Mums the Word! Cross-national
Relationship between Maternal
Employment and Gender Inequalities
at Work and at Home
Kathleen L McGinn
Elizabeth Long Lingo
Working Paper 15-094
Mayra Ruiz Castro
Mums the Word! Cross-national
Relationship between Maternal
Employment and Gender Inequalities at
Work and at Home
Kathleen L McGinn
Harvard Business School
Mayra Ruiz Castro
Harvard Business School
Elizabeth Long Lingo
Worcester Polytechnic Institute
Working Paper 15-094
Copyright © 2015, 2016 by Kathleen L McGinn, Mayra Ruiz Castro, and Elizabeth Long Lingo
Working papers are in draft form. This working paper is distributed for purposes of comment and discussion only. It may
not be reproduced without permission of the copyright holder. Copies of working papers are available from the author.
Mums the Word! Cross-national Relationship between Maternal Employment
andGenderInequalitiesatWorkandatHome
Kathleen L. McGinn, Harvard Business School
Mayra Ruiz Castro, Kingston College, London
Elizabeth Long Lingo, Worcester Polytechnic Institute
ABSTRACT
Our research considers how childhood exposure to non-traditional gender role models at
home, specifically being raised by an employed mother, relates to men’s and women’s
employment and domestic outcomes. Our analyses rely on national level archival data
and individual level survey data collected as part of the International Social Survey
Programme in 2002 and 2012 from nationally representative samples of men and women
in 24 countries. Adult daughters, but not sons, of employed mothers are more likely to be
employed and, if employed, are more likely to hold supervisory responsibility, work
more hours, and earn higher wages than women whose mothers stayed home fulltime. At
home, sons raised by an employed mother spend more time caring for family members
than men whose mothers stayed home fulltime, and daughters raised by an employed
mother spend less time on housework than women whose mothers stayed home fulltime.
The pattern of results in both fixed effects models and mixed models is consistent with
the proposition that employed mothers provide non-traditional gender role models to their
children, liberalizing gender attitudes and transmitting life skills for managing competing
responsibilities, increasing the likelihood of their daughters’ active engagement in the
workplace and their sons’active engagement in family care.
Keywords: maternal employment; work-family; gender
Gender inequality is a barrier to human development across the globe. In the public
sphere, gender inequality manifests in disadvantages for women and girls in health,
political representation and labor market participation (UNDP 2015); in the private
sphere, unequal engagement in parenting disadvantages men, women and their children
(Deutsch, 2001; Fagan & Iglesias, 1999). Gender attitudes—beliefs about appropriate
roles for men and women—both reflect and reinforce gender inequality (Davis and
Greenstein, 2009). Gender attitudes endorsing employment and domestic equality
between men and women are related to greater equality in adults’ participation at work
and at home (Alesina et al. 2013; Davis & Greenstein 2009; Farre and Vella 2013;
Frenandez & Fgli 2006; Fortin 2005; Olivetti, Pattacchini & Zenou 2015; Stickney and
Konrad 2007). Research on the intergenerational transmission of gender attitudes
Mumstheword!
1
provides evidence that parents play an essential role in shaping the gender attitudes their
children hold as adults (Thornton, Alwin & Camburn, 1983). Children raised by mothers
who are employed hold more egalitarian gender attitudes (Fernandez & Fogli 2010) and
may therefore be expected to reap the associated employment and domestic benefits, but
mothers’ employment during their sons’ and daughters’ childhood years remains a
lightening rod for emotional debate and policy discourse (Slaughter, 2015).
The objective of this study is to explore how men’s and women’s involvement in
public and private spheres relates to growing up in a household where the mother was
employed while raising children. Our analyses rely on national level archival data
combined with individual level survey data collected in 24 countries as part of the
International Social Survey Programme (ISSP) in 2002 and 2012. We find that exposure
to maternal employment during childhood is associated with enhanced labor market
outcomes for women and increased domestic involvement for men. These associations
are partially mediated through individually held gender attitudes, and moderated by
gender attitudes at the societal level. Among women, the associations hold primarily for
those with children living at home. Strong main effects remaining after controlling for
individual and country level gender attitudes, and the association with women’s parenting
status, suggest that children’s life management skills as well as their gender attitudes are
shaped by participating in households where mothers juggle demands at work and at
home.
Associations between maternal employment and children’s attitudes and
behaviors
Past research on maternal employment focuses on behavioral and cognitive outcomes
during early childhood (Baydar & Brooks-Gunn 1991; Brooks-Gunn, Han & Waldfogel
2002; Anderson, Butcher & Levine 2003; Belsky & Eggegeen 1991; Berger, Hill &
Waldfogel 2005; Bernal 2008; Brooks-Gunn, Han & Waldfogel 2002). Meta-analyses
find few consistent relationships between maternal employment and these childhood
outcomes (Goldberg, Prause, Lucas-Thompson & Himsel 2008), with the exception of
slight cognitive and behavioral benefits to maternal employment evident in lower income
children and short-term detrimental behavioral effects related to maternal employment in
the child’s first year (Lucas-Thompson, Goldberg & Prause 2010). Later studies suggest
Mumstheword!
2
detrimental effects of first year maternal employment may be limited to non-Hispanic
white children, and, where present, offset by positive indirect effects of increased
attendance in formal child-care settings (Brooks-Gunn, & Waldfogel, 2010). A smaller set of papers considers the association between maternal employment
and adult outcomes. Here, the effects appear more consistently positive. At home, adult
daughters of employed mothers report more equitable division of household work than
their counterparts raised by stay-at-home mothers (Cunningham 2001). Turning to
employment outcomes, daughters of employed mothers spend more hours in paid
employment as young adults (Olivetti et al., 2015). Marriage choices seem to reflect
maternal employment too: sons raised by employed mothers are more likely to be
married to women who are also employed (Fernandez, Olivetti et al., 2004). Maternal employment may affect children’s outcomes as adults through two
mechanisms: by shaping attitudes about what is appropriate and desirable (Fernandez &
Fogli 2010) and by transmitting skills and capacities that children can rely on later in life
(Bandura, 1977; Cunningham, 2001). Parents act as role models, shaping their children’s
attitudes and behaviors (Davis and Wills, 2010; Farréand Vella, 2013; Fernandez, 2004;
Johnston, Schurer and Shields, 2014; Moen, Erickson and Dempster-McClain, 1997;
Risman, 1998). As role models, parents influence their children’s sense of what is
desirable and possible, providing a template for how to operate in the home and in the
labor force.
Traditional gender attitudes—supporting women as homemakers and men as
breadwinners—are associated with substantial reductions in women’s human capital
investment, labor supply, and rates of return to education (Vella, 1994; Corrigall and
Konrad, 2007; Davis and Greenstein, 2009). Using data from 28 countries, Stickney and
Konrad (2007) found that, compared to individuals in their own countries, women with
traditional attitudes had significantly lower earnings than women with egalitarian
attitudes. Similarly, Fortin (2005) found in her investigation of across 25 OECD countries
that traditional views of gender roles had a strong negative association with female
employment rates and earnings. More egalitarian gender attitudes have also been associated with more equitable
division of household labor, including cooking, shopping, cleaning, and care of children
Mumstheword!
3
and parents (Davis and Greenstein, 2009). Researchers have found similar relationships
with domestic outcomes in Australia (Baxter, 1992), England (Kan, 2008), Germany
(Lavee and Katz, 2002), Israel (Lavee and Katz, 2002; Lewin-Epstein, Stier, and Braun,
2006), Sweden (Nordenmark and Nyman, 2003), Taiwan (Hu and Kamo, 2007), and the
United States (Bianchi et al., 2000; Coltrane and Ishii-Kuntz, 1992; Cunningham, 2005;
Greenstein, 1996a; 1996b), as well as in a number of cross-national studies (Batalova and
Cohen, 2002; Davis, 2007; Fuwa, 2004; Nordenmark, 2004). Both men’s and women’s
gender attitudes are important predictors of the division of household labor (Kroska,
2004; Davis and Greenstein, 2009), but women’s gender attitudes appear especially
important in maintaining a more equitable division of household labor after couples
become parents (Schober, 2011). Maternal employment may also teach children life skills. By observing their
parents’ behaviors, children learn skills and build capacities that can be drawn upon as
resources later in life (Bandura, 1977). Analyzing data from a 31-year panel study,
Cunningham (2001) found that parental division of household labor during childhood
was associated with sons' participation in routine housework as adults, while mothers’
employment during their daughters' early years was a more important predictor of adult
daughters’ behavior at home. Sons raised in homes where household labor is shared
among household members appear to learn how to do housework, and daughters whose
mothers held paid employment appear to learn how to manage a household and a job
simultaneously. Cunningham concludes that parental influences are transmitted partially
through the children's gender-role attitudes, but that life skills learned as children may
have important additional behavioral effects, especially for men and household labor.
Attitudes and skills may each play a role in the relationship between maternal
employment and adult children’s outcomes. In the study presented here, we explore
mechanisms after establishing the association between mothers’ employment status and
adult children’s employment and domestic outcomes. First, we test whether individually
held gender attitudes mediate the relationship between maternal employment and adult
outcomes. Second, we consider employment effects for women with and without children
separately; maternally influenced gender attitudes should affect both populations, while
life skills gleaned from first-hand exposure to an employed mother should be more
Mumstheword!
4
critical after having children. Third, we explore interactions between maternal
employment and country-level gender attitudes. Controlling for individual level gender
attitudes, effects that rely on learned skills should be stronger in more traditional societies
where there is little external support for women’s involvement at work and men’s
involvement at home and weaker in egalitarian societies that provide social reinforcement
for gender atypical roles. DataandMethods
Our analyses rely on individual level data from the 2002 and 2012 “Family and Changing
Gender Roles”modules of the International Social Survey Programme (ISSP). The ISSP,
a cross-national collaboration program, designs annual questionnaires across a range of
social science topics.1 Independent organizations in the participating countries collect
ISSP data, either separately or as part of ongoing national surveys, from representative
samples of the country’s adult population. Surveys are conducted primarily through faceto-face interviews and self-completion surveys. The data are documented and made
available by the Central Archive for Empirical Social Research at the University of
Cologne, Germany. ISSP publishes complete documentation of the randomization
procedures, survey protocol, and response rates, by country and year, on their website.2
(See Appendix A for text of ISSP questions used in our analyses).
The “Family and Changing Gender Roles”module of the ISSP focuses on gender
attitudes, women’s employment, marriage and children, as well as household
management and partnership (ISSP Research Group, 2013). The module consists of four
surveys, from 1988, 1994, 2002, and 2012. Because there are relatively few countries in
the first two surveys, we use data from 2002 and 2012 only. Our analyses are based on
data from all countries included in both 2002 and 2012, with the exceptions of Ireland
and Bulgaria, due to missing data on critical variables in those countries. We analyze data
from 24 countries: Australia, Austria, Chile, Czech Republic, Denmark, Finland, France,
Germany, Great Britain, Israel, Japan, Latvia, Mexico, Norway, Philippines, Poland,
Russia, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Taiwan, and the United
1SurveysaredesignedinEnglishandtranslatedintothenationallanguageineachparticipating
country.
2http://www.issp.org
Mumstheword!
5
States. We restricted our sample to working-age respondents, designated as respondents
between 18 and 60 years old. All analyses are run on males and females separately. We
excluded cases with missing data on respondent’s sex. Number of observations differs
across analyses due to missing data on outcome variables. Our dependent variables are measures of employment and domestic engagement.
Measures of employment outcomes include Employed, Supervisory Responsibility,
Hours Worked, and Z-Income. Employment is a dummy variable based on respondents’
weekly hours worked (Employed = 1 if hours worked > 0). Supervisory Responsibility is
a dummy variable reflecting whether respondents reported being directly responsible for
the work of other people (Supervisory Responsibility = 1 if yes). Hours Worked is a
continuous measure (0 to 96 hours/week). To create an income measure that is
comparable cross-nationally, we log transformed annualized earnings and standardized
within each country-year (Z-Income). Our measures of domestic engagement, Hours
Housework and Hours Care, are both continuous measures (0 to 96 hours/week). Hours
Care is based on responses to a question included in the 2012 survey only.
Our primary predictor variable, maternal employment (Mother Employed), is
based on responses to the following question on the ISSP survey: “Did your mother ever
work for pay for as long as one year, after you were born and before you were 14?”
(Mother Employed = 1 when the respondent’s mother worked before the respondent was
14 years old; 0 otherwise). Our measure of gender attitudes reflects respondents’ responses to nine survey
questions regarding women’s employment and gender roles in the household. (See
Appendix for complete list.) Seven of the nine items comprising the Gender Attitude
measure have been used in previous research on gender attitudes (e.g., Crompton and
Lyonette, 2005; Fuwa, 2004; Fuwa and Cohem, 2007; Geist, 2005; Knudsen and
Waerness, 2001; Yodanis, 2005). These include statements such as, “Being a housewife
is just as fulfilling as working for pay,” measured on a five-point scale from strongly
agree (1) to strongly disagree (5). The remaining two items ask whether “women should
work outside the home full time, part time, or not at all” before and after children start
school. Exploratory factor analysis of the nine items using principal-components analysis
suggested a one-factor solution. Factor loading for one item, “Both the man and woman
Mumstheword!
6
should contribute to the household income” was unacceptably low (.35), so we omitted
the item from our scale. Cronbach's alpha in the confirmatory analyses with the
remaining eight items was acceptable (alpha = .78; average inter-item covariance = .39).
We use the standardized scale value as our measure of Gender Attitudes. Higher scores
reflect a more egalitarian gender attitude.
Individual level controls include age, education, marital status, whether or not
there are children living in the household and religion. Because employment outcomes
and engagement in household work are likely to be curvilinear with age, we include both
Age and Age Squared in our analyses. Human capital investments in education influence
outcomes at work and at home (Becker, 1991) and are strongly associated with gender
attitudes (Desai, Chugh & Brief, 2014). We therefore control for respondents’education
in years, using a continuous variable ranging from 1 to 30 years of schooling (Years of
Education).3 We created a dichotomous variable to control for marital status, based on a
categorical response in the ISSP (Married = 1 if married/cohabiting). The presence of
children in the household is associated with movement toward more traditional gender
roles for men and women (Nomaguchi and Milkie, 2003). To control for the presence of
children at home, we transformed responses from two survey questions asking (1) how
many toddlers and (2) how many children from school age to 17 years live in the
household into a dichotomous measure (With Children = 1 if respondent reports any
children living at home; 0 otherwise). To control for religion, shown to affect female
labor supply decisions and patterns of division of household labor within the family
(Lehrer, 1995; 2004), we transformed responses to categorical survey questions on
religious affiliations into a 3-level categorical variable reflecting the two largest response
categories and grouping all other responses: No Religion (22.3%; omitted category);
Christian (88.8%); and Other Religion (11.2%).
3Respondentscouldindicatetheywere“stillinschool”,notingthelevelofschooltheywere
currentlyattending.Valuesthatindicatedrespondentwas“stillatschool”wererecodedas11years
forhighschooland14yearsforcollege,universityandvocationaltraining.
Mumstheword!
7
We estimate effects for Mother Employed using linear probability fixed effects
models.4 Fixed effects models estimate and subtract country-year means from each of the
dependent and explanatory variables, allowing us to directly compare sons and daughters
of employed mothers with those of stay-at-home mothers within each country in a given
year, rather than maternal employment effects between countries. We estimate the
following country-year fixed-effect regressions:
Yic=δMother Employedic+ βXic+ ηc+ εic
where Yic represents adult outcomes—in the workplace or at home—for the ith
respondent in country c;Mother Employedic is a dichotomous variable indicating whether
the respondent’s mother was employed for pay for one year or more between the
respondent’birth and 14th birthday (1 = yes); Xic represent respondent demographics and
family characteristics; ηc denotes country-year fixed-effects capturing factors expected to
differ by country and year, such as GDP, rates of female labor force participation, welfare
policies, and widely-held gender attitudes; εic is the error term. Our fixed-effects models
include robust standard errors clustered at the country-year level. RESULTS
Table 1 provides descriptive statistics for our demographic control variables by
country, by gender. Values vary widely between countries. Aggregating across the 24
countries studied, are more likely to have children living at home and are more likely to
identify as Christian.
------------------Place Table 1 about here-------------------Table 2 presents country by gender averages for each of our dependent variables.
Unsurprisingly, males dominate in employment outcomes and females spend more time
engaged in work at home. Men are significantly more likely than women to be employed
and, if employed, more likely to hold supervisory responsibility, spend more hours on the
job each week and report higher incomes. At home, women spend more hours than men
engaged in household tasks and caring for family members.
------------------Place Table 2 about here-------------------4Weuselinearmodelsforallofouroutcomevariables,includingdichotomousvariables,tosimplify
interpretationofthecoefficients(Angst&Pischke,2008).Inaddition,becauseourmodelsinclude
multipledichotomousandcategoricalvariables,logitmodelsoftenfailtoconverge.
Mumstheword!
8
Turning to maternal employment and gender attitudes (See Table 3), male
respondents are less likely than female respondents to have been raised by a mother who
was employed, and men report significantly more traditional gender attitudes than
women. Across men and women in the 24 countries studied, maternal employment is
positively correlated with more egalitarian gender attitudes in all countries except Latvia.
Figure 1 graphs the mean gender attitudes held by female and male adults raised by a
mother who was employed during the respondent’s childhood and those raised by a stayat-home mother, aggregating across countries.
------------------Place Table 3 and Figure 1 about here-------------------Relationshipbetweenmaternalemploymentandadultdaughters’outcomesat
workandathome
We estimate the direct and gender-attitude-mediated effects of being raised by an
employed mother using step-wise linear probability fixed effects regressions. Table 4
presents the analyses for female respondents. Model 1 shows a significant, positive
coefficient for Mother Employed on women’s Gender Attitudes, establishing the potential
for Gender Attitudes to mediate maternal employment effects in subsequent models.
Models 2 through 15 present the effects of mother’s employment on women’s
employment and domestic outcomes. For each outcome variable, the first model assesses
the strength of the association between maternal employment and the dependent variable;
the second model adds Gender Attitudes to assess mediation.
------------------Place Table 4 about here-------------------Models 2 through 11 reveal that daughters raised by mothers who worked for at
least a year during the daughter’s childhood are significantly more likely to be employed
as adults and, if employed, have a greater likelihood of holding supervisory
responsibility, work more hours weekly, and earn higher incomes. As shown in Model 2,
being raised by an employed mother is associated with an increase of 3.5 percent in the
likelihood of employment. Model 3 reveals that Gender Attitudes partially mediate the
effect of maternal employment on gender attitudes; after controlling for the positive
association between Gender Attitudes and Employed (p < .001), Mother Employed is
associated with an increase of 2.1 percent in daughters’likelihood of employment. Mumstheword!
9
Model 4 presents the effects for Mother Employed on women’s likelihood of
supervising others at work, if employed. Controlling for country-year fixed effects and
individual demographics, 23.5 percent of women raised by employed mothers supervise
others at work, relative to 19.4 percent of women raised by stay-at-home mothers.
Though effects for egalitarian Gender Attitudes are positive and significant in Model 5,
the continued predictive strength of Mother Employed suggests that maternal
employment affects daughters’leadership behavior through some mechanism in addition
to gender attitudes. Model 6 tells us that women raised by employed mothers spend roughly 45
minutes more at their jobs each week than daughters of stay-at-home mothers, and Model
7 reveals that this relationship is only partially mediated by Gender Attitudes. Turning to
earnings, Models 8 and 9 show that daughters of employed mothers earn more annually,
and this effect is fully mediated by Gender Attitudes. We see in Models 10 and 11 that
the relationship between maternal employment and daughters’ annual earnings is partially
due to greater time investment by daughters of employed mothers, but Mother Employed
remains positive and significant in Model 10 after controlling for Hours Worked.
Turning to domestic outcomes, Model 12 reveals that daughters of employed
mothers report spend approximately 45 fewer minutes on housework weekly than
daughters of stay-at-home mothers, controlling for individual demographics and fixed
country-year effects.5 These effects are mediated by Gender Attitudes: egalitarian gender
attitudes are negatively and significantly related to the amount of time women spend
doing housework, and the coefficient for Mother Employed falls below standard levels of
significance in Model 13.InModels 14 and 15, analyzing hours spent caring for family
members weekly, the coefficient for Mother Employed is positive and does not approach
significance in either model, while the coefficient for Gender Attitudes is negative and
significant.
Relationship between maternal employment and adult sons’ outcomes at
workandathome
5Inrobustnesschecks,wereplaceEmployedwithHoursWorkedinanalysesofmen’sandwomen’s
timespentonhouseworkandfamilycare;resultsintermsofdirectionandlevelofsignificance
remainessentiallyunchangedwiththealternatespecificationforemployment.
Mumstheword!
10
Table 5 presents the models for male respondents. Model 16 shows that men, like
women, raised by an employed mother hold significantly more egalitarian gender
attitudes than men raised by mothers who were not employed. The models of men’s
employment and home outcomes, however, show a marked contrast to those for women:
the coefficients for maternal employment are non-significant in all cases where they were
significant in the regressions for women, and significant only for the one dependent
variable—Hours Care—with non-significant effects for women. As expected, Models 17
through 26 reveal no significant associations between Employed Mother and men’s
employment outcomes. Gender Attitudes are significantly related to men’s hours worked;
men with more egalitarian gender attitudes spend fewer hours at work each week. Nor is
maternal employment significantly related to the time adult sons spend on housework
(Models 27 & 28), though egalitarian Gender Attitudes are associated more time spent in
housework. The coefficients for Mother Employed in Models 29 and 30 are positive and
significant: controlling for Gender Attitudes, sons of employed mothers report spending
approximately 50 additional minutes weekly caring for family members, relative to sons
of stay-at-home mothers.
------------------Place Table 5 about here-------------------Mechanisms underlying the relationship between maternal employment and
daughters’ employment outcomes
The results reported above suggest that the relationship between Mother
Employed and employment and domestic outcomes is only partially explained by the
effects of maternal employment on children’s beliefs about appropriate roles for men and
women. The effects of maternal employment on daughters’ income are fully mediated
through gender attitudes. But the likelihood of daughters’ employment and hours women
spend at work are only partially mediated, and Gender Attitudes do not mediate the
relationships between mother’s employment and daughters’ supervisory status or sons’
engagement in family care.
------------------Place Table 6 about here-------------------To explore the possibility that maternal employment may also be tied to adult
children’s outcomes through the transmission of life skills, we separated our observations
of women’s employment outcomes based on whether respondents reported that there
Mumstheword!
11
were children living in the household.6 If growing up in a home with an employed mother
teaches daughters skills useful for balancing the responsibilities of parenting with
responsibilities in the workplace, these skills would come into play when those daughters
become mothers themselves. Table 6 presents separate models for women with and
without children, controlling for individual demographics and country-year fixed effects.
Models 31 and 32 show that maternal employment is associated with more egalitarian
gender attitudes for women, regardless of whether they have children at home. The rest of
the models in Table 6 control for Gender Attitudes. Models 33 and 34 replicate the
previous models showing that maternal employment is only marginally related to the
likelihood of women’s employment after controlling for Gender Attitudes, and this holds
regardless of children at home. In Models 35 and 36, we see significant coefficients for
Employed Mother in both regressions, but the effects on Supervisory Responsibility are
marginally stronger for women with children (F(1,47)=3.52,p=.07). The pattern shown
in the remainder of the models in Table 6 is notable: for each outcome, the coefficient for
Mother Employed is significant in the model for women with children and nonsignificant in the model for women without children. Relative to the models for women
without children at home, the models for women with children at home have significantly
larger coefficients for Maternal Employment in regressions on hours worked (F(1,47) =
13.34,p<.001) and z-income after controlling for hours worked (F(1,47)=5.90,p=.02),
even after controlling for Gender Attitudes. These findings suggest that employed
mothers are affecting their daughters’ skills as well as attitudes. Daughters of employed
mothers, when faced with the opportunities and challenges of having children
themselves, appear both willing and able to emulate their mothers, managing
employment and caregiving roles simultaneously.
To further explore mechanisms, we examined the interaction between maternal
employment and broadly held gender attitudes within each country studied on each of the
work and home outcomes with significant coefficients for Mother Employed in the fixed
effects models reported above. We test for interactions between maternal employment
6Theeffectsizeformaternalemploymentonsons’hoursspentinfamilycareissignificantlylarger
formenwithchildrenlivingathomerelativetothosewithout,butthisisunsurprisinggiventhe
muchgreaterdemandforfamilycarewhentherearechildreninthehousehold.
Mumstheword!
12
and country-level gender attitudes using linear mixed-effects models including random
intercepts for individual countries and random slopes for the effect of gender attitudes at
the country level. Our mixed models allow country-specific slopes (i.e. random effects)
for gender attitudes, formally acknowledging the aggregated effects of individual
attitudes within a society (Fortrin, 2005; Wooldridge, 2003). Our mixed-effects models
include robust standard errors clustered at the country level. We controlled for
differences in maternal employment opportunities across countries with a measure of
Female Labor Force Participation Rates, by observation year (World Bank, 2014). None
of the mixed models without the interaction term generated a coefficient for Mother
Employed that was meaningfully different in magnitude or significance from the
coefficients generated by the linear probability models with country-year fixed effects
and controls for Gender Attitudes, providing a robustness check for the results reported
above. The findings from the mixed models reveal that maternal employment is more
closely related to women’s and men’s gender attitudes in countries where mothers’
involvement in the workplace is broadly supported. In spite of this, the mixed models
show significantly higher association between maternal employment and employment
and domestic outcomes in countries with more traditional gender attitudes. With the
exceptions of women’s Supervisory Responsibility and Hours spent on Housework
Weekly, which show no moderation by country level gender attitudes, the mixed models
show significant, negative coefficients for the interaction between Maternal Employment
and Mean Gender Attitudes at the country level. Figure 2 shows the marginal effects for
maternal employment by mean level of within-country Gender Attitudes for each of the
work and home outcomes significantly related to maternal employment in our fixed
effects models. Controlling for individual Gender Attitudes, the relationship between
maternal employment and daughters’ work outcomes, and again between maternal
employment and sons’ engagement in family care, is strongest in countries with more
traditional gender attitudes and weak or absent in countries with more egalitarian gender
attitudes. As seen in Figure 2, the relationship between maternal employment and adult
children’s is strongest in countries where women’s involvement in the workplace and
men’s involvment at home is less likely to be viewed as “normal.” The moderating effect
of societally held gender attitudes on employment and domestic outcomes, controlling for
Mumstheword!
13
individually held gender attitudes, suggests that being raised by an employed mother
conveys to children a set of skills and capacities for taking on roles not broadly
reinforced by others in the society around them.
------------------Place Figure 2 about here-------------------AlternativeExplanationsandRobustnessChecks
The pattern of results in both fixed effects models and mixed models is consistent
with the proposition that employed mothers provide non-traditional gender role models to
their children, liberalizing gender attitudes and transmitting life skills for managing
competing responsibilities, thereby increasing the likelihood of their daughters’ active
engagement in the workplace and their sons’ active engagement in family care. But
several alternative explanations warrant consideration. A positive statistical association
between a mother’s employment status and her children’s employment outcomes may
indicate that maternal employment is a proxy for childhood homes with more resources,
more educated parents, more emphasis on work and discipline, etc. Other than our
maternal employment predictor variable, our models have no controls for resources or
stimuli in childhood homes. If beneficial resources or stimuli associated with maternal
employment, but not inherently related to non-traditional gender role modeling, were
driving our effects, adult children of employed mothers would, on average, have
employment outcomes superior to their peers raised by stay-at-home mothers. Our
findings are in stark contrast to this supposition: a mother’s employment status appears to
have no statistical association with her sons’ employment status as adults. The lack of
association between maternal employment and sons’ employment outcomes juxtaposed
against the consistent and positive association with daughters’ employment outcomes
does not rule out the possibility that our findings reflect generic differences between
homes with employed mothers and homes with stay-at-home mothers, but it suggests that
any such differences cannot be gender neutral.
One gendered account of maternal employment as a proxy for household
differences could be that employed mothers reflect households in which women,
including daughters, are favored overall and men, including sons, suffer (Stucky et al.,
1987). If so, women raised by employed mothers may simply fare better across life and
men raised by employed mothers may simply fare worse across a spectrum of outcomes
Mumstheword!
14
than their peers raised by stay-at-home mothers. Our data offers little support for this
conjecture. Both sons and daughters raised by employed mothers have significantly more
years of education than children of stay-at-home mothers (Males: X=12.99(.03) v
12.00(.05); Females: 13.08(.03) v 11.97(.04); 1-sided t-test, both p<.001). In linear
regressions on Years of Education controlling for individual demographics discussed
above and country-year fixed effects, the coefficients for maternal employment are
positive and significant and do not differ significantly between males and females.
Providing further evidence against the “employed moms are good for daughters and bad
for sons” proposition, our findings show that men raised by employed mothers spend
more time caring for family members than men raised by stay at home mothers, while not
differing significantly in any of the employment outcomes we investigated. Past research
has found that men consider their relationships with their children as better markers of
success than their employment related outcomes (Davis & Greenstein 2009; Coltrane
1998, Gerson 1993, Hochschild & Machung 1989). We also analyzed the association
between maternal employment and self-reported overall happiness. The ISSP included a
question asking respondents, on a 7-point scale (1 = “completely happy”; 7 =
“completely unhappy”), “If you were to consider your life in general, how happy or
unhappy would you say you are, on the whole?” The coefficients for Mother Employed
do not approach significance for men or women in country-year fixed effects regressions
of overall happiness that control for individual demographics andincome. The argument
that daughters benefit and sons suffer when raised in homes with an employed mother is,
we conclude, not a convincing explanation for our findings.
One promising alternative explanation for the effect on sons’ involvement with
family care draws from past research on the effects of maternal employment on sons’
spouses’ employment. Fernandez et al. (2004) offer a convincing model and empirical
support across three US-only data sets, concluding that sons raised by mothers employed
outside the home are more likely to be married to women who work outside the home. If
this is the case across the 24 countries we study, our findings for men’s involvement in
caring for family members may be due to their wives’—rather than their mothers’—
employment. The ISSP in 2002 and 2012 included a question asking the number of hours
the respondent’s spouse/partner worked per week. In linear regressions controlling
Mumstheword!
15
individual demographics, Gender Attitudes and country-year fixed effects, Mother
Employed is a strong predictor of sons’ (but not daughters’) spouses’ likelihood of
employment (p < .001).7 We therefore reran the analyses on men’s work and home
outcomes, controlling for Spouse Employed. After controlling for Spouse Employed,
maternal employment is still a significant predictor of men’s involvement in family care
(β=.89; p=.03). In all other analyses of men’s outcomes, the coefficients for Mother
Employed remains at essentially the same levels of non-significance as those reported for
regressions without spousal employment controls.
A possible alternative explanation for the association between maternal
employment and adult daughters’ employment outcomes is that maternal employment
may simply be a proxy for the local availability of employment opportunities for women,
a feature of the place and era in which children were raised (see Goldin and Olivetti,
2013). Our measure of maternal employment could reflect the reality that women (as well
as men) are more likely to be employed in urban settings, and adult offspring, who tend
to live close to the location in which they were raised (Leopold, Geissler & Pink 2012),
are also more likely to be employed in urban settings. If so, our findings—at least those
for daughters’employment outcomes—may reflect similarities in job availability due to
mothers’ and daughters’ colocation, rather than role modeling. To test this possibility, we
reran our fixed-effects analyses on the subset of observations in our sample where
surveys included questions about respondents’ communities,8 adding a dummy variable
set to 1 if the respondent lived in an urban or suburban community (Urban). The effects
for Urban are significant in a number of the models, but the effects for Mother Employed,
and the partial mediation of those effects through Gender Attitudes, remain essentially
unchanged from those in the main analyses reported above. While endogeneity threats are inherent in cross-sectional survey data and we
cannot rule out alternative explanations, the consistent association between maternal
employment and daughters’ employment outcomes and the lack of association with men’s
7TheseeffectsarerobusttoalternativespecificationsforSpouseEmployed:=>0hours/week;=>5
hours/week;=>10hours/week.
8Surveyquestionsaboutcommunitywerenotaskedinthe2002ISSPsurveysinIsrael,Germany,
PolandandRussia.
Mumstheword!
16
employment outcomes, the magnification of influence for women with children at home,
the interaction with broadly held gender attitudes, and the stability of the effects in
additional tests introducing potential confounds mitigate concerns of omitted variables
driving our results. We conclude that our findings provide robust evidence that employed
mothers provide non-traditional gender role models affecting their sons’ and daughters’
gender attitudes and life skills, ultimately influencing their adult children’s outcomes at
work and at home. DISCUSSIONANDCONCLUSION
Our findings shed light on the relationship between maternal employment and
children’s employment and domestic outcomes as adults. Analyzing survey data from 24
countries in 2002 and 2012, we find that adult daughters of employed mothers are more
likely to be employed than adult daughters of mothers who stayed home full time when
their children were young. When employed, adult daughters of employed mothers work
more hours, are better compensated, and are more likely to hold supervisory positions
than daughters of stay-at-home mothers. At home, adult daughters of employed mothers
do fewer hours of housework each week. Maternal employment has no significant
association with the time women spent caring for family members, controlling for
employment or hours spent on the job. For sons, we see the opposite pattern: adult sons’
employment outcomes and housekeeping roles are essentially unassociated with maternal
employment, but adult sons of employed mothers spend more time caring for family
members than adult sons of stay-at-home mothers. The pattern of findings across women’s and men’s work and home outcomes in
24 countries supports the proposition that employed mothers provide non-traditional
gender role models for their children. Our exploration of mechanisms suggests that
having a non-traditional role model—being raised by an employed mother—shapes
adult outcomes through two mechanisms. The first is an influence on gender attitudes, or
beliefs about behaviors that are “right” and “normal” for men and women. We see
evidence of this in our mediation analyses: adult children of employed mothers have
significantly more egalitarian gender attitudes than adult children of mothers who stayed
home full time; in turn, gender attitudes partially or fully mediate the relationships
Mumstheword!
17
between maternal employment and adults daughters’hours worked, earnings, and hours
spent on household work each week. Yet the relationship between maternal employment and gender attitudes can only
partially account for the effects of maternal employment revealed in our analysis. The
relationship between mothers’employment and daughters’ employment and likelihood of
supervising others at work, as well as the hours sons spend engaged in caring for family
members, remains strongly significant after gender attitudes are included in the
regressions. These findings suggest that in addition to holding more egalitarian gender
attitudes, children raised by employed mothers may learn a set of skills that enable
greater participation at work and at home. The children of working mothers observe the
decisions and behaviors of their parents, learning skills and capacities that they can draw
upon as resources as they navigate gendered situations and decisions later in life
(Bandura, 1977; Cunningham, 2001). We speculate, and look forward to future research
for further exploration, that mothers who are employed may be passing information to
their daughters about important skills for exercising power and navigating career systems
outside the home, and to their sons about the skills needed for greater participation in
caring for families and homes. Our work contributes to a growing body of research exploring the effects of
maternal employment on their children’s well-being. We extend this demonstration to the
long-term impact on adult sonsanddaughters.Taken together, our findings provide an
important counterpoint to persistent beliefs and rhetoric that employed mothers are
“abandoning their children” and negatively affecting their families and society over the
long term. We find that being raised by a mother who works outside the home has no
effects on adult daughters’ or sons’ self-reported happiness. But positive associations
abound at work and at home. Adult daughters of employed mothers benefit in the
workplace relative to adult daughters of stay-at-home mothers, while spending less time
on housework and roughly the same amount of time caring for family members (even
controlling for the extra hours the adult daughters of employed mothers spend in paid
employment). Further, we see that adult sons of employed mothers spend more time
caring for family members than adult sons of stay-at-home mothers. Our research reinforces calls for national and local policies supporting parental
Mumstheword!
18
employment. Our findings suggest that policy should focus on supporting mothers who
work—part time or full time. Providing quality and reasonably-priced child care is an
important factor, but policy makers should also address workplace policies that hinder or
assist parental employment. Such policies can range from addressing the culture of
excessive work hours that drives parents—both men and women—out of the workplace
(Cha, 2010; Reid and Ely, 2015), to workplace practices that allow more women to
pursue their career aspirations (Gerson, 2011; Ramarajan, McGinn & Kolb, 2015). Future work on non-traditional gender role models and gender inequality at work
and at home could build upon our research in several ways. First, more in-depth data
collection and analysis of the actual division of labor, discourse, and negotiations among
parents and children over time is needed. For example, drawing from an in-depth
interview-based panel study, Cunningham (2001) demonstrated that sons’time spent on
household tasks as adults is associated with having a father who was more engaged at
home. Our data do not allow us to disentangle whether men’s increased care work is
driven by observation and modeling of fathers’contributions in homes where mothers are
employed (Davis and Wills, 2010). Future research could also build on analyses that have
focused on individuals’ gender attitudes as they consider or transition into parenthood
(Bass, 2014; Schober, 2013), or how couples divide the work of household management
(Treas and Tsui-o Tai, 2012). As the number of hours spent on domestic work decreases
globally, we need a better understanding of the dynamic and fluid nature of work
conducted by all family members and the long-term impact on work outcomes within and
outside the home. Finally, future research on employed mothers as role models should
also consider the larger cultural, social, and economic contexts in which gender is
negotiated and enacted in practice. This may include family and friend networks
(Olivetti, Patacchini, and Zenou 2013) or differences across countries in gender attitudes
or social welfare policies (e.g., Bittman et. al., 2003; Batalova and Cohen, 2002; Fuwa,
2004; Hook, 2006). Over the last twenty years, there have been many studies exploring the effects of
employed mothers on their children’s well-being. The consistent takeaway across these
studies is that young children of employed mothers are higher achieving and have fewer
behavioral problems than young children whose mothers are not employed, and that these
Mumstheword!
19
effects are strongest for children from low income families (Lucas-Thompson, Goldberg
and Prause, 2010). Work by economists has shown positive effects of maternal
employment on women’s work hours (Olivetti et al., 2015) and on sons’ support of
wives’ employment (Fernandez et al., 2004). But negative stereotypes persist. Our
findings add new challenges to stereotypes about employed mothers “harming” their
children. We hope the findings from our research will promote respect for the spectrum of
choices women and men make at home and at work. Whether Moms or Dads stay at
home or are employed, part time or full time, our findings suggest sons and daughters
benefit from exposure to role models offering a wide set of domestic and occupational
alternatives for leading rich and rewarding lives. Giving children opportunities to see and
know people—both men and women—making choices that defy gender stereotypes at
work and at home will help children see a wide set of options for success in their own
lives.
Mumstheword!
20
REFERENCES
Angrist, J. D., & Pischke, J. S. (2008). Mostlyharmlesseconometrics:Anempiricist's
companion. Princeton University Press.
Bandura,A.(1977).SocialLearningTheory.EnglewoodCliffs,NJ:PrenticeHall.
Basow,S.A.,&Howe,K.G.(1979).ModelInfluenceonCareerChoicesofCollege
Students.VocationalGuidanceQuarterly,27(3),239-243.
Basow,S.A.,&Howe,K.G.(1980).Role-ModelInfluence:EffectsofSexandSex-Role
AttitudeinCollegeStudents.PsychologyofWomenQuarterly,4(4),558-572.
Bass,B.C.(2014).PreparingforParenthood?:Gender,Aspirations,andthe
ReproductionofLaborMarketInequality.Gender&Society.
Batalova,J.A.,&Cohen,P.N.(2002).PremaritalCohabitationandHousework:
CouplesinCross-NationalPerspective.JournalofMarriageandFamily,64(3),
743-755.
Beaman,L.,Duflo,E.,Pande,R.,&Topalova,P.(2012).Femaleleadershipraises
aspirationsandeducationalattainmentforgirls:Apolicyexperimentin
India.science,335(6068),582-586.
Becker,G.S.(1991).ATreatiseontheFamily.Cambridge,MA:HarvardUniversity
Press.
Bianchi,S.M.,Milkie,M.A.,Sayer,L.C.,&Robinson,J.P.(2000).IsAnyoneDoingthe
Housework?TrendsintheGenderDivisionofHouseholdLabor.SocialForces,
79(1),191-228.
Bittman,M.,England,P.,Sayer,L.,Folbre,N.,&Matheson,G.(2003).WhenDoes
GenderTrumpMoney?BargainingandTimeinHouseholdWork1.American
JournalofSociology,109(1),186-214.
Brooks-Gunn,J.,Han,W.-J.,&Waldfogel,J.(2010).First-YearMaternalEmployment
andChildDevelopmentintheFirstSevenYears.MonographsoftheSociety
forResearchinChildDevelopment,75(2),7–9.
Chang,M.L.(2000).TheEvolutionofSexSegregationRegimes.AmericanJournalof
Sociology,105(6),1658-1701.
Coltrane,S.,&Ishii-Kuntz,M.(1992).Men'sHousework:ALifeCoursePerspective.
JournalofMarriageandFamily,54(1),43-57.
Corrigall,E.A.,&Konrad,A.M.(2007).Genderroleattitudesandcareers:A
longitudinalstudy.SexRoles,56(11-12),847-855.
Croft,A.,Schmader,T.,&Block,K.(2015).AnUnderexaminedInequality:Cultural
andPsychologicalBarrierstoMen'sEngagementWithCommunalRoles.Pers
SocPsycholRev.
Mumstheword!
21
Crompton,R.,&Lyonette,C.(2005).Thenewgenderessentialism–domesticand
family‘choices’andtheirrelationtoattitudes.TheBritishjournalofsociology,
56(4),601-620.
Cunningham,M.(2001).ParentalInfluencesontheGenderedDivisionof
Housework.AmericanSociologicalReview,66(2),184-203.
Cunningham,M.(2005).GenderinCohabitationandMarriage:TheInfluenceof
GenderIdeologyonHouseworkAllocationOvertheLifeCourse.Journalof
FamilyIssues,26(8),1037-1061.
Cunningham,M.(2008).Influencesofgenderideologyandhouseworkallocationon
women’semploymentoverthelifecourse.SocialScienceResearch,37(1),
254-267.
Davis,S.N.(2007).Genderideologyconstructionfromadolescencetoyoung
adulthood.SocialScienceResearch,36(3),1021-1041.
Davis,S.N.,&Greenstein,T.N.(2009).Genderideology:Components,predictors,
andconsequences.AnnualReviewofSociology,35,87-105.
Davis,S.N.,&Wills,J.B.(2010).Adolescentgenderideologysocialization:Direct
andmoderatingeffectsoffather'sbeliefs.SociologicalSpectrum,30(5),580604.
Deutsch,F.M.(2001).Equallysharedparenting.Currentdirectionsinpsychological
science,10(1),25-28.
England,P.(2005).Genderinequalityinlabormarkets:Theroleofmotherhoodand
segregation.SocialPolitics:InternationalStudiesinGender,State&Society,
12(2),264-288.
Fan,P.L.,&Marini,M.M.(2000).InfluencesonGender-RoleAttitudesduringthe
TransitiontoAdulthood.SocialScienceResearch,29(2),258-283.
Farré,L.,&Vella,F.(2013).Theintergenerationaltransmissionofgenderrole
attitudesanditsimplicationsforfemalelabourforceparticipation.
Economica,80(318),219-247.
Fagan,J.,&Iglesias,A.(1999).Fatherinvolvementprogrameffectsonfathers,father
figures,andtheirHeadStartchildren:Aquasi-experimentalstudy.Early
ChildhoodResearchQuarterly,14(2),243-269.
Fernández,R.,Fogli,A.,&Olivetti,C.(2004).MothersandSons:Preference
FormationandFemaleLaborForceDynamics.TheQuarterlyJournalof
Economics,119(4),1249-1299.
Fortin,N.M.(2005).GenderRoleAttitudesandtheLabour-marketOutcomesof
WomenacrossOECDCountries.OxfordReviewofEconomicPolicy,21(3),416438.
Fuwa,M.(2004).Macro-levelGenderInequalityandtheDivisionofHousehold
Laborin22Countries.AmericanSociologicalReview,69(6),751-767.
Mumstheword!
22
Fuwa,M.,&Cohen,P.N.(2007).Houseworkandsocialpolicy.SocialScience
Research,36(2),512-530.
Geist,C.(2005).TheWelfareStateandtheHome:RegimeDifferencesinthe
DomesticDivisionofLabour.EuropeanSociologicalReview,21(1),23-41.
Gerson,K.(2011).TheUnfinishedRevolution:ComingofAgeinaNewEraofGender,
Work,andFamily.NewYork:OxfordUniversityPress.
Goldin,C.,&Olivetti,C.(2013).ShockingLaborSupply:AReassessmentoftheRole
ofWorldWarIIonWomen’sLaborSupply.AmericanEconomicReview,
103(3),257-262.
Greenstein,T.N.(1996a).GenderIdeologyandPerceptionsoftheFairnessofthe
DivisionofHouseholdLabor:EffectsonMaritalQuality.SocialForces,74(3),
1029-1042.
Greenstein,T.N.(1996b).Husbands'ParticipationinDomesticLabor:Interactive
EffectsofWives'andHusbands'GenderIdeologies.JournalofMarriageand
Family,58(3),585-595.
Gwartney, James, Robert Lawson, and Joshua Hall. (2013). “2013 Economic Freedom
Dataset,”published in EconomicFreedomoftheWorld:2013AnnualReport,
Fraser Institute, http://www.freetheworld.com/datasets_efw.html
Hook,J.L..(2010).GenderInequalityintheWelfareState:SexSegregationin
Housework,1965–2003.AmericanJournalofSociology,115(5),1480-1523.
Hu,C.-Y.,&Kamo,Y.(2007).TheDivisionofHouseholdLaborinTaiwan.Journalof
ComparativeFamilyStudies,38(1),105-124.
ISSPResearchGroup(2013):InternationalSocialSurveyProgramme:Familyand
ChangingRolesIII–ISSP2002.GESISDataArchive,Cologne.ZA3880Data
fileVersion1.1.0,
Johnston,D.W.,Schurer,S.,&Shields,M.A.(2014).Maternalgenderroleattitudes,
humancapitalinvestment,andlaboursupplyofsonsanddaughters.Oxford
EconomicPapers,66(3),631-659.
Knudsen,K.,&Wærness,K.(2008).NationalContextandSpouses’Houseworkin34
Countries.EuropeanSociologicalReview,24(1),97-113.
Kroska,A.(2004).DivisionsofDomesticWork:RevisingandExpandingthe
TheoreticalExplanations.JournalofFamilyIssues,25(7),890-922.
Lavee,Y.,&Katz,R.(2002).DivisonofLabor,PerceivedFairness,andMarital
Quality:TheEffectofGenderIdeology.JournalofMarriageandFamily,64(1),
27-39.
Lehrer,E.L.(1995).TheEffectsofReligionontheLaborSupplyofMarriedWomen.
SocialScienceResearch,24(3),281-301.
Mumstheword!
23
Lehrer,E.L.(2004).ReligionasaDeterminantofEconomicandDemographic
BehaviorintheUnitedStates.PopulationandDevelopmentReview,30(4),
707-726.doi:10.1111/j.1728-4457.2004.00038.x
Lewin-Epstein,N.,Stier,H.,&Braun,M.(2006).TheDivisionofHouseholdLaborin
GermanyandIsrael.JournalofMarriageandFamily,68(5),1147-1164.
Lucas-Thompson,R.G.,Goldberg,W.A.,&Prause,J.(2010).Maternalworkearlyin
thelivesofchildrenanditsdistalassociationswithachievementand
behaviorproblems:ameta-analysis.Psychologicalbulletin,136(6),915.
ManYeeKan.(2008).Doesgendertrumpmoney?Houseworkhoursofhusbands
andwivesinBritain.Work,Employment&Society,22(1),45-66.
Miller,A.,&Sassler,S.(2010).StabilityandChangeintheDivisionofLaboramong
CohabitingCouples.SociologicalForum,25(4),677-702.
Moen,P.,Erickson,M.A.,&Dempster-McClain,D.(1997).Theirmother'sdaughters?
Theintergenerationaltransmissionofgenderattitudesinaworldof
changingroles.JournalofMarriageandtheFamily,281-293.
Nomaguchi,K.M.,&Milkie,M.A.(2003).CostsandRewardsofChildren:TheEffects
ofBecomingaParentonAdults'Lives.JournalofMarriageandFamily,65(2),
356-374.
Nordenmark,M.(2004).Doesgenderideologyexplaindifferencesbetween
countriesregardingtheinvolvementofwomenandofmeninpaidand
unpaidwork?InternationalJournalofSocialWelfare,13(3),233-243.
Nordenmark,M.,&Nyman,C.(2003).FairorUnfair?PerceivedFairnessof
HouseholdDivisionofLabourandGenderEqualityamongWomenandMen:
TheSwedishCase.EuropeanJournalofWomen'sStudies,10(2),181-209.
Olivetti,C.,Patacchini,E.,&Zenou,Y.(2013).Mothers,FriendsandGenderIdentity.
NBERWorkingPaperNo.19610.
Pew Social Trends. (2014). http://www.pewsocialtrends.org/2014/04/08/chapter-4public-views-on-staying-at-home-vs-working/. Accessed 27 May 2015.
Risman,B.(1998).GenderVertigo.NewHaven,CT:YaleUniversityPress.
Sassler,S.,&Miller,A.J.(2010).WaitingtoBeAsked:Gender,Power,and
RelationshipProgressionAmongCohabitingCouples.JournalofFamilyIssues.
Schober,P.S.(2011).TheParenthoodEffectonGenderInequality:Explainingthe
ChangeinPaidandDomesticWorkWhenBritishCouplesBecomeParents.
EuropeanSociologicalReview.
Stickney,L.T.,&Konrad,A.M.(2007).Gender-roleattitudesandearnings:A
multinationalstudyofmarriedwomenandmen.SexRoles,57(11-12),801811.
Mumstheword!
24
Stuckey,M.F.,McGhee,P.E.,&Bell,N.J.(1982).Parent–childinteraction:The
influenceofmaternalemployment.DevelopmentalPsychology,18(4),635644.
Thornton,A.,Alwin,D.F.&Camburn,D.(1983).CausesandConsequencesofSexRoleAttitudesandAttitudeChange.AmericanSociologicalReview,48,211227.
Treas,J.,&Tai,T.-o.(2011).HowCouplesManagetheHousehold:WorkandPower
inCross-NationalPerspective.JournalofFamilyIssues.
UnitedNationsDevelopmentProgram(UNDP).2010.HumanDevelopmentReport.
NewYork:OxfordUniversityPress.
Vella,F.(1994).GenderRolesandHumanCapitalInvestment:TheRelationship
betweenTraditionalAttitudesandFemaleLabourMarketPerformance.
Economica,61(242),191-211.
Wooldridge,J.M.(2003).Cluster-SampleMethodsinAppliedEconometrics.The
AmericanEconomicReview,93(2),133-138.
World Bank. (2009). Data retrieved February 26, 2014, from World Development
IndicatorsOnline(WDI)database.
Mumstheword!
25
Mumstheword!
26
Mumstheword!
27
Mumstheword!
28
Mumstheword!
29
Mumstheword!
30
Mumstheword!
31
Mumstheword!
32
f)
Appendix A
ISSP Survey Questions1
Survey Questions Used in Creating Measures for Control Variables
Age
Age of respondent (in years)
Years of Education
How many full years of schooling or education have you had? Please include primary
and secondary schooling, university and full-time vocational training, but do not
include repeated years.
Marital Status
What is your current legal marital status?
1=Married, or living as married; 2=Widowed; 3=Divorced; 4=Separated, after being
married; 5=Never married, single, not married
Children Living in the Household
How many children up to the age of school age live in your household?
How many children between school age and 17 years old live in your household?
Religion
Groups of religious affiliations
Do you belong to a religion and, if yes, which religion do you belong to?
Recoded:
0=No Religion; 1= Christian; 2=Jewish; 3=Islamic; 4= Buddhist; 5=Hindu; 6=Other
A man’s job is to earn money, a woman’s job is to look after the home and
family
1=strongly agree; 5=strongly disagree
What do you think is the best arrangement for women's work outside the home under
the following circumstances?
g) When there is a child under school age.
h) After the youngest child starts school.
1=stay home; 2=part-time; 3=full-time
Survey Questions Used in Creating Measures for Dependent Variables
Employed
Last week were you working full time, part time, going to school, keeping house, or
what?
1 Currently in paid work; 2 Currently not in paid work, paid work in the past; 3 Never had
paid work; 9 No answer
Hours Housework
How many hours spend on household work?
0 None, no hours, does not apply; 1 1 hour or 95 less than 1 hour; 2 2 hours; 3 3 hours,
CN: 3 hours or more; 95 hours and more; 98 Don't know, BG: can't choose; 99; No answer
Hours Care
On average, how many hours a week do you spend looking after family members (e.g.
children, elderly, ill or disabled family members)?
0 None, no hours, does not apply; 1 1 hour or 95 less than 1 hour; 2 2 hours; 3 3 hours,
CN: 3 hours or more; 95 hours and more; 98 Don't know, BG: can't choose; 99; No answer
Predictor Variables
Mother Employed
Did your mother ever work for pay for as long as one year, after you were born and
before you were 14?
1=Yes, she worked for pay; 2=No
Gender attitudes
To what extent do you agree or disagree...?
a) A working mother can establish just as warm and secure a relationship with
her children as a mother who does not work;
b) A pre-school child is likely to suffer if his or her mother works;
c) Family life suffers if a woman goes out to work;
d) Work is alright, but what a woman really wants is a home and family;
e) Being a housewife is just as fulfilling as working for pay
Mumstheword!
Supervisory Responsibility
In your main job, do you supervise anyone or are you directly responsible for the work
of other people?
1=Yes, supervise others at work; 2=No, do not supervise
Hours Worked
How many hours, on average, do you usually work for pay in a normal week,
including overtime?
Z-Income
Before taxes and other deductions, what on average is your own total monthly
income?
Country specific personal income (annualized, logged, and standardized
1
Questions phrased slightly differently across languages.
33