Differences in German Youth Gender Ideologies.

Differences in German Youth Gender Ideologies: The Relationship Between

Family Structure and Doing Gender

In recent years, there has been rapid change in family formation and romantic partnering

(Mather 2010). Furthermore, this rapid change has been associated with shifting gender

ideologies among adults (Mason and Lu 1988). This shift in gender ideologies can vary in

intensity and direction, and can have a significant impact on societal and individual level

outcomes (John, Shelton, and Luschen 1995; Christie-Mizell et al. 2007). Much of past literature

focuses on how family structure affects how adults do gender with their partners, not how they

do gender with their children. Even with shifts in gender ideologies and family formation among

adults, little research has studied how these changes in doing gender behaviors may affect

youths’ gender ideologies. This is especially important because youth may be more vulnerable to

these shifts since they spend more time with their parents than older individuals (Kiecolt and

Acock 1988).

With my research, I aim to uncover whether family structure influences youth gender

ideologies. Family structure can affect how parents model behaviors and attitudes at home, in

turn influencing their offspring’s gender ideology. In order to look at youth gender ideology

differences, I analyze German youth across an array of family structures and ages. I predict

youth in single-parent families will have the most egalitarian gender ideologies, and youth with

stepparents or social parents will be more egalitarian than those youth with two-parent biological

married parents. Furthermore, I predict that girls will be more sensitive to the potential effects of

family structure on gender ideology than boys will, and will be more egalitarian. Overall, I find

that youth with stepparents are more egalitarian, but my biggest finding suggests support for the

final hypothesis – that girls are more egalitarian than boys are.

LITERATURE REVIEW

Gender Ideologies

From the women’s social movements to the battle for same sex rights, gender, and gender

ideologies, permeate all aspects of life. Gender ideology refers to the, “system of values, beliefs,

and attitudes a person holds about the meaning of biological sex and gender” (Halpern and

Jenkins 2016:1). Furthermore, gender ideologies often include attitudes about how individuals

believe men and women should act in society. For example, if an individual believes that women

should stay at home and only have children, this reflects their traditional gender ideologies.

Egalitarian ideologies reflect equality between the sexes, in that people believe that men and

women should be allowed to do the same things in society, hold the same jobs, act in similar

ways, and not be treated badly because of it. The majority of research studying gender ideology

looks at beliefs about women working, such as whether people believe women should work

outside the home (Ciabattari 2001). In recent years, scholars have begun including gender

ideology measures that look at other areas that primarily affect women, such as child care,

fertility, relationship quality, and abuse. (Davis and Greenstein 2009). This research has

primarily focused on married women. These measures of gender ideology are important to look

at because they can influence societal policies including women’s work, women’s education, and

child care, as well as individual-level decisions about fertility and relationship quality.

Gender Ideologies’ Influence on Society

Gender ideologies can affect how a society views women in both beneficial and

detrimental ways. For example, a society that holds more traditional gender ideologies may be

more accepting of particular forms of discrimination, like sexism (Brandt 2011). Gender

ideologies vary based on the country of study (Lavee and Katz 2002; Kaufman and Bernhardt

2012; Kim and Choi 2012). These potential differences influence the opportunities women have

in their society, with some countries placing more importance on egalitarian ideologies while

others place more importance on traditional ideologies (Alesina et al. 2013; Twenge 1997). For

instance, in some countries women lead governments, while in others women are discouraged

from driving a car. Gender ideologies also help drive cross-national differences in percentages of

women in the workforce or in higher education (John, Shelton, and Luschen 1995; Kaufman and

Bernhardt 2012). One example of this takes place in Rwanda, where after their mass genocide,

policy makers implemented egalitarian laws that required a higher number of women in

parliament (Kristoff and WuDunn 2010). This implementation of egalitarian gender laws

actually helped to make their country grow and prosper (Kristoff and WuDunn 2010).

Most past research on gender ideologies focuses on the United States, or other Western

countries like the United Kingdom and Sweden. However, since women’s opportunities may

vary depending on the country of residence, it is important to understand the impact of gender

ideologies in other countries. There may be cultural or class differences within countries that are

different from the United States. In addition, since gender ideologies can have an impact on the

societal level, it is also important to look at the other impacts they can have.

Gender Ideologies’ Influence on Individuals

In addition to the impact that gender ideologies have on broader societies, they can also

have an impact at the individual level. For example, individual women’s feelings about gender

affect their choices concerning educational attainment, work force participation, and health.

Holding egalitarian views not only increases opportunities for women, but also benefits the

holders of these views, both male and female, resulting in more positive life outcomes. For

instance, individuals who hold more egalitarian views expected to gain more education (Davis

and Pearce 2007), had higher self-esteem (Weiss, Freund, and Wiese 2012), and were more

supportive of their partners in their personal goals (Tichenor 2005; Braun et al. 2008; Boudet et

al. 2013).

On the other end of the spectrum, individuals with more traditional gender ideologies

tended to be more negative towards others, and, among women, are less likely to seek

opportunities to advance their position in both the home and the workplace. For instance,

scholars have found that men who hold traditional gender ideologies are more likely to be

discriminatory or violent towards others (Totten 2003; Magallares 2016). Women with more

traditional gender ideologies tend to have lower earnings at work (Christie-Mizell et al. 2007),

suggesting they may not seek advancement. Since gender ideologies help shape opportunities for

later in life, it is important to understand why people hold the beliefs they do, and how they

develop these beliefs. Family formation, or romantic partnering, plays an important part in what

gender ideologies an individual holds.

The Impact of Family Structure: A Doing Gender Perspective

When looking at why adults hold particular ideologies, scholars have found that adults’

partnering is associated with their gender ideologies (Morgan and Walker 1983; Batalova and

Cohen 2002; Barber et al. 1989). These scholars believe this may have to do with the way adults

do gender with their partners. Doing gender is, “a complex of socially guided perceptual,

interactional, and micropolitical activities that cast particular pursuits as expressions of

masculine and feminine ‘natures’” (West and Zimmerman 1987:126). In other words, the doing

gender perspective suggest that individuals construct their gender through their behaviors in

everyday situations by performing roles fitting of their sex. The majority of past research that has

looked at how doing gender in the home impacts adult gender ideologies focuses mainly on the

division of labor between partners since the division of labor is often a measure of doing gender

between partners (Davis, Greenstein, and Marks 2007). This measure often looks at how partners

do gender together, such as with women doing housework or caring for a child, traditionally

feminine tasks, while men work outside the home (West and Zimmerman 1987; Deutsch 2007).

There may be something biological about the types of doing gender behaviors an individual

does, but for the most part, this doing gender process is largely a social construct (Goffman

1977). What goes on in the home, specifically the division of labor, impacts gender ideologies,

and doing gender can vary based on the family structure an individual resides in.

Adult gender ideologies: family structure and doing gender. Looking even further at why

family structure may affect adult gender ideologies, scholars have turned to doing gender and

found that doing gender behaviors vary depending on the family structure an adult resides in,

which then affects their gender ideologies. With marriage, divorce, and remarriage come norms

about doing gender behavior. There are differences in gender ideologies across stable coupled

adults, uncoupled adults, and re-coupled adults (Kiecolt and Acock 1988; Vespa 2009), likely

attributable to the distribution of doing gender behaviors, such as the division of labor, that

adults perform with their partners (Davis, Greenstein, and Marks 2007). Traditionally, scholars

have studied gender ideologies through the division of housework among couples. In other

words, the division of gender roles, and ultimately the gender ideologies adults hold vary based

on their relationship status because their division of housework varies based on relationship

status.

There are differences between the types of gender ideologies adults hold if they live in

traditional marriages or if they reside in more egalitarian cohabiting relationships. Much of past

research has found that married individuals are more likely to be traditional in all aspects of life

(Morgan and Walker 1983; Mason and Lu 1988; Wilkie 1993). Some research has found that this

traditionalism has lessened over time, but an effect remains (Harris and Firestone 1998). A

possible reason that many individuals in married relationships hold more traditional gender

ideologies is because of the traditional divisions of labor between partners.

While a single person must accomplish household tasks regardless of their gendered

position in society, married individuals divide household tasks in a much more traditional

gendered way (Vespa 2009). One example of this is that women in married households may be

more likely to perform housework while men are more likely to work outside the home (Wilkie

1993). As a result, if adults in these marriages split up labor in traditional ways, men are much

more likely to hold traditional ideologies than single men are (Wilkie 1993). While these patterns

may show a selectivity effect, it may very well be that the act of doing gender in ways that are

more normative cements gender ideologies on an everyday basis. However, people who have

egalitarian ideologies shift towards traditional ones after being married, often seen through their

division of labor, and this goes against the possible selectivity argument (Gupta 1999). This is an

important example of how family structure can lead to changes in behavior and beliefs. Past

research has found support for this when they have found that marriage brings with it a shift in

roles; for example, women who marry often do more housework while men who marry do less

housework, a reflection of more traditional gender ideologies (Gupta 1999).

Married people may hold traditional ideologies because they have a partner with whom to

do gender with, but what types of ideologies do people who cohabit hold? Cohabiting adults also

have a partner with whom to do gender with on a daily basis, so one might expect that they will

hold more traditional ideologies and will do gender in ways that are more traditional. However,

by choosing to cohabit instead of marry they are eschewing traditional means of doing gender.

As a result, these cohabiting couples appear to be going against traditional means of doing

gender, and therefore demonstrate more egalitarian ideologies than traditionally stable-married-

two-adult couples do. For example, scholars have found that cohabiting men contribute more to

housework, reducing the housework their partner would have to do and ultimately promoting

equitable living arrangements (Shelton and John 1993; Batalova and Cohen 2002; Davis,

Greenstein, and Marks 2007; Dominquez-Folgueras 2013). Other research has found that being

in a cohabiting relationship decreases the amount of hours of housework for women, but that it

matters less for men (Shelton and John 1993). Overall, scholars have found that cohabiting

women are not like married women, but are actually much more like single women in terms of

both their egalitarian ideologies and division of labor (Shelton and John 1993).

In addition to coupled adults, research also suggests that uncoupled adults hold different

gender ideologies. For instance, single women may go against traditional ways of doing gender

by eschewing settings that reinforce traditional gender ideologies. They do so by moving toward

more egalitarian gender ideologies, both by entering the workforce and gaining more education.

Scholars have found support for this claim as single women who have never been married spend

the least amount of time doing housework, suggesting a shift towards more egalitarian ideologies

(Shelton and John 1993). Looking at another type of single parents, scholars have found that

divorced women hold more egalitarian ideologies, especially about the division of labor in the

home, than married women do (Kiecolt and Acock 1988; Barber et al. 1989; Amato and Booth

1991; Moors 2003). Going deeper, scholars have found that divorce brings with it stressors such

as a restructuring of gender roles or changes in the levels of support outside the home (Lucier-

Greer and Adler-Baeder 2016). Since divorced women experience these particular stressors, they

may be less likely to enact traditional doing gender behaviors, thereby shifting their own gender

ideologies towards egalitarian ideologies. As a result, women who have divorced or chosen not

to marry may exhibit a willingness to break away from traditional means of doing gender, such

as getting married or raising children in a two-parent home (Ellman 2000). This suggests that

these divorced and single women may hold more egalitarian ideologies than women who are

married. Overall, single women, regardless of how they got there, tend to hold more egalitarian

ideologies compared to married or cohabiting women (Lucier-Greer and Adler-Baeder 2016).

In my research, in addition to looking at the potential differences between people who

have a partner with whom to do gender and those who do not have a partner, I also look at adults

who have both lived with a partner and been single. For these individuals, I expect that they will

be more egalitarian than married stable adults will because they did not always have a partner to

do gender with, similar to both single and divorced women. I also expect that they will be more

traditional than single adults will because they have been in coupled relationships where two

individuals split the housework and labor in traditional ways. Research on these re-coupled

adults tends to be limited, or looks more explicitly at how parents interact with children, not how

adults interact with each other (Sullivan 1997; Lucier-Greer and Adler-Baeder 2016). For

instance, some scholars found that women in remarried households actually do less housework,

which represents egalitarian forms of doing gender (Sullivan 1997). Other scholars have found

that this particular relationship structure is more complex, since these adults are more egalitarian

than if they were still single, but they also exhibit similar doing gender behaviors as married

couples (Lucier-Greer and Adler-Baeder 2016). Taken as a whole, regardless of the fact that

remarried adults are a more complex case, the idea that family structure has a powerful impact

on adult gender ideologies because they have a partner with whom to do gender is a very

powerful argument that warrants further research, especially for groups other than adults. I argue

that these doing gender behaviors of couples are both similar to and different from the doing

gender behaviors of parents. Parents do gender differently than couples, and since these doing

gender behaviors matter for gender ideology development, children may develop their ideologies

differently depending on the family structure they live in. Before I can argue that youth are

affected by the doing gender behaviors of their parents, we must first understand how doing

gender operates differently for youth and their parents than it does for couples.

Doing gender among parents and children. Doing gender can operate similarly for

parents and children as it does for couples in a few different ways. For example, partners may

negotiate with each other on whether women will perform traditionally feminine tasks such as

housework or child care; similarly, parents negotiate with their children when deciding whether

to have girls perform indoor chores while boys perform outdoor chores (Duckett, Raffaelli, and

Richards 1989). One example is that girls are more often likely to help in housework as a whole

than boys are, and much of girls’ involvement in the home is influenced by the mother’s work

outside the home (Gill 1998). Mothers working outside the home may impact the doing gender

behaviors they perform with their children; this is similar to how divorced women often work

outside the home more, an example of a less traditional doing gender behavior, and how this may

make them more egalitarian (Shelton and John 1993). Another example is that parents may

negotiate with their children when deciding whether their girls will wash dishes while their boys

will make home repairs (McHale, Crouter, and Tucker 1999). Much of this decision comes from

the influence of the father over his children (McHale, Crouter, and Tucker 1999).

As a result, it makes sense to assume that these doing gender behaviors and the

negotiation of them—for both couples and parents and children—may be influenced by family

structure. For example, single parents may need more child labor around the house, so they may

not have the luxury of choosing whether their children perform traditionally gendered tasks. This

is very similar to how single parents may be doing gender. Past research has shown us that single

adults do gender differently, possibly by choosing to be single or by working outside of the home

and, as a result, they hold more egalitarian ideologies (Kiecolt and Acock 1988; Barber et al.

1989; Shelton and John 1993).

At the same time, there are dynamics that differ across these kinds of relationships, so we

can see that doing gender may operate differently for parents and children than for couples. The

very doing gender behaviors that parents performed when they did not have children may shift

when there is another person in the picture, particularly a child (Katz-Wise, Priess, and Hyde

2013). In addition, individuals in coupled relationships may have different reasons for doing

gender with their partner than they do once they become parents, or they may have different

reasons for doing gender than their children. For example, women in coupled relationships may

be more willing to do gender in a traditional way in order to maintain their desirability to their

more traditional partners. On the other hand, youth are not likely to have similar reasons when

dealing with doing gender; their main reason stems from having a model within the home of

gender appropriate behaviors. This difference in reasons, as well as the shift in doing gender that

occurs when partners become parents, leads to the conclusion that doing gender operates

differently for partners than it does parents. For instance, within romantic relationships, doing

gender may be more of a push-pull relationship, in that partners give and take while they are

negotiating doing gender behaviors between each other, affecting the types of gender ideologies

they hold. On the other hand, doing gender in the home between parents and children is not a

push-pull relationship. Instead, this may be a more hierarchical relationship because children

look up to their parents as role models, not that parents pull their ideologies or behavior from

their children. In this hierarchical relationship, parents hold more of the power, which may make

them more influential concerning their child’s modeling behavior or ideologies.

While we see that there are differences and similarities between the doing gender

behaviors of couples and the doing gender behaviors of parents and their children, much of this

area is unexplored. It is important to study youth because this is the period where they embody

these gender ideologies, and this is when the socialization process is occurring. By studying

them, I can come to understand how and why they adopt particular views. It is also important to

study youth because youth may be more susceptible to the doing gender behaviors of their

parents, especially since youth look to their parents as models for their own attitudes and

behaviors and these models for behavior vary based on the types of doing gender parents

perform. For instance, youth may be more vulnerable to their parent’s gendered behaviors since

they spend most of their time within the home and, at such a young age, are more likely to look

to their parents as models for appropriate gender behavior and attitudes, whereas older youth

may look more towards other models of behaviors (Davis and Greenstein 2009). In addition,

since more youth today are living in diverse family structures, such as living with cohabiting

parents, divorced parents, or with parents who married after the child was born, it is important to

understand the differences between these family structures (Bumpass and Lu 2000; Mather

2010). Past research has only looked at youth gender ideologies on a basic level, examining

briefly both the types of ideologies they hold and how they developed, finding that parental

beliefs play a part in the ideologies a youth may hold (Davis and Greenstein 2009; Davis and

Wills 2010; Boehnke 2011; Goldberg et al. 2012). However, there has been limited research

focusing specifically on how family structure creates an environment for the doing gender

behaviors that parents perform with their children and how that influences the youth’s gender

ideologies (Kiecolt and Acock 1988; Russell and Ellis 1991).

Youth gender ideologies: the impact of family structure and doing gender. Since we

know that doing gender behaviors operate differently for parents and children than for couples,

and since family structure affects the doing gender behaviors of adults, it makes sense to assume

the same for adults. Research suggests that family structure the youth resides in affects their

gender ideologies (Kiecolt and Acock 1988; Russell and Ellis 1991). Gender ideologies seem to

vary from more traditional to egalitarian as family structure changes from two-parent bio to a

single. For example, if a youth resides with two biological parents who have always been

married, they are more likely to experience exposure to traditional ways of doing gender. Past

findings suggest that women with traditional beliefs are more likely to get married, have

children, and stay married to their child’s biological father, which are all actions reflective of

traditional ideologies (Kaufman 2000). Since these biological two-parent married families are

more likely to be traditional (Morgan and Walker 1983; Mason and Lu 1988; Kaufman 2000),

youth may be more likely to hold similar traditional ideologies because two parents of opposite

sexes expose them to traditional ways of doing gender in the home. These traditional parents

may be more likely to enact traditional means of doing gender with each other, but they are also

more likely to enact traditional means of doing gender with their children, thereby affecting their

child’s ideology.

Whereas for children raised by single parents we would expect to see the opposite, that

they would be more likely to hold egalitarian ideologies. For example, if a youth resides in a

single-parent family, they may hold the least traditional gender ideologies. This may be because

children experience less exposure to traditional types of doing gender because the parent with

whom they live has no partner with whom to enact traditional gendered roles and behaviors. In

addition to experiencing exposure to less traditional ways of doing gender, these youths may

have less reinforcement because they spend less time with their single parent because of their

single parent working. These youths may have less reinforcement because they have seen their

single parent restructure his or her life because of a divorce or separation. Past research has

found that youth with single mothers held egalitarian ideologies (Russell and Ellis 1991), since

less traditional women head them and expose them to this dual-role of doing gender (Kiecolt and

Acock 1988; Russell and Ellis 1991).

In addition to studying stable two-parent biological married families and single-parent

families, having stepparents or social parents may expose youth to alternative ways of doing

gender. While these youth have parents of both sexes who perform traditional means of doing

gender with their children, residing in stepfamilies may result in egalitarian ideologies since

stepparents may be less physically or emotionally available to stepchildren and less willing to

parent a child that is not theirs (Lampard and Peggs 1999; Carlson and Knoester 2011). This can

lead to weaker reinforcement of traditional ideologies since children are not modeling after their

stepparent with whom they may have no relationship. As such, youth in these families

experience more exposure to less traditional examples of doing gender than they would if they

lived with both of their biological parents. However, some past research finds that having

stepparents results in more traditional child gender ideologies than having a single parent, but

results in more egalitarian ideologies than if the youth lived in a stable biological married family

(Janning, Collins and Kamm 2011; Carlson and Knoester 2011; Halpern and Jenkins 2016).

Others have found that there is no difference in gender ideologies when children have either a

biological father or a stepfather, since these children have a same-sex model in the home (Kiecolt

and Acock 1988).

Youth residing with cohabiting parents may hold more egalitarian gender ideologies than

those in stable two-biological married parent families, and may hold gender views closer to

youth residing with stepparents. These youths may have more egalitarian gender ideologies

because their primary role models for doing gender, their cohabiting parents, have made choices

in their life to reject traditional family formation norms. Cohabiting individuals are more likely

to share housework equally (Lye and Waldron 1997; Davis, Greenstein, and Marks 2007), and

since this is a reflection of more egalitarian ways of doing gender in the home. Children exposed

to that lifestyle may be more likely to model their attitudes similarly. I do not measure

housework, but instead I examine family structure as a representation of these doing gender

behaviors, which then affect their children’s gender ideologies.

Some past research has found support for my current research project, in that they have

found that family structure does matter for the development of youth gender ideologies. When

looking at single parents for example, past scholars have found that youth with single parents are

more likely to hold egalitarian ideologies (Russell and Ellis 1991). I argue that most of this

research may not reflect current trends in gender ideologies and family structure, hence why

current research is necessary. Past research has also tended to lump together parents who have

been single since the birth of their child with parents who become single after the birth of their

child (Kiecolt and Acock 1988); this conflation may miss important nuances regarding the

impact of stability on doing gender behaviors. In addition, some research has claimed that family

structure and social background do not matter for youth gender ideology especially as they age,

but these were smaller sample sizes and may also not be reflective of current gender ideology

trends (Davis 2007). Past research has primarily focused on single parents, two-parent families,

or stepfamilies, did not look at all family structures in one model, only included small-scale

studies that do not lend power to their argument, or found mixed results on whether or not family

structure matters (Kiecolt and Acock 1988; Davis 2007; Janning, Collins and Kamm 2011).

Overall, few recent studies have looked at how family structure affects doing gender behaviors,

or parents influences and how they do gender with their children, ultimately affecting their

child’s gender ideologies. Since there are few studies that look at this specifically, what is

especially about my research is that because I have access to better data it allows me to examine

family structures beyond the dichotomy of married, single, or recoupled. It allows me to get at

the nuances that exist within past research’s broad categories.

Youth gender ideology: what about gender? Another formative component of youth

gender ideology that may be affected by family structure and the doing gender behaviors of

parents is gender itself. Girls experience the world differently than boys do, and have different

outcomes. For example, when looking at popularity, scholars have found that girls tend to gain

popularity because of feminine and ascribed characteristics such as their looks, while boys

tended to gain popularity because of their athleticism, toughness, or success (Adler, Kless and

Adler 1992). In addition, women are less likely to be in leadership positions, partly because men

wanted more power in their lives (Schuh et al. 2014). Another example is that in recent years,

females have become more likely to finish college (Buchmann and DiPrete 2006); this may be a

reflection of their gender ideologies.

In addition to experiencing socialization through societal influences, family, especially

parents, also influences boy and girls and, consequently, their ideologies. Fathers and mothers

may parent their children differently based on sex (Raley and Bianchi 2006; Davis and

Greenstein 2009). For instance, research shows that parents teach their boys to take more risks

while they teach girls to take fewer risks (Morrongiello and Dawber 1999). In addition, past

research has shown that same-sex role models are incredibly important (Bozett 1985). Since boys

and girls have different outcomes in life, and are treated differently based on their gender, it is

important to look at how gender influences gender ideology—since gender ideology can

influence their outcomes, and is affected by family structure and doing gender. I argue females

will be more susceptible to the effects of doing gender and family structure. This may tie back to

the power division between parents and children; perhaps parents have more power over their

daughters, which would then give them more influence over their daughter’s gender ideologies.

It may tie back to how parents do gender differently with their children than with their partners.

Furthermore, I believe that females will be more likely to hold egalitarian ideologies because

females may invest more in the outcomes of gender norms, such as workforce participation,

household division of labor, and educational opportunities. By providing a setting for doing

gender, family structure affects youth gender ideology; these ideologies may in turn affect the

opportunities youth encounter. This may be especially true for girls who wish to enter higher

education or find higher paying jobs, as traditional ideologies can become barriers to doing so

and as they may experience exposure to different doing gender behaviors from their parents.

Overall, we can see that family structure does influence doing gender behaviors of

couples, which affects the gender ideologies they hold (Kiecolt and Acock 1988; Vespa 2009).

They find that single adults tend to be egalitarian and married adults tend to be traditional

(Amato and Booth 1991; Wilkie 1993). In addition, some research has also shown that family

structure matters for youth gender ideologies, though this research is limited (Russell and Ellis

1991; Kaufman 2000). What they have found reflects what researchers have found for adults. We

also see that females may have different gender ideologies than males will because of their

family structure and their parents’ doing gender behaviors (Bozett 1985). However, this past

research does not really look at how family structure may operate in different ways for couples

than it does for parents, especially concerning doing gender behaviors in the home. I argue that it

makes sense to assume that family structure operates through doing gender behaviors in slightly

different ways with parents and children than it does with couples. To examine the potential

relationship between family structure and youth gender ideology through the lens of doing

gender, I look at married two-parent families, single-parent families, stepparent families, and

two-parent cohabiting families.

HYPOTHESES

Since I believe that opportunities to do gender vary based on family structure, and since

youth are more vulnerable to these effects as they develop, I lay out the following hypotheses:

1 – Youth living in single-parent families will have more egalitarian gender ideologies

than youth in stable two-parent married families. Youth exposed to only one parent have less

exposure to traditional ways of doing gender than if they experienced exposure to both sexes.

2 – Youth living in stepfamilies will have more egalitarian ideologies than stable two-

parent married families. Youth exposed to two married parents will have traditional gender roles

reinforced more frequently and intensely, but because stepparents may be reluctant to parent a

child who is not biologically theirs (Lampard and Peggs 1999), they may present a weaker

version of traditionally gendered family roles than would two married biological parents. In

addition, exposure to a single-parent lifestyle, where their ideologies may become more

egalitarian, may create effects that last into the remarriage or re-partnership of their parents.

3 – Youth living in cohabiting families (social parents) will have more egalitarian gender

ideologies than youth in stable two-parent married families. While cohabiters have access to

another person to do gender with, they have made an affirmative choice to reject traditional

family formation norms, many of which come with gendered role expectations already attached.

4 – Girls will be more sensitive to these family structure and doing gender effects than

boys will, and will therefore be more egalitarian than boys will. Family structure may affect girls

more strongly than boys since societal pressures revolving around gender norms may personally

affect females, and since they may be more susceptible to parental doing gender behaviors.

DATA & METHODS

For this project, I use the German National Educational Panel Study (NEPS), a nationally

representative study, which surveyed and interviewed students at German schools using a

longitudinal design. The NEPS dataset has two youth cohorts: one starting in fifth grade and

ending in eighth grade and another starting in ninth grade and ending in twelfth grade. The

survey focused mainly on educational pathways and outcomes, but information on gender

ideologies is also available. Students and their parents participated in the survey through their

school systems. I use the NEPS dataset because not only do they capture ages that are vulnerable

to the effects of family structure on gender ideology development, but they also include

interesting measures of gender ideology. My current research project studies gender ideology by

looking at perceptions of division of labor, technical skills, and career types. I also use NEPS

data because Germany is an interesting case to study in that Germans tend to have more

egalitarian ideologies than Eastern countries or the United States, and these more egalitarian

ideologies have led to more national-level childcare and welfare opportunities (Dirksmeier

2015). In addition, gender ideology research tends to focus mainly on the United States, so

Germany is an interesting case to study.

From the first cohort, I use the second wave (average age 11); from the second cohort, I

use the third wave (average age 15). I use both cohorts because I wanted to see if the

socialization process occurs differently as the youth age. After excluding cases that were missing

information on gender ideology outcomes and family structure, the total sample size for the first

cohort is 4,181, and the total sample size for the second cohort is 9,913. I analyze each cohort

separately and compare the effect sizes across the two age groups as a partial test of how gender

ideologies develop as children age.

Gender Ideology Measures

Gender ideology can refer to “the underlying concept of an individual’s level of support

for a division of paid work and family responsibilities that is based on the notion of separate

spheres” (Davis and Greenstein 2009:89). The cohorts I use touch on beliefs about gender in

education or the workplace, femininity, and household labor divisions, which is similar to

previous work (Davis and Greenstein 2009). Within the NEPS, I measure gender ideology

through a set of questions that ask respondents to record their responses on a scale of 1 “disagree

completely” to 4 “agree completely.” NEPS gender ideology statements for 6th and 10th graders

include statements such as “boys and girls should have the same chores at home,” “girls can

handle technical devices just as well as boys,” “girls should be able to learn the same professions

as boys,” and “for some professions, men are better suited than women.” I reverse recoded the

last statements so that higher values reflect more egalitarian gender ideologies. I then constructed

a gender ideology scale by using the row means of the gender ideology statements and dividing

by the total number of statements, resulting in a final scale of 1 to 4: “completely disagree” to

“completely agree” where higher values represent more egalitarian gender ideologies. Finally, I

dropped any cases that were missing information on more than two of the questions. See Table 1

for a full description of all gender ideology measures, as well as other variables used in analysis.

[Table 1 about here]

Family Structure Measures

To capture family structure, I created my own categories from available NEPS data. To

do so for each cohort, I use the available parent data (including marital status, whether the parent

lives with a partner, relationship to the youth, and the year they began living with a partner); if

parent data were missing, then I used information on youth (including household composition

and role of parents) to fill family structure. If both information on the parent, the parent-child

relationship, or household composition data were missing, but the main parent was living with

the same partner as before, I filled in missing data using family structure data from the previous

waves. After completing that process, if family structure was still unknown, I dropped the case.

Cohort two family structure questions varied slightly from cohort one family structure questions.

Within cohort two, family structure questions, such as parental marital status, parent-child

relationship, household composition, and role of parents, are present in some waves. As a result,

I used all of the questions available in waves one through three to create these family structures.

The final family structure categories I created using the NEPS include “biological married

stable,” “biological cohabiting stable,” “biological single stable,” “post-birth single,” “post-birth

stepfamily,” “post-birth social family,” and “post-birth biological cohabit/married family.”

Three categories capture stability within family structures. “Biological Married Stable”

applies to youth who have lived, and continue to live, with two married biological parents.

“Biological Cohabit Stable” captures whether the youth currently lives with two cohabiting

biological parents, and has lived with them since birth. “Biological Single Stable” applies to

youth who have lived, and continue to live with one single biological parent.

Four categories capture instability within family structures. “Post-Birth Single” captures

whether the youth currently lives with a single parent who became single after the youth was

born, through divorce, separation, or death. “Post-Birth Stepfamily” applies to youth who

currently live with one biological and a stepparent. For example, a youth living with a biological

father who remarried someone other than the biological mother fits this category. “Post-Birth

Social Family” applies to youth who live with one biological parent and one social parent who

started cohabiting after the youth was born. This is similar in nature to the “post-birth

stepfamily” with the sole difference being that the parents in the previous category are married.

My final family structure variable, “post-birth biological married/cohabit,” includes children who

reside with their biological parents but whose parents began living together after they were born.

Table 2 presents the percentages of youth within these family structures.

Youth Gender

I also include youth gender as part of my interaction effect because I believe females are

more likely to hold egalitarian gender ideologies since societal laws revolving around gender

norms in the home and workplace may affect females more personally. In addition, I include

gender because past research has shown that parents parent their sons and daughters differently

(Morrongiello and Dawber 1999), and I believe this affects the development of gender

ideologies. I recoded child gender so that 0 equaled “female” and 1 equaled “male.”

Control Measures

In addition to including gender as a separate control, I also include parental education,

parental employment status, household income, child age, youth religious affiliation, and youth

religiosity as general controls. Since parents provide a role model for youth, I control for

parental education. For example, adults with more education tend to have more egalitarian

gender ideologies (Alwin 1989; see also Davis and Greenstein 2009). As a result, parental

education may influence child gender ideologies, so I control for it and other factors in my final

models. I constructed maternal education to reflect major credentials: “lower general education –

less than high school,” “high school equivalent,” “some college (or other similar qualifications),”

“bachelor’s degree,” and “doctoral degree or similar.” In addition to education, occupation and

income are two indicators of socioeconomic status that can influence parental gender ideologies

(Alwin 1989).

I use maternal employment instead of both partner and main parent employment because

I focus mainly on attitudes about female equality and opportunities. Maternal employment was

recoded to indicate levels of participation in the paid work force: “full time," “part time,"

“unemployed but looking for work,” “student/in training,” “homemaker," and “no mother in

home.” Income in the NEPS is a continuous variable that I logged due to skewness. This resulted

in a range from about 7 to about 14. In addition, I also include religious denomination since

some religions may subscribe to more traditional gender ideologies than others. I also included a

measure of religiosity because individuals who actively participate in religious activities may

hold more traditional gender ideologies. Religious denomination categories include “Roman

Catholic,” “Protestant,” and “other,” while religiosity categories range from “not at all religious”

to “very religious.” I originally included youth nationality because I believed that individuals

who are immigrants into Germany might have different gender ideologies, but due to small cell

size, I was unable to use this measure. I attempted to get at nationality through language spoken

at home, but it left me with the same problem as mentioned above. I do acknowledge there are

limitations associated with not including nationality measures in my models.

Missing Values

As with all research, missing values can affect the accuracy of estimates. I fill in missing

values for family structure using available parent data, such as marital status and relationship to

youth; if parent data was missing, then I filled in missing using youth data, such as household

rosters and role of parent data. If both youth and parent data was missing, these family structure

cases were dropped. For missing values on the gender ideology measure, I dropped if they were

missing two or more items from the scale. For control variables that had missing information, I

performed single imputations where I used other variables to predict missing values on controls

such as education or occupation. I compared means of control variables before and after

imputations and found similar distributions before and after.

Analysis Plan

I perform separate analysis for each of the two cohorts. I first present the percentage of

youth in various family structures; I also present descriptive statistics on the gender ideology

scale, as well as other variables, across family structures (Table 2). This allows me to examine

whether there are differences in average gender ideologies scores across family structures. I then

conduct multivariate analyses (Table 3) using ordinal logistic regression, presented in odds

ratios. To calculate the percentage of the likelihood of holding more egalitarian gender

ideologies, I subtract the odds ratio coefficient from one and then multiply by 100 ((1-odds ratio

coefficient) *100). As a result, odds ratios higher than one have a positive effect while odds

ratios lower than one have a negative effect. First, I present a model with only family structure

and gender ideology (Model 1). I then present a model with only youth gender and gender

ideology (Model 2). The third model includes family structure, youth gender, gender ideology,

and all controls (Model 3). I also run another model with an interaction term between youth

gender and family structure (Model 4). I do this in order to see whether girls are more sensitive

to the effects of family structure on the development of their gender ideologies than boys are.

RESULTS

My first analysis looks at the number and percentage of youth in various family structures

(Table 2).

[Table 2 about here]

My findings suggest that the majority of youth in this sample (about 70 percent) reside in a stable

two-parent biological married family, regardless of cohort. The next largest percentage (about 13

percent) captures youth who reside in post-birth single families–in families where the child’s

main parent was divorced, separated, or widowed after the child’s birth. When looking at other

research, results suggest that Germany has a different distribution of youth across family

structures than other countries (Dufur et al. 2016). This could mean that any potential differences

we see in the effects of family structure on gender ideology may occur because of cultural

differences in Germany. It also reiterates the point that Germany is an interesting case to study

since youth are living in different families than youth in other countries are. Older youth are

more likely to reside in post-birth stepfamilies, biological cohabit stable families, post-birth

single families, or post-birth biological families than younger individuals are (Table 2).

My second analysis looks at the means of gender ideologies across family structures for

both cohort one and two (Table 2). I found that the majority of youth, regardless of their family

structure, have middle-range gender ideologies in both cohort one and cohort two (Table 2).

There is little variation across the family structures, calling into question my assertion that family

structure would help shape youths’ gender ideologies. However, I do see that younger kids (the

first cohort) have less egalitarian ideologies in biological cohabiting families, stepfamilies, and

post-birth biological married/cohabit families (Table 2). On the other hand, older youth in post-

birth stepfamilies and biological cohabiting families are more egalitarian (Table 2). Regardless

of cohort, youth residing in single families tend to be the most egalitarian, and that those in post-

birth social families are less egalitarian compared to those residing in stable biological married

families (Table 2). This suggests that perhaps there are cultural factors in Germany that shape

both family formation practices and gender ideologies that differ substantially from otherwise

similar countries, making Germany an even more intriguing setting for future study. Older kids,

cohort two, are more likely to hold egalitarian gender ideologies regardless of family structure,

except for post-birth social families where the gender ideology mean is similar across both

cohorts (Table 2).

I then examine the number and percentage of females, slightly religious youth, as well as

the number of youth with mothers working full time, mothers with bachelor’s degrees (Table 2).

The majority of youth across both cohorts in biological two-parent married stable or in post-birth

single parent families are female (about 50 percent). Females may be more susceptible to the

doing gender behaviors of their parents, so if more children in these traditional family structures

are female, they may be more likely to hold similar traditional gender ideologies. In addition,

mothers in single parent families are much less likely to hold bachelor’s degrees. In post-birth

social families, single stable families, and cohabiting stable families, youth are more likely to

have mothers who are working full-time than in other family structures (Table 2). I do find that

the older cohort has a higher percentage of mothers working, suggesting that there may have

been a shift in gender ideologies towards more egalitarian views as the youth age, or that women

with older children are more likely to work (Table 2). From my sample, it seems that a majority

of youth in Germany tend be slightly religious, but that youth in biological two-parent stable

married families are more likely to be “slightly religious” than youth in other family structures.

This may affect the socialization process in the home. However, the older cohort is slightly less

likely to be “slightly religious” than the younger cohort is (Table 2).

My multivariate analysis (Table 3) provides little support for my hypotheses. Before

including controls, within cohort one, children residing in a post-birth biological cohabit/married

family are less likely to have egalitarian gender ideologies, and children in post-birth

stepfamilies are more likely to hold egalitarian gender ideologies (Model 1). Not only did I find

no support for hypothesis four from these results, but I also found the opposite. Children residing

in post-birth biological cohabit or married families are actually less egalitarian than those in

stable biological married families are. I do find initial support for hypothesis two for cohort one

when I find that youth in stepfamilies are more egalitarian than youth in stable biological married

families. On the other hand, in cohort two, before including controls, no family structure was

significant (Model 1). This suggests that there is a change in the socialization process within

families as youth get older, or that younger youth are more affected by family structure and their

parents’ doing gender behaviors than older youth are.

[Table 3 about here]

After including controls, I find that living in a post-birth biological cohabit/married

family significantly results in a youth being less egalitarian, and living in a post-birth stepfamily

results in the opposite for cohort one (Model 3). Once again, I find no support for hypothesis

four, but do find support for hypothesis two. For cohort two, living in a biological cohabit stable

family results in youth being more traditional (Model 3). I do caution the reader because this is a

small group. This result initially rejects hypothesis 3, though I believe there are limitations

associated with this result that I will discuss later; I originally claimed that living in a cohabiting

family would make youth more egalitarian, but it actually does the opposite. For cohort one,

having a mother who works part time results in a youth being more traditional compared to youth

living with a mother who works full-time. This result makes sense when considering that women

who work full-time are more likely to hold more egalitarian gender ideologies that they model

for their children. Surprisingly, greater maternal education was not associated with more

egalitarian gender attitudes among youth for either cohort. In addition, having an older mother

increases support for egalitarian gender ideologies, while being slightly non-religious, slightly

religious, and very religious decreases support for egalitarian gender ideologies as compared to

being not religious. Both of these results are significant only in cohort two (Model 3). This

suggests that family structure matters less for older youth, and that other parental and youth

characteristics matter more. Overall, my first multivariate analyses suggest support for

hypotheses two and three, and rejects hypothesis one and four. Hypothesis two claimed that

residing in a stepfamily would result in youth holding egalitarian gender ideologies than if they

lived in a stable two-parent biological married family. Hypothesis three claimed that youth living

in cohabiting families would be more egalitarian than if they had been in stable two-parent

biological married families, but I found the opposite was true for the older cohort. This category

is smaller than the other categories, so if future research was to look at this category with a larger

sample size, the results may support the original hypothesis.

The most notable effect comes into play with youth gender. Females, even before

including family structure or other controls, are more likely to hold egalitarian gender ideologies

than males (Model 2). After including family structure and controls, for cohort one, females are

five and a half times more likely to hold egalitarian gender ideologies than males; in cohort two,

females are eight and a half times more likely to hold egalitarian gender ideologies than males

(Model 3). This gender effect remains strong after controlling for maternal employment and

education, as well as family income, religiosity, and religion, suggesting that gender itself is a

large motivating factor in whether an individual will hold egalitarian gender ideologies. This

shows partial support for my final hypothesis that females are more egalitarian.

Another part of my final hypothesis was that females would be more susceptible to the

effects of family structure and doing gender. We assume that since gender mattered for previous

models, that it would matter with combined with family structure. In other words, I assumed that

parents would parent their sons and daughters differently, since past literature has shown similar

results (Raley and Bianchi 2006; Davis and Greenstein 2009), and that this would then affect

their youth’s gender ideology. After running models with the interaction between youth gender

and family structure and performing various statistical checks, I found that the cell sizes of the

interactions prevented me from finding support for my original claim. Therefore, I do not present

those results.

DISCUSSION

Looking back at the past research, we see that there are a few main arguments that are

important to reiterate before delving into what this all means. One is that gender ideologies

matter for a variety of outcomes, both at the societal and individual level (Christie-Mizell et al.

2007; Brandt 2011), and this is why gender ideologies are so important to study. Second is that

the family structure adults currently reside in affects the doing gender behaviors they perform

with their partner, which then affects their gender ideology. Past scholars have found support for

this claim when they found that single adults are more egalitarian while married adults are less

egalitarian (Kiecolt and Acock 1988; Amato and Booth 1991; Wilkie 1993; Vespa 2009). The

third argument is that this argument from adults can also apply to youth, but in slightly different

ways. I argued that the family structure a youth lived in affected the doing gender behaviors that

their parents perform with them, which then impacts the youth’s gender ideology development

through both reinforcement and exposure to traditional or egalitarian ideologies. Some scholars

have found results to support parts of this claim, but the majority have focused on very limited

family structures or does not delve deeper into the doing gender argument (Russell and Ellis

1991; Kaufman 2000). The final argument I made, which scholars found support for, is family

structure affects youth gender in different ways, which then affects the development of youth

gender ideologies. Scholars have found that parents raise females differently than males, and that

this can affect their gender ideology development (Bozett 1985). Overall, I originally argued that

family structure would matter for the development of youth gender ideologies because of the

doing gender behaviors that parents perform with their children. My results show support for

some of these arguments, but does not show support for my main argument – that family

structure matters.

My second and third argument revolve around the idea that family structure affects

gender ideology through doing gender behaviors. When looking at adults, previous research has

found that parental coupling affects doing gender, which in turn affects gender ideology (Wilkie

1993; Lucier-Greer & Adler-Baeder 2016). For instance, some past research found that married

adults are more traditional than single, divorced or cohabiting adults are which is evident when

looking at their division of housework (Morgan and Walker 1983; Wilkie 1993). Furthermore,

past research has also found that some family structures exert an egalitarian influence on youth

gender ideologies through the doing gender behaviors, mainly division of housework (Shelton

and John 1993; Vespa 2009; Lucier-Greer and Adler-Baeder 2016). Doing gender operates

through family structure for couples like mentioned above, so one would assume that it does

operate through family structure for parents and children, even if it was in slightly different

ways. I did find that youth living with stepparents are more egalitarian, which supports one of

my original hypotheses. This result may occur because children are more likely to have exposure

to less traditional doing gender behaviors with stepparents, or because they do not have a

relationship with the stepparent which allows for the doing gender behaviors to translate to

gender ideology development (Carlson and Knoester 2011). However, my results overall suggest

that other than for younger adults living in stepfamilies, family structure really plays no part in

affecting the doing gender behaviors parents perform with their children, or that it does so in

other ways, and, therefore, family structure does not affect the development of gender ideologies

among youth in Germany. This is interesting because it goes against what I thought would

happen based on past research (Russell and Ellis 1991; Kaufman 2000). My null findings shows

us that because society is constantly changing, that socialization can occur through different

pathways, and that perhaps family structure operates in different ways than how past scholars

have imagined.

What my null findings do suggest is that what couples do in their home with their

children to do gender does not affect youth gender ideologies. These shifts in partnering that

adults experience may force them to confront how they do gender, but this does not pass down to

their children. Even using the most comprehensive family structure models, family structure still

explains very little of the differences in gender ideologies for youth in Germany, which goes

against past research (Russell and Ellis 1991; Kaufman 2000). This allows us to rule out family

structure, and to look elsewhere for why gender ideologies vary across youth. Perhaps this null

finding has to do with how we are describing the relationship of doing gender between parents

and children– maybe this relationship is not hierarchical, but actually much more push and pull

like for adult couples. Perhaps the relationship between parents, doing gender, children, and

youth gender ideology is similar to the relationship of couples – a push-pull relationship between

individuals instead of a hierarchical one. This gap is one area where future research could shed

more light on.

Furthermore, it makes sense to claim that what actually matters for the development of

youth gender ideologies may be how adults do gender in other ways, specific personal

characteristics of the youth or parent, or even societal influences I cannot capture. In addition, a

youth’s characteristics, such as their gender or religion, matters more than family structure for

the development of gender ideologies. For instance, I found that having a mother who works part

or full time increases the odds that the youth will hold egalitarian gender ideologies. Past

research echoes this result, because women working has often been associated with egalitarian

ideologies (Brewster and Padavic 2000). It makes sense to assume that this effect may transfer to

the child through modeling that occurs in ways other than doing gender negotiation between

parents and children. In addition to having a working mother, my results suggest that having an

older mother also increases the odds that the youth will hold egalitarian ideologies. One

theoretical explanation for this is that women who put off having children may be choosing to

eschew more traditional doing gender behaviors since what much of society expects of women is

to have children or to have them earlier. These results are prime examples of showing that the

ways adults do gender outside the home, such as having children later or working outside the

home, matter more than the ways they do gender in the home or with a partner. These doing

gender behaviors outside of the home or relationship may still provide models for children in

ways I did not capture.

In terms of finding that youth characteristics matter for the development of gender

ideologies, I find that individual youth characteristics such as religion and religiosity influence

gender ideology development of youth in Germany. For instance, compared to those who are

very not religious, those who are more religious are more likely to hold traditional gender

ideologies. Past scholars agree that those who are less religious are more likely to be supportive

of LGBTQ relationships, one marker of an egalitarian ideology (Harbaugh and Lindsey 2015). In

addition, I also found that older kids, who are a part of an “other” religion, which includes more

conservative religions like Muslim or Jewish, are more traditional than those who are Roman

Catholic. The fact that this finding is only present for the older cohort suggests religiosity and

religious identity matter more as youth get older. Furthermore, the finding that religion and

religiousness matter suggests that the experiences youths have in their own lives may be more

important for developing their ideologies than the family structure they lived in growing up. It

also suggests that perhaps religion is more of a socializing agent than family, at least in this

context. Past research echoes this, finding personal experiences of daughters matter more than

their mother’s ideologies (Moens, Erickson, and Dempster-McClain 1997). This may show

support for the first part of my argument, in that family structure of adults does matter for their

gender ideologies, since the family structure they currently reside in may reflect personal

experiences that have affected their gender ideology, as opposed to the family structure they

lived in growing up.

Concerning my final argument, I found support for it when I found females are more

egalitarian. As I discussed earlier, one theoretical explanation for this result may be that females

invest more in the outcomes of gender norms, such as workforce participation, household

division of labor, and even educational opportunities (John, Shelton, and Luschen 1995;

Kaufman and Bernhardt 2012). Past research has shown that more egalitarian gender ideologies

result in higher educational attainment, higher self-esteem, more involvement in the workforce,

in government, and in education, especially for women (Davis and Pearce 2007; Weiss et al.

2012). For instance, females may be more likely to benefit from higher educational attainment or

more involvement in society. Since females may be more invested in these outcomes, may

benefit more from them, and are more likely to hold egalitarian gender ideologies, researchers

should be aware of the bigger impact gender has on gender ideologies. I will mention more

specific policy implications later on. Policy makers should be aware that gender plays a huge

part in the acceptance of particular gender ideologies, which can influence the opportunities

individuals have in society. If females are more likely to hold egalitarian gender ideologies,

politicians and lawmakers should be conscious of this when creating laws that involve gender.

This gender component transcends cultural barriers as well, given that our results are for

youth in Germany, but we may see similar trends in the United States. Perhaps these findings are

occurring in Germany because of factors I cannot measure. German culture may be more

accepting of women as a whole, or that women in Germany may have more opportunities open

to them, so they become more egalitarian. Perhaps socialization in German families is different

from American families. This is one area future research can explore further.

Another part of my original argument dealing with youth gender revolved around family

structure and gender. Since I found that being female influences the ideologies one holds, I also

thought that youth gender would specifically play a part in how parents do gender. Past research

has shown that fathers and mothers parent their children differently depending on sex (Raley and

Bianchi 2006; Davis and Greenstein 2009). For example, as I discussed earlier, parents teach

their boys to act differently than they teach their daughters to (Morrongiello and Dawber 1999). I

had expected that females living in less traditional families would be more egalitarian, or that

males living in stable married families would be more traditional. After running these

interactions between youth gender and family structure, I find that youth gender on its own plays

a part in the ideologies a youth holds, but that family structure does not interact with gender to

influence gender ideology. In other words, I found no support for the idea that parents raise their

children differently based on sex, or that living in particular family structures has more impact on

females than males. This result may be because my cell size may have caused problems with the

analysis, so this does not suggest that future researchers should not study gender and family

interactions, only that they need better data so that they can still capture the detail of various

family structures, while still being able to run analyses.

I acknowledge that there are limitations to my research; for instance, missing control

variables, such as nationality/race, may also play a part in the development of gender ideology

for youth. If future researchers were able to create surveys that captured both measures of gender

ideology as well as control variables connected to ideology development, such as nationality,

then perhaps scholars could get a more complete picture of how gender ideology develops. In

addition, while my results suggest that family structure does not play a part in the development

of gender ideology, I acknowledge there are other aspects of family structure I cannot capture

with my dataset. For instance, shifts in family structure or the time spent in various family

structures may influence gender ideology development more than the actual family structure. I

was also not able to include cross-sex analysis between parents and children, and I was not able

to know much about non-custodial or joint custody arrangements. These may have impacts on

the doing gender behaviors that children are exposed to. Future research may be able to help

resolve my research’s limitations. In the future, I would like to study these extra areas relating to

family structure, but my current data are not set up for studying other aspects of family structure.

In addition, my research does not look at the intergenerational transmission of gender ideology,

mainly because I do not have data on how the parents themselves grew up. As a result, I suggest

that future research looks at whether adults are more susceptible to gender ideology shifts as

their relationships and family partnering change.

Overall, when looking back at the outline of my arguments (that family structure affects

adult gender ideologies through gender behavior, that family structure does the same for youth in

slightly different ways, and that youth gender matters for gender ideology development), I find

limited support. Family structure does not appear to affect gender ideology through doing

gender. However, due to important limitations, future research on the impact of family structure

on gender ideology, should examine how societal trends influence families and ideologies over

time. The fact that previous research has found effects of adult romantic coupling on gender

ideologies, but has claimed family structure effects, is an important distinction that much of the

research has not made clear. In addition, my null findings suggest family structure does not

matter for gender ideology development, but it is possibly some other way of doing gender that

matters. What I mean by this is that the doing gender behaviors that parents perform with their

children, such as how they split the division of labor or even other sex-typed modeling, may not

matter as much as the doing gender behaviors they do outside of the home, such as having a

more educated mother or having a working mother.

My findings have important consequences for policy makers as well. For instance, policy

makers should be aware of religious differences in the implementations of gender policies,

possibly making more religion-specific gender policies in the workplace. In addition, we know

that parental characteristics really matter for the development of gender ideologies among youth.

I believe that policies can be implemented (or improved) that can influence parents, which can

then affect their children. For instance, if society implements policies so that more women

entered the workforce, then perhaps when women become mothers, their egalitarian ideologies

will trickle down to their children because they provide a model of egalitarian doing gender

outside the home. In addition, they could implement laws that allow women to return to the

workforce after having children, or policies that allow women with younger children to work,

such as having childcare facilities available. In addition, one of my other findings was that older

mothers are more likely to have egalitarian youth. One possible implication from this revolves

around birth control and pregnancy policies. I am not advocating that all women have children

later, but that gender oriented policies should make this option available to women. In less

developed societies, many women have children earlier in their life course, which may prevent

them from gaining an education or from working (Kristoff and WuDunn 2010). If other countries

could implement similar policies that allow women to gain an education, or work while they are

mothers, this modeling can then affect their children. This intergenerational transmission could

possibly resulting in a more egalitarian society as a whole. These are just a few examples of

policy implications that can come from my research findings, but this only goes to show that

society need to make more strides in making things equal for all.

possibly making more religion-specific gender policies in the workplace. In addition, we know

that parental characteristics really matter for the development of gender ideologies among youth.

I believe that policies can be implemented (or improved) that can influence parents, which can

then affect their children. For instance, if society implements policies so that more women

entered the workforce, then perhaps when women become mothers, their egalitarian ideologies