93*** −1.93*** −1.93*** −2.04***  Vocational educ. certificate or equivalent −0.75 −0.76 −0.76 −0.77 −0.82  Lower than secondary education −0.66 −0.67 −0.66 −0.68 −0.67 Religiosity 0.18*** 0.18*** 0.18*** 0.18*** 0.19*** Age −0.54*** −0.53*** −0.54*** −0.53*** −0.49*** Age2 0.00 0.00 0.00 0.00 0.00 Constant 57.82*** 57.67*** 57.82*** 57.65*** 56.55*** R2 (within) 0.04 0.04 0.04 0.04 0.04 N (observations) 29,918 29,918 29,918 29,918 29,918 N (individuals) 9,583 9,583 9,583 9,583 9,583 Wald tests (p values)  βMother = βFather 0.03  βDaughter = βSon 0.05 βMother of daughter = βFather of daughter 0.04 0.04  βMother of son = βFather of son 0.35 0.94  βMother of daughter = βMother of son 0.36 0.76  βFather of daughter = βFather of son 0.05 0.03  βMother of daughter, trajectory = βFather of daughter, trajectory 0.69  βMother of son, trajectory = βFather of son, trajectory 0.13  βMother of daughter, trajectory = βMother of son, trajectory 0.33  βMother of daughter, trajectory = βFather of son, trajectory 0.35 Note: Fixed-effects models using HILDA Survey data. Higher values represent more traditional gender-role attitudes. Models also control for missing information on religion. Significance levels: ***p < 0.001 **p < 0.01 *p < 0.05 Table 2. Predictors of Gender-Role Attitudes, Fixed-Effects Models, HILDA Survey (Australia) Model 1 2 3 4 5 Effect of the transition to parenthood for…  All parents 1.91***  Male parent 2.50***  Female parent 1.39***  Parent of daughter 2.41***  Parent of son 1.39***  Female parent of daughter 1.70** 1.93***  Male parent of daughter 3.20*** 3.60***  Female parent of son 1.06* 1.68**  Male parent of son 1.75** 1.74** Effect of years since parenthood for …  Female parent of daughter −0.10  Male parent of daughter −0.15  Female parent of son −0.21**  Male parent of son −0.03 Marital status (ref. married)  De facto relationship −0.53 −0.52 −0.53 −0.53 −0.50  Divorced, separated, widowed −0.59 −0.58 −0.59 −0.57 −0.54  Single −0.45 −0.42 −0.44 −0.42 −0.28 Highest educ. qualification (ref. secondary educ.)  University degree or higher −1.93*** −1.93*** −1.93*** −1.93*** −2.04***  Vocational educ. certificate or equivalent −0.75 −0.76 −0.76 −0.77 −0.82  Lower than secondary education −0.66 −0.67 −0.66 −0.68 −0.67 Religiosity 0.18*** 0.18*** 0.18*** 0.18*** 0.19*** Age −0.54*** −0.53*** −0.54*** −0.53*** −0.49*** Age2 0.00 0.00 0.00 0.00 0.00 Constant 57.82*** 57.67*** 57.82*** 57.65*** 56.55*** R2 (within) 0.04 0.04 0.04 0.04 0.04 N (observations) 29,918 29,918 29,918 29,918 29,918 N (individuals) 9,583 9,583 9,583 9,583 9,583 Wald tests (p values)  βMother = βFather 0.03  βDaughter = βSon 0.05 βMother of daughter = βFather of daughter 0.04 0.04  βMother of son = βFather of son 0.35 0.94  βMother of daughter = βMother of son 0.36 0.76  βFather of daughter = βFather of son 0.05 0.03  βMother of daughter, trajectory = βFather of daughter, trajectory 0.69  βMother of son, trajectory = βFather of son, trajectory 0.13  βMother of daughter, trajectory = βMother of son, trajectory 0.33  βMother of daughter, trajectory = βFather of son, trajectory 0.35 Model 1 2 3 4 5 Effect of the transition to parenthood for…  All parents 1.91***  Male parent 2.50***  Female parent 1.39***  Parent of daughter 2.41***  Parent of son 1.39***  Female parent of daughter 1.70** 1.93***  Male parent of daughter 3.20*** 3.60***  Female parent of son 1.06* 1.68**  Male parent of son 1.75** 1.74** Effect of years since parenthood for …  Female parent of daughter −0.10  Male parent of daughter −0.15  Female parent of son −0.21**  Male parent of son −0.03 Marital status (ref. married)  De facto relationship −0.53 −0.52 −0.53 −0.53 −0.50  Divorced, separated, widowed −0.59 −0.58 −0.59 −0.57 −0.54  Single −0.45 −0.42 −0.44 −0.42 −0.28 Highest educ. qualification (ref. secondary educ.)  University degree or higher −1.93*** −1.93*** −1.93*** −1.93*** −2.04***  Vocational educ. certificate or equivalent −0.75 −0.76 −0.76 −0.77 −0.82  Lower than secondary education −0.66 −0.67 −0.66 −0.68 −0.67 Religiosity 0.18*** 0.18*** 0.18*** 0.18*** 0.19*** Age −0.54*** −0.53*** −0.54*** −0.53*** −0.49*** Age2 0.00 0.00 0.00 0.00 0.00 Constant 57.82*** 57.67*** 57.82*** 57.65*** 56.55*** R2 (within) 0.04 0.04 0.04 0.04 0.04 N (observations) 29,918 29,918 29,918 29,918 29,918 N (individuals) 9,583 9,583 9,583 9,583 9,583 Wald tests (p values)  βMother = βFather 0.03  βDaughter = βSon 0.05 βMother of daughter = βFather of daughter 0.04 0.04  βMother of son = βFather of son 0.35 0.94  βMother of daughter = βMother of son 0.36 0.76  βFather of daughter = βFather of son 0.05 0.03  βMother of daughter, trajectory = βFather of daughter, trajectory 0.69  βMother of son, trajectory = βFather of son, trajectory 0.13  βMother of daughter, trajectory = βMother of son, trajectory 0.33  βMother of daughter, trajectory = βFather of son, trajectory 0.35 Note: Fixed-effects models using HILDA Survey data. Higher values represent more traditional gender-role attitudes. Models also control for missing information on religion. Significance levels: ***p < 0.001 **p < 0.01 *p < 0.05 Model 1 estimates the effect on gender-role attitudes of the “base” parenthood measure. On a scale from 0 to 100, the transition to parenthood is associated with an increase of 1.91 units (p < 0.001) in the degree of traditionalism in individuals’ gender-role attitudes. Model 2 provides evidence on whether and how this parenthood effect is moderated by parental gender. Becoming a parent leads to an increase in support for traditional gender-role attitudes of 2.5 units (p < 0.001) among men and 1.39 units (p < 0.001) among women. This gender difference, whereby men traditionalize more than women upon the experience of parenthood, is statistically significant in Wald tests (p < 0.05). These results are consistent with previous studies of changes in gender-role attitudes across the transition to parenthood (see, e.g., Baxter et al. 2015; Evertsson 2013; Kroska and Elman 2009; Schober and Scott 2012). Model 3 presents the results of an initial model examining moderation in the effect of parenthood on gender-role attitudes by child’s gender. In this model, parental gender is not (yet) taken into account. Results indicate that the gender-role attitudes of individuals who become parents of firstborn daughters become more traditional across the transition to parenthood (β = 2.41, p < 0.001), and so do the gender-role attitudes of individuals who become parents of firstborn sons (β = 1.39, p < 0.001). Therefore, having a firstborn daughter seems to be more strongly associated with a traditionalization of gender-role attitudes than having a firstborn son, with this difference being statistically significant in a Wald test (p = 0.05). This pattern of results is inconsistent with Hypothesis 1 (based on interest-/exposure-based theories), but consistent with Hypothesis 2 (based on gendered societal expectations and backfire effect theories). Model 4 further considers how different permutations of parental and child’s gender affect gender-role attitudes. The coefficients on all of the parent-child gender variables are positive and statistically significant, which suggests shifts toward more traditional gender-role attitudes irrespective of parental and child’s gender. Having a firstborn daughter is associated with an increase in support for traditional gender-role attitudes of 3.2 units (p < 0.001) among men and 1.7 units (p < 0.01) among women. This gender difference, whereby men become more traditional than women upon the arrival of a firstborn girl, is statistically significant (p < 0.05). This is consistent with Hypothesis 3. Having a firstborn son is associated with an increase in support for traditional gender-role attitudes of 1.7 units (p < 0.01) among men and 1.06 units (p < 0.05) among women, with the gender difference not being statistically significant (p > 0.05). Of particular interest is whether having a daughter relative to having a son has a differential effect on attitude shifts across the transition to parenthood. For men, we find evidence that having a firstborn daughter (β = 3.2) is associated with a significantly larger increase in support for traditional gender-role attitudes than having a firstborn son (β = 1.75) (p = 0.05). For women, there are no statistically significant differences (p > 0.05) in the effects of having a firstborn daughter (β = 1.7) and having a firstborn son (β = 1.06). Model 5 examines men’s and women’s gender-role attitude trajectories after the birth of their first child. In discussing the results of this model, we focus on the estimated coefficients on the interactions between each of the parent-child gender variables and the variable capturing the number of years since the birth of the first child. These coefficients give the expected change in parental gender-role attitudes associated with an additional year since the transition to parenthood, ceteris paribus. The effects are estimated separately for individuals in each of our four parent-child gender groups to allow for heterogeneity in trends. For female and male parents of firstborn daughters as well as for male parents of firstborn sons, the number of years after the birth of the firstborn child is not statistically related to gender-role attitudes (p > 0.05). For female parents of firstborn sons, however, the model suggests a trend toward less traditional gender attitudes over time (β = −0.21, p < 0.01). When this is considered in conjunction with the estimated effect of the transition to parenthood for this subgroup (β = 1.68, p < 0.001), this suggests that, on average, mothers of firstborn sons return to their pre-parenthood gender-role attitudes when their children turn eight years of age (1.68/0.21 = 8). Nevertheless, results from Wald tests reveal that the gender-attitude trajectories for the different child-parent gender groups are not statistically different from each other. Altogether, these analyses yield virtually no evidence of trends toward more or less egalitarian gender attitudes after parenthood. Discussion & Conclusion Our base results resemble those from previous studies: becoming a parent shifts individuals’ gender-role attitudes toward more traditional views, with the effect being larger among men than women (Baxter et al. 2015; Evertsson 2013; Kroska and Elman 2009; Schober and Scott 2012). This finding is consistent with Australian and international scholarship demonstrating that children are often a turning point in how couples distribute and rationalize household labor: the arrival of children (and particularly the first child) is associated with the emergence of more traditional gender divisions (Baxter, Hewitt, and Haynes 2008; Perales, Baxter, and Tai 2015; Pollmann-Schult 2015). The traditionalizing effect of parenthood on gender-role attitudes is apparent, to different degrees, across all four permutations of parental and child’s gender. This finding extends the previous evidence base by demonstrating that parenthood is associated with a shift toward more traditional gender beliefs irrespective of the gender of firstborn children. When considered collectively, the magnitude of the parenthood effects is moderate to large, equivalent to 7 to 25 percent of the standard deviation in the gender-role attitude scale. Such effects are also seemingly larger than those of other variables, including education level and partnership status. Using interest-based and exposure-based theories of within-individual attitude change, we hypothesized that individuals with firstborn daughters would experience less traditionalization in their gender-role attitudes across the transition to parenthood than individuals with firstborn sons (Hypothesis 1). Our results provided no support for this proposition. Instead, they yielded evidence in favor of the predictions of gendered societal expectations and backfire effect theories: individuals with firstborn daughters experienced more traditionalization in their gender-role attitudes after the transition to parenthood than those with firstborn sons (Hypothesis 2). This pattern of results was apparent when considering male and female parents separately, although the pre-/post-parenthood difference in gender-role attitudes was only statistically significant among men. The latter constitutes evidence in favor of Hypothesis 3. The pattern of results in our Australian national sample is therefore inconsistent with that reported by Shafer and Malhotra (2011), the only other longitudinal study available on this topic. Their results indicated that, in one US cohort, having a daughter reduced men’s support for traditional gender roles slightly, and did not affect women’s support for such roles. However, the effect on men reported by Shafer and Malhotra was small in magnitude (about 11 percent of the variable’s standard deviation) and only marginally statistically significant (p = 0.09). While it is not possible to identify the reasons behind these differences in results, contextual and design differences across studies may have contributed to these. For example, the sample in Shafer and Malhotra (2011) represents a single cohort from the United States born between 1958 and 1965 and interviewed between 1979 and 2004, whereas our sample comprises multiple cohorts from Australia born between 1955 and 1993 and interviewed between 2001 and 2015. It is thus possible that their results apply to a bygone time in which prevailing societal-level gender attitudes and arrangements in the United States were comparatively more traditional, and the arrival of a firstborn girl would trigger new lived experiences among first-time fathers. Our finding that daughters shift parents’ (and particularly fathers’) gender-role attitudes toward less egalitarian standpoints aligns instead with predictions based on gendered societal expectations, which poses the question of why these may operate comparatively strongly in contemporary Australia. As previously hinted, this pattern of results is highly consistent with the idea that the Australian institutional context, more than that in the United States, places importance on parental (and particularly maternal) childcare. Similar to Shafer and Malhotra (2011) for the United States, our analyses of post-parenthood gender-attitude trajectories using Australian data were not highly patterned. All but one of the estimated trajectories for the different parent-child gender combinations and all of the differences in trajectories across subgroups were statistically insignificant. For mothers of firstborn sons, their attitudes were found to revert to pre-parenthood levels after their children turned eight years of age. For the remaining parent-child gender subgroups, the finding of no trends in parental gender-role attitudes after birth suggests that attitude shifts accompanying the transition to parenthood are long-lasting: individuals’ gender-role attitudes generally do not revert back to their pre-birth levels as time elapses, at least to the extent that we can observe. If parents become progressively more exposed to circumstances that challenge their gender attitudes as children grow older, these results suggest that such exposure is unlikely to be a driver of parental shifts in gender attitudes. Our analyses are, however, not without shortcomings. First, the alpha score denoting internal consistency for our attitude scale is only moderate (0.60), which may have introduced some statistical noise and effect attenuation in our estimates. However, our parameters of interest remain moderate to large in magnitude when evaluated against the standard deviation of the outcome variable (they account for 7 to 25 percent of it) and precisely estimated, which adds confidence to our findings. Second, the gender-attitude items in the HILDA Survey spread “only” over 15 years (2001–2015). Hence, individuals who become parents over the life of the survey can only be subsequently tracked for one to 14 years. This means that we can only observe changes in their post-parenthood attitudes for a limited amount of time, and that individuals who become parents early in the observation window are observed for longer spells of time. In addition, it is likely that the influence of sons and daughters in shifting parental gender-role attitudes becomes more pronounced when children are older than 14 years (McHale, Crouter, and Whiteman 2003). For example, concerns about equal access to educational opportunities, gender pay gaps, sex life, and domestic violence may not influence parents’ beliefs until their sons and daughters are old enough to have encountered these (Weitzman 2015). It follows that future studies in this area should aim to leverage longer panel datasets that track parents’ attitudes as their children move into adolescence and young adulthood. Such data may not currently exist, and so pursuing these methodological refinements may require the collection and maturation of new fit-for-purpose datasets. Another important avenue for further research consists of identifying whether the estimated effects of parent-child gender dyads on gender-role attitudes operate consistently across social strata (Lee and Conley 2016). For example, it is possible that the moderating effect of child’s gender is stronger among lowly than highly educated parents, or individuals who held comparatively less traditional gender attitudes prior to parenthood. Third, the differences in the results we find for Australia and those previously reported for the United States suggest that cross-national differences in cultural and institutional regimes may shape the processes under consideration. Hence, future studies should also examine the moderating effect of child’s gender on parental gender-role attitudes in other country contexts. More broadly, our findings illustrate the need for further studies that follow a life-course approach to the study of gender-attitude change. The principle of “linked lives” hints at the need to move away from a focus on how personal circumstances and dynamics shape individual gender-role attitudes, and into the role played by social contexts (in our application, family context). We show that there are spillover effects across family members, whereby a personal trait of children (their gender) influences the gender attitudes of their parents. Taking a broader view, it is likely that parents and children act as mutual co-influencers on each other’s worldviews as their lives unfold. Future research applying a life-course perspective to the analysis of gender-attitude change should therefore pay attention to how the attitudes of parents and children evolve in response not only to their own life events and transitions, but also in response to the life experiences of one another. Altogether, our findings indicate that in Australia a child’s gender makes a difference to how parents experience and react to parenthood, with daughters being raised in more traditionalizing households. This process may be problematic if it means that Australian girls are raised in family environments in which parents are less likely to appreciate and invest in their talents, for example, by tracking them into gender-typical educational pathways. In this scenario, the comparatively higher rates of gender-role traditionalization observed for parents of firstborn girls would result in their daughters encountering obstacles that limit their life chances not only outside but also within the family home beginning early in their life course, even if their parents are well intentioned. Such a situation may constitute an important factor hampering much needed progress toward gender equality in Australia. Appendix Table A1. Summary of Theoretical Predictions Theory/Predictions Main effect of having a daughter vs. having a son Differences by parental gender Interest-based theories The interest structures of parents of daughters shift so that they more strongly support a gender-egalitarian society: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women already benefit from gender equality prior to having a daughter: Stronger effect among men Women feel more attached to their daughters: Stronger effect among women Exposure-based theories Exposure to discriminatory behavior against daughters will make parents reassess their gender attitudes: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women are already knowledgeable about gender-based discrimination prior to parenthood: Stronger effect among men Women spend more time with their daughters and are more likely to witness gender discrimination: Stronger effect among women Gendered societal expectations The social construction of femininity leads parents to more strongly support intensive parenting of daughters: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) It is less costly for men to adhere to normative gender scripts about intensive parenting of daughters: Stronger effect among men (Hypothesis 3) Backfire effect theories People hold more feverously on to their attitudes when presented with information challenging them: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) Men’s gender-role attitudes are more conservative and thus more prone to “backfire effects”: Stronger effect among men (Hypothesis 3) Theory/Predictions Main effect of having a daughter vs. having a son Differences by parental gender Interest-based theories The interest structures of parents of daughters shift so that they more strongly support a gender-egalitarian society: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women already benefit from gender equality prior to having a daughter: Stronger effect among men Women feel more attached to their daughters: Stronger effect among women Exposure-based theories Exposure to discriminatory behavior against daughters will make parents reassess their gender attitudes: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women are already knowledgeable about gender-based discrimination prior to parenthood: Stronger effect among men Women spend more time with their daughters and are more likely to witness gender discrimination: Stronger effect among women Gendered societal expectations The social construction of femininity leads parents to more strongly support intensive parenting of daughters: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) It is less costly for men to adhere to normative gender scripts about intensive parenting of daughters: Stronger effect among men (Hypothesis 3) Backfire effect theories People hold more feverously on to their attitudes when presented with information challenging them: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) Men’s gender-role attitudes are more conservative and thus more prone to “backfire effects”: Stronger effect among men (Hypothesis 3) Table A1. Summary of Theoretical Predictions Theory/Predictions Main effect of having a daughter vs. having a son Differences by parental gender Interest-based theories The interest structures of parents of daughters shift so that they more strongly support a gender-egalitarian society: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women already benefit from gender equality prior to having a daughter: Stronger effect among men Women feel more attached to their daughters: Stronger effect among women Exposure-based theories Exposure to discriminatory behavior against daughters will make parents reassess their gender attitudes: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women are already knowledgeable about gender-based discrimination prior to parenthood: Stronger effect among men Women spend more time with their daughters and are more likely to witness gender discrimination: Stronger effect among women Gendered societal expectations The social construction of femininity leads parents to more strongly support intensive parenting of daughters: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) It is less costly for men to adhere to normative gender scripts about intensive parenting of daughters: Stronger effect among men (Hypothesis 3) Backfire effect theories People hold more feverously on to their attitudes when presented with information challenging them: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) Men’s gender-role attitudes are more conservative and thus more prone to “backfire effects”: Stronger effect among men (Hypothesis 3) Theory/Predictions Main effect of having a daughter vs. having a son Differences by parental gender Interest-based theories The interest structures of parents of daughters shift so that they more strongly support a gender-egalitarian society: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women already benefit from gender equality prior to having a daughter: Stronger effect among men Women feel more attached to their daughters: Stronger effect among women Exposure-based theories Exposure to discriminatory behavior against daughters will make parents reassess their gender attitudes: Less traditionalization in the gender-role attitudes of parents of daughters (Hypothesis 1) Women are already knowledgeable about gender-based discrimination prior to parenthood: Stronger effect among men Women spend more time with their daughters and are more likely to witness gender discrimination: Stronger effect among women Gendered societal expectations The social construction of femininity leads parents to more strongly support intensive parenting of daughters: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) It is less costly for men to adhere to normative gender scripts about intensive parenting of daughters: Stronger effect among men (Hypothesis 3) Backfire effect theories People hold more feverously on to their attitudes when presented with information challenging them: Less traditionalization in the gender-role attitudes of parents of sons (Hypothesis 2) Men’s gender-role attitudes are more conservative and thus more prone to “backfire effects”: Stronger effect among men (Hypothesis 3) Table A2. Individual Items Used to Measure Gender-Role Attitudes No. Statement Reverse-coded 1 “Many working mothers seem to care more about being successful at work than meeting the needs of their children” No 2 “If both partners in a couple work, they should share equally in the housework and care of children” Yes 3 “Whatever career a woman may have, her most important role in life is still that of being a mother” No 4 “Mothers who don’t really need the money shouldn’t work” No 5 “Children do just as well if the mother earns the money and the father cares for the home and the children” Yes 6 “As long as the care is good, it is fine for children under 3 years of age to be placed in childcare all day for 5 days a week” Yes 7 “A working mother can establish just as good a relationship with her children as a mother who does not work for pay” Yes No. Statement Reverse-coded 1 “Many working mothers seem to care more about being successful at work than meeting the needs of their children” No 2 “If both partners in a couple work, they should share equally in the housework and care of children” Yes 3 “Whatever career a woman may have, her most important role in life is still that of being a mother” No 4 “Mothers who don’t really need the money shouldn’t work” No 5 “Children do just as well if the mother earns the money and the father cares for the home and the children” Yes 6 “As long as the care is good, it is fine for children under 3 years of age to be placed in childcare all day for 5 days a week” Yes 7 “A working mother can establish just as good a relationship with her children as a mother who does not work for pay” Yes Note: HILDA Survey. Table A2. Individual Items Used to Measure Gender-Role Attitudes No. Statement Reverse-coded 1 “Many working mothers seem to care more about being successful at work than meeting the needs of their children” No 2 “If both partners in a couple work, they should share equally in the housework and care of children” Yes 3 “Whatever career a woman may have, her most important role in life is still that of being a mother” No 4 “Mothers who don’t really need the money shouldn’t work” No 5 “Children do just as well if the mother earns the money and the father cares for the home and the children” Yes 6 “As long as the care is good, it is fine for children under 3 years of age to be placed in childcare all day for 5 days a week” Yes 7 “A working mother can establish just as good a relationship with her children as a mother who does not work for pay” Yes No. Statement Reverse-coded 1 “Many working mothers seem to care more about being successful at work than meeting the needs of their children” No 2 “If both partners in a couple work, they should share equally in the housework and care of children” Yes 3 “Whatever career a woman may have, her most important role in life is still that of being a mother” No 4 “Mothers who don’t really need the money shouldn’t work” No 5 “Children do just as well if the mother earns the money and the father cares for the home and the children” Yes 6 “As long as the care is good, it is fine for children under 3 years of age to be placed in childcare all day for 5 days a week” Yes 7 “A working mother can establish just as good a relationship with her children as a mother who does not work for pay” Yes Note: HILDA Survey. About the Authors Francisco (Paco) Perales is Senior Research Fellow and ARC DECRA Fellow at the Life Course Centre, Institute for Social Science Research (University of Queensland). His research focuses on understanding socio-economic inequalities by gender and sexual identity and relies on longitudinal and life-course approaches. His recent work has been published in outlets such as Social Forces, Journal of Marriage and Family, Sex Roles, European Sociological Review, Work, Employment & Society, and Social Science Research. Yara Jarallah is Postdoctoral Research Fellow at the Life Course Centre, Institute for Social Science Research (University of Queensland). Her research investigates changes in the family, including fertility and union formation, in light of political conflict and structural forces of control in the Arab world. She also works on other research fields, including gender-role attitudes, union dissolution and childbearing, adolescent health and well-being, women’s health and well-being, domestic labor, and forced migration. Janeen Baxter is Professor of Sociology and Director of the Life Course Centre, Institute for Social Science Research (University of Queensland). She has research interests in gender inequality, unpaid work, social disadvantage, well-being, and life-course and longitudinal research. She has published widely in these areas, including Negotiating the Life Course: Stability and Change in Life Pathways (Springer, 2013). References Allison , Paul . 2009 . Fixed Effects Regression Models . London : Sage . Google Scholar CrossRef Search ADS Baxter , Janeen , Sandra Buchler , Francisco Perales , and Mark Western . 2015 . “ A Life-Changing Event: First Births and Men’s and Women’s Attitudes to Mothering and Gender Divisions of Labor .” Social Forces 93 ( 3 ): 989 – 1014 . Google Scholar CrossRef Search ADS Baxter , Janeen , Belinda Hewitt , and Michele Haynes . 2008 . “ Life Course Transitions and Housework: Marriage, Parenthood, and Time on Housework .” Journal of Marriage and Family 70 ( 2 ): 259 – 72 . Google Scholar CrossRef Search ADS Bianchi , Suzanne , and Melissa Milkie . 2010 . “ Work and Family Research in the First Decade of the 21st Century .” Journal of Marriage and Family 72 ( 3 ): 705 – 25 . Google Scholar CrossRef Search ADS Bolzendahl , Catherine , and Daniel Myers . 2004 . “ Feminist Attitudes and Support for Gender Equality: Opinion Change in Women and Men, 1974–1998 .” Social Forces 83 ( 2 ): 759 – 89 . Google Scholar CrossRef Search ADS Brewster , Karin , and Irene Padavic . 2000 . “ Change in Gender-Ideology, 1977–1996: The Contributions of Intracohort Change and Population Turnover .” Journal of Marriage and Family 62 ( 2 ): 477 – 87 . Google Scholar CrossRef Search ADS Buchler , Sandra , Francisco Perales , and Janeen Baxter . 2017 . “ Does Parenthood Change Attitudes to Fathering? Evidence from Australia and Britain .” Sex Roles 77 ( 9–10 ): 663 – 675 . Google Scholar CrossRef Search ADS Cha , Youngjoo , and Sarah Thébaud . 2009 . “ Labor Markets, Breadwinning, and Beliefs: How Economic Context Shapes Men’s Gender Ideology .” Gender and Society 23 ( 2 ): 215 – 43 . Google Scholar CrossRef Search ADS Conley , Dalton , and Emily Rauscher . 2013 . “ The Effect of Daughters on Partisanship and Social Attitudes toward Women .” Sociological Forum 28 ( 4 ): 700 – 718 . Google Scholar CrossRef Search ADS Connell , R.W. , and James Messerschmidt . 2005 . “ Hegemonic Masculinity: Rethinking the Concept .” Gender and Society 19 ( 6 ): 829 – 59 . Google Scholar CrossRef Search ADS Cunningham , Mick , Ann Beutel , Jennifer Barber , and Arland Thornton . 2005 . “ Reciprocal Relationship between Attitudes about Gender and Social Contexts during Young Adulthood .” Social Science Research 34 ( 4 ): 862 – 92 . Google Scholar CrossRef Search ADS Dahl , Michael , Cristian Dezso , and David Ross . 2012 . “ Fatherhood and Managerial Style: How a Male CEO’s Children Affect the Wages of His Employees .” Administrative Science Quarterly 57 ( 4 ): 1 – 25 . Google Scholar CrossRef Search ADS Dahl , Gordon , and Enrico Moretti . 2008 . “ The Demand for Sons .” Review of Economic Studies 75 : 1085 – 1120 . Google Scholar CrossRef Search ADS Danigelis , Nicholas , Melissa Hardy , and Stephen Cutler . 2007 . “ Population Aging, Intracohort Aging, and Socio-Political Attitudes .” American Sociological Review 72 ( 5 ): 812 – 30 . Google Scholar CrossRef Search ADS Davis , Shannon , and Theodore Greenstein . 2009 . “ Gender Ideology: Components, Predictors, and Consequences .” Annual Review of Sociology 35 : 87 – 105 . Google Scholar CrossRef Search ADS Deaux , Kay . 1985 . “ Sex and Gender .” Annual Review of Psychology 36 : 49 – 81 . Google Scholar CrossRef Search ADS Downey , Douglas , Pamela Jackson , and Brian Powell . 1994 . “ Sons versus Daughters: Sex Composition of Children and Maternal Views on Socialization .” Sociological Quarterly 35 ( 1 ): 33 – 50 . Google Scholar CrossRef Search ADS Evertsson , Marie . 2013 . “ The Importance of Work: Changing Work Commitment Following the Transition to Motherhood .” Acta Sociologica 56 ( 2 ): 139 – 53 . Google Scholar CrossRef Search ADS Freese , Jeremy , and Brian Powell . 1999 . “ Sociobiology, Status, and Parental Investment in Sons and Daughters: Testing the Trivers-Willard Hypothesis .” American Journal of Sociology 104 ( 6 ): 1704 – 43 . Google Scholar CrossRef Search ADS Glynn , Adam , and Maya Sen . 2015 . “ Identifying Judicial Empathy: Does Having Daughters Cause Judges to Rule for Women’s Issues? ” American Journal of Political Science 59 ( 1 ): 37 – 54 . Google Scholar CrossRef Search ADS Kane , Emily . 2012 . The Gender Trap: Parents and the Pitfalls of Raising Boys and Girls . New York : New York University Press . Kroska , Amy , and Cheryl Elman . 2009 . “ Change in Attitudes about Employed Mothers: Exposure, Interests, and Gender Ideology Discrepancies .” Social Science Research 38 ( 2 ): 366 – 82 . Google Scholar CrossRef Search ADS PubMed Lee , Byungkyu , and Dalton Conley . 2016 . “ Does the Gender of Offspring Affect Parental Political Orientation? ” Social Forces 94 ( 3 ): 1103 – 27 . Google Scholar CrossRef Search ADS Lips , Hilary . 2001 . Sex and Gender: An Introduction , 4th ed . Mountain View : Mayfield Publishing . Lorber , Judith . 1995 . Paradoxes of Gender . New Haven, CT : Yale University Press . Lundberg , Shelly , and Elaina Rose . 2002 . “ The Effects of Sons and Daughters on Men’s Labor Supply and Wages .” Review of Economics and Statistics 84 ( 2 ): 251 – 68 . Google Scholar CrossRef Search ADS Lytton , Hugh , and David Romney . 1991 . “ Parents’ Differential Socialization of Boys and Girls: A Meta-Analysis .” Pyschological Bulletin 109 ( 2 ): 267 – 96 . Google Scholar CrossRef Search ADS McHale , Susan , Ann Crouter , and Shawn Whiteman . 2003 . “ The Family Contexts of Gender Development in Childhood and Adolescence .” Social Development 12 ( 1 ): 125 – 48 . Google Scholar CrossRef Search ADS Mikołajczak , Malgorzata , and Janina Pietrzak . 2014 . “ Ambivalent Sexism and Religion: Connected through Values .” Sex Roles 70 ( 9–10 ): 387 – 99 . Google Scholar CrossRef Search ADS PubMed Morgan , S. Phillip , Diane Lye , and Gretchen Condran . 1988 . “ Sons, Daughters, and the Risk of Marital Disruption .” American Journal of Sociology 94 ( 1 ): 110 – 29 . Google Scholar CrossRef Search ADS Moseley , Kathryn , Gary Freed , and Susan Goold . 2010 . “ Which Sources of Child Health Advice Do Parents Follow? ” Clinical Pediatrics 50 ( 1 ): 50 – 56 . Google Scholar CrossRef Search ADS PubMed Nyhan , Brendan , and Jason Reifler . 2010 . “ When Corrections Fail: The Persistence of Political Misperceptions .” Political Behavior 32 ( 2 ): 303 – 30 . Google Scholar CrossRef Search ADS Nyhan , Brendan , Jason Reifler , and P. A. Ubel . 2013 . “ The Hazards of Correcting Myths about Health Care Reform .” Medical Care 51 ( 2 ): 127 – 32 . Google Scholar CrossRef Search ADS PubMed OECD . 2016 a. Family Database: Time Used for Work, Care and Daily Household Chores. Retrieved from http://www.oecd.org/els/family/database.htm. OECD . 2016 b. Family Database: Key Characteristics of Parental Leave Systems. Retrieved from http://www.oecd.org/social/family/database.htm. Oswald , Andrew , and Nattavudh Powdthavee . 2010 . “ Daughters and Left-Wing Voting .” Review of Economics and Statistics 92 ( 2 ): 213 – 27 . Google Scholar CrossRef Search ADS Perales , Francisco , Janeen Baxter , and Tsui-o Tai . 2015 . “ Gender, Justice and Work: A Distributive Approach to Perceptions of Housework Fairness .” Social Science Research 51 : 51 – 63 . Google Scholar CrossRef Search ADS PubMed Pollmann-Schult , Matthias . 2015 . “ Sons, Daughters, and the Parental Division of Paid Work and Housework .” Journal of Family Issues 38 ( 1 ): 100 – 123 . Google Scholar CrossRef Search ADS Raley , Sara , and Suzanne Bianchi . 2006 . “ Sons, Daughters, and Family Processes: Does Gender of Children Matter? ” Annual Review of Sociology 32 : 401 – 21 . Google Scholar CrossRef Search ADS Rose , Katherine , and James Elicker . 2010 . “ Maternal Child Care Preferences for Infants, Toddlers, and Preschoolers: The Disconnect Between Policy and Preference in the USA .” Community, Work and Family 13 ( 2 ): 205 – 29 . Google Scholar CrossRef Search ADS Rossi , Alice , and Peter Rossi . 1990 . Of Human Bonding: Parent-Child Relations across the Life Course . New York : De Gruyter . Schober , Pia , and Jacqueline Scott . 2012 . “ Maternal Employment and Gender Role Attitudes: Dissonance among British Men and Women in the Transition to Parenthood .” Work, Employment and Society 26 ( 3 ): 514 – 30 . Google Scholar CrossRef Search ADS Scott , Jacqueline , Duane Alwin , and Michael Braun . 1996 . “ Generational Changes in Gender-Role Attitudes: Britain in a Cross-National Perspective .” Sociology 30 ( 3 ): 471 – 92 . Google Scholar CrossRef Search ADS Shafer , Emily , and Neil Malhotra . 2011 . “ The Effect of a Child’s Sex on Support for Traditional Gender Roles .” Social Forces 90 ( 1 ): 209 – 22 . Google Scholar CrossRef Search ADS Steiner , Leslie , ed. 2007 . Mommy Wars: Stay-at-Home and Career Moms Face Off on Their Choices, Their Lives, Their Families . New York : Random House . Summerfield , Michelle , Simon Freidin , Markus Hahn , Nathan La , Ning Li , Ninette Macalalad , Mark O’Shea , Nicole Watson , Roger Wilkins , and Mark Wooden . 2016 . “HILDA User Manual—Release 15.” Melbourne Institute of Applied Economic and Social Research, University of Melbourne. Urbatsch , R. 2009 . “ Interdependent Preferences, Militarism, and Child Gender .” International Studies Quarterly 53 ( 1 ): 1 – 21 . Google Scholar CrossRef Search ADS Vozmediano , Laura , Cesar San-Juan , Ana I. Vergara , and Natalia Alonso-Alberca . 2017 . “ Watch out, Sweetie: The Impact of Gender and Offence Type on Parents’ Altruistic Fear of Crime .” Sex Roles 77 ( 9–10 ): 676 – 86 . doi:10.1007/s11199-017-0758-7 . Google Scholar CrossRef Search ADS Warner , Rebecca L. 1991 . “ Does the Sex of Your Children Matter? Support for Feminism among Women and Men in the United States and Canada .” Journal of Marriage and the Family 53 : 1051 – 56 . Google Scholar CrossRef Search ADS Warner , Rebecca L. , and Brent Steel . 1999 . “ Child Rearing as a Mechanism for Social Change: The Relationship of Child Gender to Parent’s Commitment to Gender Equity .” Gender and Society 13 : 503 – 17 . Google Scholar CrossRef Search ADS Washington , Ebonya . 2008 . “ Female Socialization: How Daughters Affect Their Legislator Fathers’ Voting on Women’s Issues .” American Economic Review 98 ( 1 ): 311 – 32 . Google Scholar CrossRef Search ADS Weitzman , Abigail . 2015 . “ Do Fathers’ Sexual Behaviors Vary with the Sex of Firstborns? Evidence from 37 Countries .” Journal of Health and Social Behavior 56 ( 4 ): 495 – 513 . Google Scholar CrossRef Search ADS PubMed World Economic Forum . 2016 . “Rankings.” Retrieved from http://reports.weforum.org/global-gender-gap-report-2016/rankings/. © The Author(s) 2018. Published by Oxford University Press on behalf of the University of North Carolina at Chapel Hill. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Social ForcesOxford University Press

Published: Mar 15, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off