Gender-Role Attitudes and Behavior Across the Transition to Parenthood

Abstract

Based on social structural theory and identity theory, the current study examined changes in gender-role attitudes and behavior across the first-time transition to parenthood, and following the birth of a second child for experienced mothers and fathers. Data were analyzed from the ongoing longitudinal Wisconsin Study of Families and Work (WSFW). Gender-role attitudes, work and family identity salience, and division of household labor were measured for 205 first-time and 198 experienced mothers and fathers across four time points from five months pregnant to 12 months postpartum. Multi-level latent growth curve analysis was used to analyze the data. In general, parents became more traditional in their gender-role attitudes and behavior following the birth of a child, women changed more than men, and first-time parents changed more than experienced parents. Findings suggest that changes in gender-role attitudes and behavior following the birth of a child may be attributed both to transitioning to parenthood for the first time, and to negotiating the demands of having a new baby in the family.

The transition to parenthood is one of the most salient markers of development in adults. Approximately 81% of women in the United States give birth to a child at some point in their lives (Dye, 2005). Normative life events, such as becoming a parent, are related to changes in personality and social relationships because the transition to parenthood involves tasks and demands that challenge the existing familial context (Lang, Reschke, & Neyer, 2006). The transition to parenthood, defined as giving birth to a first child, has implications for changes in gender-role attitudes, identity salience, and division of labor for both the mother and father in heterosexual couples. In addition to the couple’s level of parenting experience, each parent’s gender might also influence changes in attitudes and behavior. Most previous research has examined gender-differentiated behavior in first-time parents only; consequently, changes in parents following the birth of a child have been attributed to transitioning to parenthood. This study allowed for exploration of an alternative hypothesis positing that changes in parents following the birth of a child are more due to negotiating the demands of having a new baby in the family than to transitioning to parenthood. The current study examined gender-role attitudes, work and family identity salience, and division of household labor following the birth of a child for both first-time and experienced mothers and fathers.

Theoretical Framework

This study draws on social structural theory, which expands role theory to address the differentiation of power between men and women. Eagly and Wood (1999) developed social structural theory to challenge evolutionary theories of gender differences. The theory posits that the roles people occupy, whether due to individual choice, sociocultural pressures, or biological potentials, lead them to develop psychological qualities, and in turn behaviors to fit those roles. For example, women’s greater biological role in childbearing (pregnancy, childbirth, and lactation) coupled with cultural expectations for motherhood, place them in a different parenting role than men, who have a smaller biological contribution but still face marked cultural expectations for fatherhood, such as serving as a breadwinner. Thus, social structural theory would acknowledge that women and men may differ psychologically prior to the birth of a child because they already occupy different gender roles, but would also predict a greater divergence in psychological and behavioral characteristics with the birth of a child, which would better align individuals with their role as parent.

According to social structural theory, men and women undergo psychological change based on the degree to which their social roles are altered. Therefore, parents who are transitioning to parenthood for the first time would be expected to change more dramatically than parents having an additional child (Transition to Parenthood Hypothesis). Indeed, first-time parents have been found to perceive more change following the transition to parenthood than parents having another child (Harriman, 1983). Alternatively, with the birth of any child, parents experience the demands of having a new baby in the family and perhaps it is the negotiation of these demands that prompts psychological and behavioral change. This pattern, hereafter referred to as the New Baby Hypothesis, will be explored in the current study as an alternative to the Transition to Parenthood Hypothesis.

Transition to Parenthood and Gender-Role Attitudes

Gender-role attitudes are one aspect of psychological change that might be observed in individuals as they become parents. Gender-role attitudes can be thought of on a continuum from traditional to egalitarian (nontraditional). Traditional gender-role attitudes focus on the interdependence of partners and the distinction in power between the husband’s breadwinner role and the wife’s homemaker and mother roles, whereas nontraditional gender-role attitudes allow for egalitarian power relations and less gender-differentiated roles so that both partners can contribute equally both economically and in child-rearing (Rogers & Amato, 2000). Over the past few decades, gender-role attitudes for men and women have become less traditional (Rogers & Amato, 2000; Spain & Bianchi, 1996; Thornton, 1989); and women’s attitudes have changed more than men’s, such that women hold more egalitarian attitudes (Twenge, 1997).

Couples may shift to more traditional ways of thinking following a first-time transition to parenthood (Hoffman & Manis, 1978; Lamb, 1978), although it should be noted that these studies are more than 30 years old and phenomena may have changed. In a cross-sectional study comparing gender-role attitudes in couples who were cohabitating, married, anticipating a first child, and parenting, the parenting group had the most traditional attitudes (Abrahams, Feldman, & Nash, 1978). In addition, the gender-role attitudes of women with children under age six are more traditional than women without young children (Harris & Firestone, 1998). In the current study, we expected first-time parents to have more egalitarian attitudes than experienced parents before the birth of a child, and in general, we expected individuals to become more traditional in their gender-role attitudes across time. Based on Twenge’s (1997) findings that women’s attitudes have changed more than men’s, we also anticipated that women in the current study would have more egalitarian gender-role attitudes than men. The experience of transitioning to parenthood may depend on contextual factors, such as socioeconomic status (Antonucci & Mikus, 1988; Russell, 1974). Belsky and Kelly (1994) found that for the couples in their study, money concerns were as much a part of the transition to parenthood as navigating the demands of having a new baby in the house, such as feedings and sleep schedules. In addition, both income and education have been found to be related to gender-role attitudes (Acock & Edwards, 1982; Thornton, Alwin, & Camburn, 1983). Recognizing these potentially influential factors, we ran analyses both with and without education and income as control variables in the current study.

Parental Identity Salience

Another aspect of psychological change that might take place in individuals becoming parents is identity salience. The concept of identity salience is based on identity theory, in which an individual has multiple identities, such as parent, spouse, or worker, that are organized in a salience hierarchy, within which some identities are more important or salient than others (Pasley, Kerpelman, & Guilbert, 2001; Stryker & Serpe, 1982). Identity salience is based on commitment to a role, such that if an individual is more committed to a specific role, such as parent, that role will be more salient (Stryker & Serpe). Men and women may be more committed to specific roles because these roles are supported by society (Stryker & Serpe); and socioculturally appropriate roles may be more salient in a person’s identity hierarchy than less normative roles (Thoits, 1983). According to social structural theory, society maintains restrictions and differential opportunities for men and women, which lead to a gender-differentiated division of labor (Eagly & Wood, 1999). Social structural theory contends that gender-differentiated workforce participation results in men and women psychologically adapting to these roles. Part of this adaptation might be related to identity salience such that family salience becomes more important for women and work salience becomes more important for men.

The role of motherhood is seen by society as central to a woman’s identity. Parenthood is more salient for women’s self-conceptions than for men’s (Simon, 1992); and men perceive fathering as something they “do,” whereas women experience mothering as something they “are” (Ehrensaft, 1987). Stueve and Pleck (2001) found that in couples with children aged two to five years, compared with mothers, fathers constructed their parental identity as relatively more co-parental than solo-parental. This is consistent with previous research suggesting that fathers are seen by both parents as “helping” rather than “sharing” parental responsibilities (Cowan & Cowan, 1992; LaRossa & LaRossa, 1981). Parenthood is also salient for men, though not as salient as the worker role. The role of economic provider for men is supported by society through opportunities for work and higher pay, thus leading men to be more committed to the provider role than to the parental role.

The degree of developmental change taking place following the birth of a child depends on how much the individual invests in the parental role (Palkovitz, 1996). For individuals for whom the parental role is more salient than other roles, developmental change is expected to be more dramatic than for individuals for whom the parental role is less salient. As noted above, parenthood is more central for a woman’s self-concept than for a man’s self-concept (Simon, 1992); and women perceive more change than men following the first-time transition to parenthood (Harriman, 1983). Therefore, in the current study, while we expected both parents to increase in family salience and decrease in work salience following the birth of a child, we expected women to change more than men. Since the parental role is becoming particularly salient in first-time parents across the transition to parenthood as men and women adopt this new role, we expected identity salience to change more in first-time parents than in experienced parents.

Transition to Parenthood and Division of Labor

Changes in division of household labor often take place following the birth of a child. When men and women become parents, they tend to become more differentiated in work and family roles (Belsky & Kelly, 1994; Cowan & Cowan, 1992), and this differentiation tends to become more traditional (Goldberg & Perry-Jenkins, 2004). Specifically, women begin to work less outside of the home and perform more housework than men following the birth of a first child (Gjerdingen & Center, 2005), a pattern that continues with the addition of a subsequent child (Sanchez & Thomson, 1997). Even in couples who expected a more equal division of labor, both the mothers and fathers reported that the mothers did more of the housework and the fathers did less after the birth of a first child (Belsky & Kelly, 1994; Cowan & Cowan, 1992; Hackel & Ruble, 1992; Ruble, Fleming, Hackel, & Stangor, 1988).

Gender-differentiated patterns of division of labor may also be due to societal constraints related to parental leave, childcare availability, gender differences in income, and social role expectations. Smith and Reid (1986) suggested that societal constraints and socialization often make it hard for men and women to be egalitarian in divisions of family responsibility, even when they want to be. For instance, the gender wage gap, in which women earn approximately 81% as much as men (U.S. Department of Labor, 2005), may result in the mother staying home to care for the children and household because it makes more financial sense to lose her income than the father’s income. Indeed, although the majority of women continue to work after having children (Cohany & Sok, 2007), women are more likely to stop working or become part-time workers, whereas men spend more time working (Cowan & Cowan, 1992; Cowan et al., 1985). Likewise, although division of household labor has become more equal over time in that men’s contribution to housework has increased and women’s contribution has decreased (Rogers & Amato, 2000), women are still more likely to perform household labor than men, and in practice, the division of work and household labor is still gender-differentiated (Duncan, Edwards, Reynolds, & Alldred, 2003). Therefore, in the current study, we expected women to do a greater share of housework across the transition to parenthood than men.

The Current Study

Across existing studies on gender and the transition to parenthood, the majority of research studied first-time parents only, and this body of research has concluded that changes in parents following the birth of a child are due to transitioning to parenthood (Transition to Parenthood Hypothesis). By including both first-time parents and experienced parents having a second child, the current study allowed for exploration of an alternative explanation that changes are more due to negotiation of the demands of having a new baby in the family (New Baby Hypothesis), and that changes will also occur in parents having a second child. The experienced parents group was limited to couples having a second child to avoid the potential confound of including parents with more than two children, who may have more traditional attitudes, manifested in having a larger family. Most previous studies have examined gender-differentiated behavior instead of gender-role attitudes, and past research has not examined identity salience and gender-role attitudes together. The current study addressed this gap in the research by examining gender-role attitudes, identity salience, and gender-differentiated behavior, as measured by division of labor. The inclusion of all three of these variables allowed for a more in-depth examination of the psychological and behavioral changes associated with having a child.

The aim of the current study was to examine changes in gender-role attitudes, identity salience, and division of labor across the first-time transition to parenthood and following the birth of a second child for experienced mothers and fathers. This study also examined how these processes may be influenced by each parent’s gender and by the couple’s level of parenting experience. Gender-role attitudes, identity salience, and division of household labor were expected to become more traditional for both first-time and experienced mothers and fathers from pre-birth through the first year postpartum. Attitudes, identity salience, and division of labor were expected to change more for first-time parents than for experienced parents, and more for mothers than for fathers.

To address the research questions, data were analyzed from the longitudinal Wisconsin Study of Families and Work (WSFW) (Hyde, Klein, Essex, & Clark, 1995). Gender-role attitudes, work and family identity salience, and division of household labor were measured in both first-time and experienced mothers and fathers across four time points from five months pregnant to 12 months postpartum.

Method

Participants

570 pregnant women and 550 male partners were originally recruited for participation in the Wisconsin Maternity Leave and Health (WMLH) Project, now called the Wisconsin Study of Families and Work (WSFW) (Hyde et al., 1995). In the current study, participants were limited to first-time parents and experienced parents having a second child; participants were excluded from analyses if the couple had more than two children, had any non-biological children living with them, or if the index child was a twin. Data were analyzed for 403 couples: 205 first-time parents and 198 experienced parents having a second child. Background characteristics of the sample are reported in Table 1. Participants were recruited in the Milwaukee and Madison Wisconsin Standard Metropolitan Statistical Area (SMSA) through obstetrical and hospital clinics.

Table 1.

Background Characteristics of the Sample at Time 1

	First-Time		Experienced
Measures	Mothers	Fathers	Mothers	Fathers
Age (Mean)	27.71 (4.04)	28.78 (4.23)	29.71 (5.68)	30.96 (4.38)
Ethnicity (%)
White (Not Hispanic)	92.6	94.0	93.5	91.5
African American (Not Hispanic)	2.4	4.0	3.5	4.8
Hispanic	1.0	0	0	0
Native American	1.5	0	1.5	1.6
Asian American	1.0	0	0	1.6
Other	1.5	2.0	1.5	.5
Years of Education (%)
8–12	11.4	17.5	13.5	17.5
13–16	68.1	55.4	61.6	57.4
17–20	20.5	27.1	24.9	25.1
Household Income (%)
<$20,000	5.0	−	3.6	−
$20–50,000	50.0	−	51.1	−
$50–80,000	45.0	−	45.3	−
Marital Status of Mother (%)
Married	90.7	−	98.5	−
Never Married	8.3	−	1.5	−
Divorced	1.0	−	0	−
Duration of Partnership (Mean)	3.33 (2.37)	−	5.68 (2.74)	−

Note. Where means are noted, standard deviations are in parentheses. Age range was 20–41 for mothers and 20–48 for fathers. Household income was measured for the couple and is only reported for mothers. Marital status was only measured for the mother and duration of partnership was measured by number of years the mother was living with the father at the time of recruitment. Range of duration of partnership was <1–18 years for experienced mothers and <1–15 years for first-time mothers.

To be eligible for participation, women had to meet the following criteria: (a) over the age of 18; (b) between weeks 12 and 21 of pregnancy (second trimester); (c) not disabled in any way that would significantly alter physical functioning as a parent (e.g., in a wheelchair); (d) living with the partner, though not necessarily married; (e) at least one member of the couple working for pay or profit; (f) not a student; (g) not self-reported as “unemployed”; (h) in possession of a telephone in order to set up interview appointments; (i) fluency in English well enough to understand the interviewer; and (j) sufficiently literate to complete paper-and-pencil questionnaires.

All patients in the second trimester of pregnancy were identified daily by clinic staff and approached for participation in the study by a recruiter from the WMLH Project staff. Of the women eligible for participation, 75% agreed to be in the study. Overall, the attrition rate from Time 1 (T1) to Time 4 (T4) was 10% for mothers and 14% for fathers. Attrition analyses were conducted to compare participants who dropped out of the study by T4 with the original sample at T1. T-tests on gender-role attitudes, work and family salience, and division of household labor revealed that the scores of mothers and fathers who dropped out did not differ significantly from those of participants who did not drop out. A chi-square test on parental experience was also non-significant and revealed that first-time and experienced parents were equally likely to have dropped out of the study by T4.

Design and Procedure

The women were interviewed in their homes by a female interviewer, and the men were interviewed via telephone by a male interviewer on four separate occasions: (a) during the second trimester of pregnancy (T1), (b) one month postpartum (T2), (c) four months postpartum (T3), and (d) 12 months postpartum (T4). In addition, mothers and fathers completed mail-out questionnaires at each of the interview times. Division of household labor was collected via home and telephone interviews, and gender-role attitudes and work and family identity salience were collected via mail-out questionnaires. Differences in data quality based on mode of administration of the measures in the current study were not expected based on findings from a meta-analysis on data quality in face-to-face interviews versus telephone interviews, in which the authors found that any differences in the two modes were small (de Leeuw & van der Zouwen, 1988).

Measures

Gender-role attitudes

Gender-role attitudes were measured using the Traditional-Egalitarian Sex Role (TESR) scale (Larsen & Long, 1988). The scale consists of 20 items, measured on a 5-point scale from strongly agree (1) to strongly disagree (5). High scores indicate egalitarian gender-role attitudes and agreement with statements such as “It is just as important to educate daughters as it is to educate sons.” Low scores indicate traditional gender-role attitudes and agreement with statements such as “Ultimately a woman should submit to her husband’s decision.” For the WMLH sample, reliability at T1 as measured by coefficient alpha was .84 for mothers and .87 for fathers.

Identity salience

To measure work and family identity salience, two scales from the Salience Inventory (Nevill & Super, 1986) were used. Each scale consists of 10 items, one set for the home and family role and one set for the work role. Each item is measured on a 4-point scale from little or none (1) to a great deal (4). An example of a family salience item is “I am very much involved in home and family activities.” An example of a work salience item is “I really feel personally involved in work.” For the WMLH sample, reliability at T1 for fathers as measured by coefficient alpha was .92 for family salience and .92 for work salience; for mothers, family salience was .92 and work salience was .91.

Division of labor

Division of household labor was measured using questions developed for the WMLH Project. These questions included nine items assessing division of labor for household tasks. Each item was rated on a 5-point scale from you do all of it (1) to your partner does all of it (5), with (3) being you divide it equally, and another option of you hire someone to do it, which was treated as missing data. Examples of division of household labor items are “How do you and your wife/husband/partner divide each of the following household and family jobs: Cleaning the house; paying the bills?”

Analytic Technique

The following analyses were conducted using latent growth curve models in Mplus (Muthén & Muthén, 2007), which have several distinct advantages over traditional statistical techniques based on ordinary least squares and the generalized linear model, such as analysis of variance (ANOVA) (Kaplan, 2000). First, latent growth curves allow one to model trajectories in a given construct, such as the pattern of changes over time in gender-role attitudes, division of labor, or identity salience, if measures were collected at multiple times (Willett & Sayer, 1994). Second, path models allow for simultaneous analysis of all parts of a model, such as mediating processes (Kaplan, 2000), rather than needing to parse tests of mediation into three or more analyses. Finally, sophisticated statistical programs such as Mplus offer state-of-the-art methods for handling missing values using an approach proposed by Arbuckle (1996), in which separate means and covariances are calculated for each group of participants who have the same pattern of data completion and then combined to predict the log likelihood for the full sample using full information maximum likelihood. This technique is particularly useful given the common problem of missing values in longitudinal studies, and allowed us to use data from all participants who completed at least one assessment.

Observations gathered in the present study violate the assumption of independence because two sets of measures come from each couple. We accounted for this nested design in two ways. Analyses concerning gender-role attitudes and identity salience were conducted using a clustering option available in Mplus. In analyses that examined division of labor, father’s scores were reverse coded and then scores for mothers and fathers were averaged because they measure a single question of interest – the gender-differentiated division of labor. Notably, couples’ reports of division of labor were significantly negatively correlated, such that both parents agreed that the mother did more than the father. Correlations between mothers’ and fathers’ reports were −.70 at T1, −.69 at T2, −.73 at T3, and −.73 at T4, p < .001.

In each latent growth curve model, data from assessment waves were used to estimate latent “intercepts” and “slopes”. The intercept represents scores at one point in time, and can be set to a time point of interest, such as the beginning of the current study, at the second trimester of pregnancy. The slope represents the trajectory or pattern of change over time in a construct, such as linear or quadratic effects. Linear and quadratic effects were constructed in a manner to account for the non-equidistant lengths of time between assessment waves. Multiple fit indices were used to assess goodness of fit, including the Comparative Fit Index (CFI; ideal >.95), the Tucker-Lewis Index (TLI; ideal >.95), the root mean squared error of approximation (RMSEA; ideal <.06), and the χ² test of model fit (ideally non-significant).

Finally, there does not yet exist a standard way to calculate effect sizes in latent growth curve models. In an effort to provide effect sizes, however, we have reported a special type of standardized coefficient available in Mplus (called “StdY”) that represents the difference between groups (e.g., gender difference) in standard deviation units of the dependent variable (e.g., intercept, linear effect, or quadratic effect) (Muthén & Muthén, 2007). Therefore, these effect sizes can be interpreted in a manner similar to that of Cohen’s d. The signs of the effect sizes match the signs of the coefficients.

Results

The main objective of the current study was to examine changes in gender-role attitudes and behavior across the first-time transition to parenthood and following the birth of a second child for experienced mothers and fathers, by examining measures of gender-role attitudes, identity salience, and division of labor. Patterns of overall change are depicted in Figure 1, and correlations among all variables for mothers and fathers are provided in Table 2.

Figure 1.

Changes in gender-role attitudes, work and family salience, and division of household labor for first-time (FT) and experienced (Exper) mothers and fathers across four assessment waves from pre-birth to one year postpartum: second trimester (2^nd Tri), one month postpartum (1 Mo PP), four months postpartum (4 Mo PP), and 12 months postpartum (12 Mo PP).

Table 2.

Descriptive Statistics and Correlations Among Gender-Role Attitudes, Family and Work Salience, and Division of Household Labor for Mothers and Fathers at 5 Months Pregnant (T1) to 12 Months Postpartum (T4)

Measures	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16
1. T1 Attitudes	.48**	.84**	.83**	.80**	−.00	.03	.04	.01	.38**	.33**	.31**	.23**	−.01	−.02	−.01	−.00
2. T2 Attitudes	.82**	.52**	.83**	.81**	−.01	.04	.02	−.04	.30**	.32**	.29**	.21**	−.02	−.03	.05	.03
3. T3 Attitudes	.80**	.81**	.45**	.83**	.02	.04	.03	.02	.29**	.29**	.30**	.24**	.01	−.01	.03	.07
4. T4 Attitudes	.79**	.82**	.80**	.50**	−.03	−.03	−.03	−.04	.29**	.28**	.24**	.22**	.02	−.02	.01	.02
5. T1 Fam Sal	.20**	.20**	.16**	.17**	.18**	.51**	.55**	.49**	.10*	.04	.04	.06	−.08	−.03	−.09	−.04
6. T2 Fam Sal	.18**	.27**	.26**	.22**	.60**	.11*	.61**	.54**	.06	.12*	.12*	.10	−.20**	−.09	−.15**	−.12*
7. T3 Fam Sal	.17**	.19**	.25**	.19**	.61**	.70**	.03	.57**	.12*	.12*	.10*	.09	−.15**	−.02	−.12*	−.06
8. T4 Fam Sal	.14*	.11*	.17**	.15**	.55**	.67**	.72**	.07	.05	.05	.05	.11*	−.03	.02	−.06	.01
9. T1 Wk Sal	.00	.08	.06	.02	.24**	.15**	.22**	.18**	.03	.67**	.67**	.57**	.04	.00	−.03	.05
10. T2 Wk Sal	.06	.15**	.12*	.06	.18**	.33**	.28**	.17**	.66**	.16**	.67**	.62**	−.05	−.11*	−.05	.03
11. T3 Wk Sal	.02	.07	.12*	.03	.19**	.25**	.28**	.20**	.66**	.71**	.16**	.69**	−.05	−.08	−.05	−.00
12. T4 Wk Sal	.05	.09	.08	.05	.20**	.25**	.27**	.27**	.68**	.69**	.72**	.13*	−.01	−.02	−.00	.05
13. T1 Div HH	−.03	.01	−.01	−.01	−.04	−.07	−.10	−.10	.11*	.08	.13*	.15**	.70**	.66**	.71**	.71**
14. T2 Div HH	−.02	.01	.00	.01	−.02	−.07	−.02	−.02	.12*	.07	.15**	.17**	.63**	.69**	.67**	.62**
15. T3 Div HH	−.05	−.07	−.04	−.07	−.04	−.06	−.08	−.08	.13*	.11*	.16**	.17**	.67**	.68**	.73**	.74**
16. T4 Div HH	−.02	−.05	−.02	−.01	.00	−.08	−.08	−.08	.20**	.14*	.16**	.17**	.67**	.67**	.74**	.73**
Mothers M(SD)	87.81 (9.27)	87.22 (9.57)	86.07 (9.81)	86.37 (9.74)	35.17 (4.69)	36.10 (4.66)	36.54 (4.50)	36.07 (4.34)	29.42 (5.74)	27.05 (6.54)	26.79 (6.30)	26.65 (6.39)	2.88 (.48)	3.00 (.57)	2.79 (.55)	2.75 (.53)
Fathers M(SD)	79.50 (11.64)	78.07 (11.97)	76.81 (11.68)	77.06 (12.19)	34.80 (4.89)	35.03 (4.75)	34.54 (5.04)	34.56 (4.80)	32.08 (5.63)	31.40 (5.98)	30.85 (6.01)	30.29 (5.76)	2.81 (.41)	2.72 (.45)	2.82 (.45)	2.86 (.45)

Note. The data used to compute these descriptive statistics and correlations included first-time parents and experienced parents with 1 previous biological child. Correlations for mothers are above the diagonal; correlations for fathers are below the diagonal. Correlations on the diagonal are between the mothers’ and fathers’ scores. Wk Sal: Work Salience, Fam Sal: Family Salience, Div HH: Division of Household Labor.

^*p < .05,

^**p < .0.1. Mean scores for all measures are reported at the bottom of the table, with standard deviations in parentheses. Higher scores represent more egalitarian gender-role attitudes, Range: 20–100; more commitment to the home and family role or work role, Range 10–40; and completion of a greater share of housework by the father than by the mother with 3 representing equal division of housework, Range: 1–5.

The coefficients, standard errors, and effect sizes for the latent growth curve models are presented in Table 3. The first results reported are the overall intercept, linear effect, and quadratic effect for the measure of interest (attitudes, salience, or division of labor). The overall intercept represents the latent mean during the second trimester of pregnancy (Time 1) across all participants and tests whether the group average on the measure of interest was significantly different from zero at this time. The overall linear and quadratic effects test whether there was linear or quadratic change over time, and the sign of the coefficients indicates whether that change was increasing (a positive coefficient) or decreasing (a negative coefficient) over time.

Table 3.

Raw Coefficients and Model Fit Statistics for Latent Growth Models Examining Gender-Role (GR) Attitudes and Identity Salience Across the Transition to Parenthood

	Gender-Role Attitudes			Family Salience			Work Salience
	Coefficient	Std. Error	Effect Size	Coefficient	Std. Error	Effect Size	Coefficient	Std. Error	Effect Size
Overall Intercept	87.605*	1.429		34.992*	0.615		29.285*	0.753
Overall Linear Effect	−0.389 *	0.157		0.233*	0.102		−0.541*	0.121
Overall Quadratic Effect	0.023*	0.008		−0.010‡	0.006		0.018*	0.007
Intercept regressed onto
Parent Gender	−8.855*	0.838	−1.148	−0.007	0.454	−0.002	2.107*	0.572	0.447
Parenting Experience	−1.374	0.930	−0.191	1.214*	0.460	0.654	−1.437*	0.573	−0.803
Parent Gender X Exper	0.515	1.137	0.067	−0.729	0.618	−0.217	1.131	0.807	0.240
Child Gender	0.701	0.911	0.097	−0.305	0.375	−0.164	0.557	0.411	0.311
Linear Effect regressed onto
Parent Gender	−0.012	0.126	−0.062	−0.186*	0.084	−1.157	0.371*	0.090	1.862
Parenting Experience	0.169	0.104	2.110	−0.180*	0.086	−1.777	0.270*	0.091	1.983
Parent Gender X Exper	−0.256	0.176	−1.314	−0.098	0.117	−0.612	−0.066	0.130	−0.330
Child Gender	0.020	0.088	0.250	0.065	0.063	0.639	−0.046	0.069	−0.337
Quadratic Effect regressed onto
Parent Gender	−0.002	0.007	−0.453	0.008‡	0.004	1.521	−0.015*	0.005	−1.909
Parenting Experience	−0.012*	0.006	−1.840	0.009‡	0.005	1.880	−0.009‡	0.005	−1.884
Parent Gender X Exper	0.013	0.010	2.616	0.004	0.006	0.818	−0.001	0.007	−0.181
Child Gender	−0.005	0.005	−0.777	−0.003	0.003	−0.693	0.003	0.004	0.681
CFI	0.995			0.983			0.987
TLI	0.992			0.972			0.978
RMSEA	0.035			0.040			0.043
χ2 test of model fit	χ2 (17) = 32.616*			χ2 (17) = 37.784*			χ2 (17) = 40.465*

Note.

Significance at α = .05 level is denoted by *. Marginal significance at p < .10 is denoted by †. A non-significant χ² statistic indicates goodness-of-fit.

Gender was coded as female = 0, male = 1; therefore, positive coefficients indicate that fathers were higher on a latent variable. Parental experience was coded as first-time = 0, experienced = 1; therefore, positive coefficients indicate that experienced parents were higher on a latent variable. Effect size refers to the magnitude of the group difference in standard-deviation units of the dependent variable. With the addition of control variables (parents’ education and income) some effects became non-significant, as discussed in the text.

The next set of results (“Intercept, Linear Effect, and Quadratic Effect regressed onto…”) reports whether the patterns shown in the overall intercept, and linear and quadratic effects varied by group membership. For instance, a significant effect of the “intercept regressed onto parent gender” indicates that mothers and fathers differed on the measure of interest during the second trimester of pregnancy, whereas a significant finding of the “linear effect regressed onto parent gender” indicates that mothers and fathers showed different patterns of change over time. In a sense, this second set of results represents moderators to the overall intercept and linear and quadratic effects, and when significant indicates that the pattern of overall results depends on gender and prior parenting experience. In the third and final portion of Table 3, model fit indices are reported to assess how well the latent growth curve models fit the data to which they were applied.

Changes in Attitudes and Identity

The models analyzed in this section took the form of latent growth curves and were used to examine overall patterns of change and differences by gender, parenting experience, and a gender-by-experience interaction. Child gender was analyzed as a covariate in these analyses to account for the effects of child gender on parental trajectories, but this variable did not exert a significant effect in any model. The growth curves were constructed to test differences in “intercept” (second trimester of pregnancy), as well as linear and quadratic change across time.

Gender-Role Attitudes

First, we analyzed a latent growth curve model to assess changes in gender-role attitudes across the transition to parenthood overall and by gender and parenting experience. Across participants, there was an overall negative curvilinear effect in gender-role attitudes, indicating that individuals became more traditional during the transition to parenthood, although the rate of change lessened over time (linear effect b = −0.389, quadratic b = 0.023, Table 3). As expected, women’s gender-role attitudes were more egalitarian than men’s before and after childbirth (b = −8.855), with women reporting egalitarian attitudes that were more than one standard deviation above that of men’s. Although all groups declined in egalitarian attitudes across time, a large parenting experience effect on quadratic change indicated that this decline lessened over time for first-time parents (b = −0.012); that is, first-time parents appeared to partially “recover” from their changes in attitudes, whereas experienced parents did not (Figure 1).

After controlling for education and income, overall linear and quadratic effects were no longer significant. This is qualified by the fact that the effect of parenting experience on quadratic change remained significant, as did the effect of parent gender on gender-role attitudes during pregnancy. Higher parental education (number of years) was associated with more egalitarian gender-role attitudes before childbirth, b = .69, p < .001, as was income (range: <$5000/year to >$80,000/year), b = 1.20, p < .01.

Family Salience

To assess change over time in family salience, we constructed a latent growth curve model identical to that used to examine gender-role attitudes, simply substituting in the four family salience measurements. As shown in Table 3, overall family salience increased in a curvilinear fashion among participants across time and began to level by 12 months postpartum, as indicated by a significant linear effect and a marginally significant quadratic effect (linear effect b = 0.233, quadratic b = −0.010). A large significant gender difference in the linear effect suggested that mothers increased in family salience at a greater rate than did fathers (b = −0.186). Compared to parents with one previous child, first-time parents reported less family salience during pregnancy (b = 1.214) but a sharper increase in family salience across the transition to parenthood (b = −0.180), as evidenced by significant effects of parenting experience on the intercept and linear effect. Both effects were of large magnitude, which indicated the substantial role of parenting experience on family salience.

After controlling for education and income, several effects were no longer significant, including the overall linear and quadratic effects. The effect of parenting experience on the intercept remained significant, with its effects on the linear and quadratic effects marginally significant. Parent gender was no longer a significant predictor of the linear and quadratic effects. Higher parental education was associated with higher family salience before childbirth, b = .20, p < .05.

Work Salience

Across groups, work salience decreased over time, leveling off several months after the birth of a child, as indicated by significant linear and quadratic effects (linear effect b = −0.541, quadratic b = 0.018, Table 3). Large gender differences in the intercept and linear and quadratic effects indicated that fathers reported a higher average work salience during pregnancy than mothers, and mothers experienced a sharper curvilinear decline than fathers in work salience (b = 2.107, 0.371, and −0.015, respectively). Compared with experienced parents, first-time parents reported higher work salience during pregnancy (b = −1.437), but a steeper decline in work salience following childbirth (b = 0.270), as shown by the large effects of parenting experience on the intercept and linear effect.

After controlling for education and income, the overall intercept and linear and quadratic effects remained significant. Parent gender continued to significantly predict the intercept and linear and quadratic effects. Parenting experience continued to predict the intercept and linear effect, but its marginal quadratic effect was reduced to non-significance. Higher parental education was associated with higher work salience during pregnancy, b = .48, p < .001. In addition, high family income was associated with a smaller decline in work salience following childbirth, linear b = .06, p < .05; quadratic b = −.003, p < .05.

Changes in Division of Labor

The model analyzed in this section took the form of a latent growth curve to examine both overall patterns of change in gender-differentiated division of labor, as well as differences due to parenting experience. Child gender was included as a covariate. The growth curve for division of household labor was constructed to test differences in the “intercept” (second trimester of pregnancy) and in linear and quadratic change across time. Scores were centered around three so that a score of zero represented an equal division of labor. Model fit indices indicated that this model provided a poor fit to the data; therefore, the model could not be interpreted with confidence and is not presented in this paper.

Discussion

The goal of this study was to examine changes in gender-role attitudes, identity salience, and division of labor across the first-time transition to parenthood and following the birth of a second child for experienced mothers and fathers. As predicted, gender-role attitudes became more traditional for all participants from pregnancy to 12 months postpartum; and they changed differently for first-time parents than for experienced parents, such that first-time parents began to become more egalitarian again over time, whereas experiences parents remained more traditional. The findings did not demonstrate a difference between mothers and fathers in attitudinal change over time. Also as expected, family salience increased and work salience decreased for both parents during this time; and identity salience changed more over time for first-time parents than for experienced parents, and more for mothers than for fathers.

Gender-Role Attitudes and Transition to Parenthood

The transition to parenthood and giving birth to subsequent children are major developmental milestones for adults. The finding that parents on average showed changes in gender-role attitudes following the birth of a child aligns with social structural theory, which contends that occupying the role of parent is expected to prompt psychological change, such as attitudinal change, to adapt to this role (Eagly & Wood, 1999). Societal pressures to conform to the role of motherhood for women and the role of provider for men may alter attitudes about parenthood in support of these gender-differentiated roles. When education and income were controlled, gender-role attitudes no longer showed change over time overall. However, a gender difference was still seen in gender-role attitudes during pregnancy, and parenting experience still affected change in attitudes over time. Consistent with other research, couples with more education and a higher income had more egalitarian gender-role attitudes before the birth of a child (Thornton et al., 1983). It should be noted that the addition of control variables increased complexity of the model, and therefore reduced the power to detect effects that were present before adding the controls.

In the current study, gender-role attitudes changed differently for first-time parents than for experienced parents, providing support for the Transition to Parenthood Hypothesis, that attitudinal change over time for parents is due to transitioning to parenthood. However, the results also demonstrated that gender-role attitudes became more traditional over time, especially for experienced parents. Additionally, gender-role attitudes during pregnancy were similar for first-time and experienced parents, and first-time parents appeared to become more egalitarian again after they became more traditional, whereas experienced parents remained more traditional. These patterns may signal that first-time parents return to higher levels of egalitarian attitudes over time, although the design of the current study did not allow us to test this possibility. Nonetheless, the findings that experienced parents’ attitudes became more traditional across time suggest that changes in attitudes postpartum may be related in part to the more direct experience of having a new baby in the house and negotiating the demands associated with this situation, providing support for the New Baby Hypothesis.

A number of processes are likely to account for the new baby effect, where parents become more traditional in their attitudes and behavior following the birth of a child. One factor may be the labor involved in caring for a baby (e.g., breastfeeding), combined with societal attitudes about a woman’s competency in the role of motherhood. When a new mother adeptly cares for her child, these attitudes are confirmed. At the same time, having a new baby is financially demanding, both immediately and for many years following infancy. This may cause the father’s provider role to become more salient within his identity hierarchy, leading to more traditional gender-role attitudes and identity salience.

In short, the results supported both the Transition to Parenthood Hypothesis and the New Baby Hypothesis. Insofar as both first-time and experienced parents showed trends toward more traditional attitudes, the New Baby Hypothesis was supported. Insofar as first-time parents displayed a different pattern of change than experienced parents did, such that first-time parents began to become more egalitarian again over time, whereas experiences parents remained more traditional, the Transition to Parenthood Hypothesis was supported.

Identity Salience and Transition to Parenthood

The differential patterns in family and work salience seen in first-time and experienced parents, provide support for the Transition to Parenthood Hypothesis; greater change was found for first-time parents undergoing a life-changing transition, whereas experienced parents had already done so previously. The results also support the New Baby Hypothesis in that family salience increased and work salience decreased for both first-time and experienced parents following the birth of a child, indicating that some change may be due to having a new baby in the household.

Gender differences in identity salience support previous research suggesting that the role of parenthood is more central to women’s identities than to men’s identities, and that the importance of family to women is shown through fulfillment of the motherhood role, whereas the importance of work and family to men is shown through fulfillment of the provider role (Burke & Cast, 1997; Erickson, 2005; Simon, 1992). According to social structural and identity theories, women who are in the role of mother psychologically adjust to this role through making family more salient and work less salient to their identity. Although family salience also increased for fathers following the birth of a child, it changed to a lesser degree than for mothers. Perhaps this is because the male parental role is defined as breadwinner by society. This might also provide an explanation for why work salience was higher for fathers than for mothers at all time points.

When education and income were controlled, family salience no longer showed change over time, and gender differences were no longer seen in change over time for the sample as a whole. However, parenting experience continued to exert an effect on family salience during pregnancy and across the transition to parenthood. For work salience, gender and parenting experience still had an effect both during pregnancy and over time, even when education and income were controlled. As previously mentioned, the addition of control variables to an already complex model may have reduced the power to defect effects that were formerly seen in the model.

Division of Labor and Transition to Parenthood

The model analyzed to test changes in division of labor over time by both gender and parenting experience did not provide a good fit to the data and thus could not be interpreted with confidence. One reason why the model fit was poor could be related to the items used to test division of household labor. This measure asked parents about division of labor regarding a number of different household tasks. While some of the items were gender-stereotypical items such as taking out the trash for men or cleaning the house for women, many of the items were gender-neutral, such as paying the bills. It is possible that couples divide these tasks between themselves in different ways, but in ways that seem fair to the couple. Therefore, a measure that averages across all tasks does not capture variation between families. An additional statistical problem with the model fit was that the data did not show substantial variance between families or over time in the proportion of household tasks completed by mothers versus fathers. In short, we cannot come to conclusions about division of household labor in the current study.

Limitations

The current study has a number of limitations that should be recognized. The first is that the data were collected in the early 1990s. It is possible that gender-role attitudes, division of labor, and identity salience have changed since that time. However, the study still provides important findings regarding these variables across the transition to parenthood. Regarding the study design, following the same families through a subsequent birth, rather than comparing different families, would allow for a cleaner picture of change over time in addition to controlling for any differences inherent in people who decide to have only one versus two children. This would be a good area for future research to further clarify the findings from the current study.

Another limitation was that the items used to measure division of labor in the current study were self-report. It is possible that participants underestimated or overestimated the relative contribution of each partner to household tasks. More objective measures of division of labor, such as time use measures, could produce different results. However, in the current study there was considerable agreement between parents on each person’s contribution, and the division of labor measure asked about specific behavior regarding household tasks, which showed good validity for the division of household labor measure.

Regarding sampling, the nature of the current study necessitated recruitment of participants who were members of a heterosexual couple living together. The inclusion of both genders allowed for examination of gender-related attitudes and behavior, such as exploring patterns of gender-differentiated division of labor. The experiences of alternative families such as single-parent families, same-sex couples, or non-biological families created through remarriage, re-partnership, or adoption may be different from those of a heterosexual couple following the birth of a child. For example, same-sex couples may negotiate divisions of labor differently, or a single mother might have a more equal work and family salience because she alone occupies the roles of both parent and provider for her family. Thus, there are limits to the extent to which the findings from this study can be generalized beyond heterosexual couples.

Another possible limitation of the study is that the mothers in the sample were older and more stable than the general population, probably because teenage mothers were excluded from the sample. In the United States in 1990, the average age of first-time mothers was 24.2 and the average age of mothers having a second child was 26.9 (Mathews & Hamilton, 2002). In the current study, the average age of first-time mothers was 27.71 and the average age of mothers having a second child was 29.88. All of the couples were living together as required for participation in the study, and had been living together for 5.4 years on average; and 95% of the women were married, therefore conferring relationship stability that may not be characteristic of the general population.

A final limitation of the sample is that it is not diverse in terms of ethnicity, although the sample accurately represents the area from which the families were recruited. Future research could use a more diverse sample to examine whether differences exist in gender-role attitudes, identity salience, and division of labor in parents from different ethnic groups.

Conclusions

Findings in the current study showed changes in gender-role attitudes, identity salience, and division of labor that were linked to each parent’s gender and to the couple’s level of parenting experience. Overall, parents became more traditional in their attitudes and behavior from pregnancy through the first year postpartum. For the most part, the pattern of results demonstrated that mothers changed more than fathers, and that first-time parents changed more than experienced parents, providing support for the Transition to Parenthood Hypothesis. The finding that women change more than men and that first-time parents change more than experienced parents suggests that becoming a parent is a particularly salient experience for women that is associated with significant psychological and behavioral changes.

Although some of the variables changed differentially for each group of parents, both first-time parents and experienced parents showed change over time, providing support for social structural theory such that a change in roles, such as becoming parents, prompts subsequent psychological change to adjust to the role. In focusing on first-time parents only, previous studies attributed changes in parents following the birth of a child to transitioning to parenthood (Transition to Parenthood Hypothesis). Findings from this study suggest that psychological changes in parents from five months pregnant to 12 months postpartum occur for experienced parents as well and therefore may also be due to negotiating the demands of having a new baby in the family, providing support for the New Baby Hypothesis. The inclusion of gender-role attitudes, identity salience, and division of labor, and the findings that these variables change over time, expand our understanding about psychological and behavioral changes that take place when individuals have a child.