Introduction

Every generation appears to have their unique ideas and way of thinking; that is, generational cohorts tend to exhibit common views or attitudes based on a shared historical and social context (Stewart and Healy 1989). When it comes to issues nested within feminism, generational differences have been found in gender attitudes toward tomboy behavior (Morgan 1998), marital and childrearing roles (Burt and Scott 2002; Cichy et al. 2007), and women’s role in society (Slevin and Wingrove 1983). The purpose of the present study was to further explore attitudes toward feminism between intergenerational groups of family members in the U.S. College students and their older relatives, including parents and grandparents, answered questions about religiosity, political orientation, gender, and education level, based on prior research identifying variables affecting feminist attitudes (e.g., Clifton et al. 1976; Duncan and Agronick 1995; Twenge 1997). For the purposes of this study, feminism was defined as the movement organized around the doctrine that women should have the same economic, social, and political rights as men; attitudes toward feminism were measured by using the Attitudes Toward Feminism and the Women’s Movement (FWM) scale (Fassinger 1994).

Although the current study was conducted in the U.S., and unless otherwise specified the research reviewed below is from studies conducted in the U.S., many similarities can be drawn to other western countries. Whether it be supporting a women to be elected as the president of the U.S., women gaining more positions in parliament throughout Africa, or striving for the right for women to vote in Saudi Arabia, feminism and the feminist movement is apparent across the world.

Generational Differences

When compared to older adults, younger people tend to rate the women’s movement more positively (Huddy et al. 2000). Perhaps the differences between age groups or generational attitudes may best be viewed through the lens of cohort effect, as differences have been found between cohorts in feminist self-identification (Schnittker et al. 2003), rates of activism (Duncan and Stewart 2000), and correlates of feminist identity (Peltola et al. 2004). Using Stewart and Healy’s (1989) model linking development to social events, the age when a social event is experienced (i.e., childhood, early adulthood, mature adulthood, later adulthood) changes the impact of the event. For example, an incident that occurs during childhood will likely affect the child’s values and expectations, but there will be no revision of identity because of this experience. In contrast, if the same incident occurs to someone in late adulthood, Stewart and Healy theorized that some sort of identity transformation might occur, with a revision of life choices. Several studies (e.g., Duncan and Agronick 1995; Zucker and Stewart 2007) have found support for the interaction between stage of development and time of event, as social events become particularly meaningful when they coincided with early adulthood versus other life stages. For example, an individual’s attitudes towards feminism may be influenced based on one’s stage of life when the first woman was elected to the U.S. Senate (1932), voted as a state governor (1975), nominated for the U.S. Supreme Court justice (1981), appointed the U.S. Secretary of State (1996), served as the U.S. National Security Advisor (2001), or a serious presidential candidate (2008).

Lyons et al. (2005) identified specific values that are associated with different generations, but recommended that gender and cohort never be isolated from one another. Although views on gender equality, beliefs about women’s rights, and endorsement of feminist values apply to men as well, females consistently participate more in the women’s movement (Huddy et al. 2000) and feel more positive about feminist labels (Jacobson 1979). However, men are becoming more profeminist over time (Mason and Lu 1988), and a smaller gender difference on views relating to the rights of women was noted in the mid-1990s compared to a historical sample twenty years previously (Loo and Thorpe 1998).

Religion and feminism appear to be inextricably intertwined, with increased levels of religiosity associated with traditional attitudes toward women (Etaugh 1975), family values (Blanchard-Fields et al. 2001), gender traditionalism (Read 2003), and attitudes toward the women’s movement (Lottes and Kuriloff 1992; Tavris 1973) and women’s work (Sevim 2006). Religion and politics have also been found to be related to attitudes about women’s roles (Bryant 2003). Although group identification, rather than specific political affiliation, may most impact views on feminism (Rhodebeck 1996), liberal political orientation are more likely to identify as feminist (McCabe 2005; Peltola et al. 2004).

In addition, increased educational attainment is generally associated with more liberal views (McCabe 2005; Thornton and Freedman 1979), higher feminist consciousness (Reingold and Foust 1998), less traditional gender role attitudes (Marks et al. 2009), and less likelihood to ascribe to traditional family values (Blanchard-Fields et al. 2001; Willits and Funk 1989). However, some researchers (e.g., McCabe 2005; Schnittker et al. 2003) have found this relationship only with extreme differences in the education levels.

Most studies exploring feminism have focused on college students, often concluding that attending college tends to have a liberalizing effect on individuals in several domains related to gender attitudes (Bryant 2003.) However, attitudes towards gender roles also appear to be influenced by the education of family members, (Weinberg et al. 1997), including the education of both mothers and fathers (Thornton et al. 1983). Looking at multiple generations within a family, high education was a significant predictor of feminist self identity across cohorts (Peltola et al. 2004).

Family Influence

General trends in age differences of attitudes appear to be consistent across familial generations, as parents tend to be more traditional than their offspring (Thornton et al. 1983). In Britain, adolescent girls were found to have the most nontraditional attitudes when compared to other family members (Burt and Scott 2002), and in the U.S. the greatest difference was reported between mothers and daughters (Cichy et al. 2007).

Similar to the cohort effect, Moen et al. (1997) used status attainment (e.g., social class, education) to explore the family socialization processes. They noted a life course perspective, “which focuses on trajectories and transitions in roles and relationships and places them in historical and cultural contexts” (p. 283). This theory appears to be particularly useful when the focus of the research is on a construct that involves a complex interplay between unique individual experiences (e.g., different upbringings in gendered environments, educational attainment) and broader societal events (e.g., the women’s movement, presidential elections).

Mookherjee (1995) found that college students’ attitudes toward women were strongly related to their mothers’ religion and education. Using the life course perspective, Moen et al. (1997) examined the transmission of two gender attitudes (gender role ideology and work role identity) from mothers to daughters. In a longitudinal study, they found that mothers in the 1950s holding egalitarian gender attitudes were more likely to have daughters in the late 1980s with similar egalitarian beliefs. Highly educated mothers are associated with daughters who endorse less traditional feminine values, while the opposite is found with highly educated fathers (Hitlin 2006). With respect to the effect of family socialization on religion and gender attitudes, Thornton et al. (1983) found that gender-role attitudes in 18-year-olds were affected by mother’s religious identification.

In one of the first three-generational family studies, Slevin and Wingrove (1983) explored sex-role attitudes, finding that with each younger generation attitudes became more liberal, with larger differences between grandmothers and granddaughters, than between mothers and daughters. Within three generations of female college students, their mothers, and their maternal grandmothers, students were the most liberal in their attitudes towards women, grandmothers were the least liberal, and mothers scored in between these two groups (Dambrot et al. 1984).

Miller and Glass (1989) explored attitude similarity in grandparents, parents, and grandchildren for religious, political, and gender variables. Their findings suggested that attitudes diverged over time in the older parent-child dyad (i.e., grandparents and parents), while attitudes remained stable between the younger dyad (i.e., parents and grandchildren). The authors attributed these findings to period effects (e.g., women’s movement) and developmental trajectories (i.e., maybe attitudes begin to diverge in the adult child’s midlife). Sabatier and Lannegrand-Willems (2005) investigated the transmission of values in French families and found an indirect influence of grandparents on grandchildren. In other words, mothers directly influenced their children, while grandmothers indirectly impacted grandchildren by influencing mothers.

Also including three generations, Glass et al. (1986) discovered that when social status was controlled for, grandparents predicted parents’ scores on three attitude scales (gender, religion, politics) less well than parents’ predicted college-aged children’s scores. Looking at the younger dyad in more detail, parents’ views on religious and political ideology were more predictive for children than their attitudes reflected on a gender scale. Attitudes about gender were still predictive, but less so than the other variables. In fact, it appeared that gender ideology was “upwardly transmitted” throughout the generations, with parents being influenced by children in both dyads. Unfortunately, Glass and his colleagues (1986) did not investigate the influence of grandparents on grandchildren.

Purpose of the Present Study

The purpose of the present study was to investigate intergenerational attitudes toward feminism in a related younger and older adult population. Group differences were explored on variables (e.g., religion, gender, education, political affiliation, attitudes toward feminism) that have been found to be related to attitudes toward feminism in younger adults and older adults in previous studies (e.g., Clifton et al. 1976; Duncan and Agronick 1995; Twenge 1997). The study also sought to explore what variables were most predictive for each age group, something which has not been previously empirically identified. In addition, this study investigated the importance of family socialization on attitudes regarding the feminist movement, and in particular which family members (e.g., parents, grandparents) exerted the most influence on their young relative. The following hypotheses were investigated:

  1. Hypothesis 1

    Older adults will be more conservative on the feminism, religion, and political orientation measures than younger adults; overall, men will be more conservative than women.

  2. Hypothesis 2

    Religiosity, political orientation, and gender will be predictive of attitudes towards feminism in younger adults; religiosity, political orientation, gender, and education will be predictive of attitudes towards feminism in older adults.

  3. Hypothesis 3

    Including older family members’ responses on attitudes towards feminism, religiosity, political orientation and gender will increase the prediction of younger adults’ attitudes towards feminism.

  4. Hypothesis 4

    Parents’ attitudes towards feminism, religiosity, political orientation, and their education level will be better predictors of the young adults’ attitudes towards feminism than grandparents’ beliefs.

Method

Participants

The sample was obtained as part of a larger research project investigating feminism in college aged adults (N = 374). The current study included 245 undergraduate college students who participated to earn credits for their Introduction to Psychology course at a U.S. large western university, and who were selected for this study because they were linked to an older relative. The young adults were predominantly women (65.7%) and ranged in age from 18 to 26 (mean = 19.0, SD = 1.3). The majority of older adult relatives were also women (73.9%) and ranged in age from 50 to 87 (mean = 64.4, SD = 10.5). One hundred and six relatives identified themselves as parents (68 mothers, 38 fathers), and 139 self-identified as grandparents (114 grandmothers, 25 grandfathers). The majority of participants from both age groups identified as being White, Non-Hispanic. The older adult sample tended to be well educated (42% had a Bachelor’s degree).

Instruments

A similar questionnaire packet was used for both the younger and older adults. The packet contained a demographic questionnaire, a religiosity scale, a political orientation scale, and the FWM.

Demographic Questionnaire

A series of questions was similar for both the younger and older adults, including gender, age, ethnicity, and education. For the older adults, information about their familial connection to the student was added (i.e., mother, father, grandmother, grandfather). Educational span was a continuous variable, with seven options ranging from elementary school to graduate degree.

Religiosity

This construct was measured using eight items adapted from the Springfield Religiousness Scale (SRS), assessing the influence and experience of faith for each individual (Koenig et al. 1988). These items incorporated seven items from Hoge’s (1972) Intrinsic Religious Motivation Scale. This scale asks participants to rate their feelings about concepts such as faith, religion, God, and the Divine. An overall score was derived ranging from 1 (very religious) to 7 (not at all religious). The final Cronbach alpha for the present study was α = .94 for the younger adults, and α = .93 for the older adults.

Political Orientation

Participants were asked to define their political beliefs out of three options. These were coded as 1 for “more liberal” and 2 as “more conservative”.

Attitudes Toward Feminism

The Attitudes toward Feminism and the Women’s Movement (FWM) scale provided a brief measure of affective attitudes toward the feminist movement (Fassinger 1994). Borne out of measurement inadequacy in the field (Fassinger 1994), the FWM measures profeminist attitudes and favorability toward the women’s movement. It does not measure personal gender role characteristics or dogmatism; nor is the FWM susceptible to social desirability influences. Given the expected range of the participants’ age in the present study (e.g., college students to grandparents), as well as the need to mail the instruments to older family members, the FWM was selected based on the measure’s brevity, ability to assess a range of feminist positions, and straightforward assessment (Fassinger 1994).

Participants responded to 10-items on a 7-point Likert scale ranging from “strongly disagree” to “not sure/no opinion” to “strongly agree.” A high score on the FWM represents favorable attitudes toward feminism. The Cronbach alpha for the present study was α = .87 for the younger adults, and α = .88 for the older adults.

Procedure

College students enrolled in introductory psychology courses signed up to participate in the present study for course credit. After completing their own packet during class, students addressed an envelope provided by the researchers and signed their name to a cover letter be sent to the older relative. A code number originally assigned to the young adult packet was written on the questionnaire to the older adult, ensuring that the information would be linked, but would remain anonymous. The cover letter, questionnaires, and a stamped return envelope was sent to the older adult family member identified by the student. The return rate was 73% for older adult packets.

Results

Table 1 provides average responses and standard deviations for the FWM, religiosity, and political orientation. On the FWM, all groups scored near the “not sure” or middle range of the scale. Overall, younger adults described themselves as less religious and more liberal in comparison to their older adult relatives. Pearson correlations for each gender by age group are displayed in Table 2. Results suggested that scores on the FWM were related to political orientation for both ages and genders.

Table 1 Summary of average response and standard deviation on the attitudes toward Feminism and the Women’s Movement (FWM) scale, religiosity, and political orientation
Full size table
Table 2 Pearson correlations between measures for each age group by gender
Full size table

Hypothesis 1: Are Older Adults More Conservative than Younger Adults on the Measures of Feminism, Religiosity, and Political Orientation?

To explore attitudes about feminism, younger adults scores on the FWM (M = 4.58, SD = .86) were compared to their related older adults scores (M = 4.33, SD = .98) using a two-tailed paired t-test, with an alpha value of .05. The comparison was statistically significant (t, 170 = 2.61, p < .05); younger adults endorsed a significantly higher level of feminism. An independent samples t-test was used to compare younger adult women’s (M = 4.80, SD = .82) and older adult women’s (M = 4.41, SD = 1.00) attitudes towards feminism. The comparison was statistically significant (t, 335 = 3.87, p < .05), indicating that younger women endorsed more favorable attitudes towards feminism. With respect to men, the comparison of younger adults (M = 4.25, SD = .85) to older adults (M = 4.38, SD = 1.05) was not significant (t, 144 = −.85, p > .05), and the mean values were not in the expected direction.

Within each age group, scores on the FWM were compared between genders. The comparison of young adult women (M = 4.80, SD = .82) to young adult men (M = 4.25, SD = .85) was statistically significant (t, 242 = 4.92, p < .05), indicating that younger women endorsed more favorable attitudes towards feminism than their male peers. In contrast, older adult women (M = 4.41, SD = 1.00) were not significantly different from older adult men (M = 4.38, SD = 1.05; (t(240) = 2.10, p > .05).

Scores were also compared to investigate levels of religiosity (non-transformed scale scores were used for these analyses). Younger adults scores (M = 2.87, SD = .93) were compared to older adults scores (M = 2.12, SD = .92) using a two-tailed paired t-test, with an alpha value of .05. The comparison was statistically significant (t, 174 = 8.64, p < .05); younger adults described themselves as less religious than their older adult relatives. To investigate further, an independent samples t-test was used to compare younger adult women’s (M = 2.81, SD = .97) and older adult women’s (M = 2.00, SD = .82) religiosity. The comparison was statistically significant (t, 338 = 8.27, p < .05), demonstrating that younger women considered themselves to be less religious. With respect to men, the comparison of younger adults (M = 2.90, SD = .98) to older adults (M = 2.46, SD = 1.03) was also significant (t, 145 = 2.56, p < .05) and showed the same pattern as found in younger women. Thus, both younger women and men described themselves as less religious when compared to older adults of both genders.

Within each age group, scores on the religiosity scale were compared between genders. The comparison of young adult women (M = 2.81, SD = .97) to young adult men (M = 2.90, SD = .98) was not statistically significant (t, 243 = −.61, p > .05). In contrast, older adult women (M = 2.00, SD = .82) were significantly more religious than older adult men (M = 2.46, SD = 1.03; (t, 240) = −3.54, p < .05).

Political orientation was also found to be statistically different between the two age groups. Using a two-tailed paired t-test, with an alpha value of .05, younger adults scores (M = 2.11, SD = 1.03) were compared to older adults scores (M = 2.12, SD = .92). The comparison was statistically significant (t, 170 = −3.43, p < .05). There were three options for political orientation; selecting a higher number indicated a leaning towards conservative views. Thus, older adults endorsed a significantly higher level of political conservatism; both means fell closest to the middle option on the scale, “neither”. An independent samples t-test was used to compare younger adult women’s (M = 2.07, SD = 1.05) and older adult women’s (M = 2.36, SD = .92) political orientation. The comparison was statistically significant (t, 335 = −2.80, p < .05), demonstrating that younger women considered themselves to be more liberal politically. With respect to men, the comparison of younger adults (M = 2.15, SD = .95) to older adults (M = 2.27, SD = .93) was not significant (t, 144 = −.76, p > .05).

Political orientation scores were also compared between genders within each age group. The comparison of young adult women (M = 2.07, SD = 1.05) to young adult men (M = 2.15, SD = .95) was not statistically significant (t, 243 = −.63, p > .05). Similarly, the political orientation of older adult women (M = 2.36, SD = .92) was not significantly different from that of older adult men (M = 2.27, SD = .93); (t, 236 = .70, p > .05).

Hypothesis 2: What Predicts Attitudes Towards Feminism in Younger Adults? What Predicts Attitudes Towards Feminism in Older Adults? Is There a Gender Difference?

Hierarchical linear regression was used to test whether religiosity and political orientation were predictive of attitudes towards feminism in younger adults above and beyond gender; and then again using data collected from the older adults. None of the results for the Tolerance or the Variance Inflation Factor (VIF) were above 1 in any of the regressions; therefore, multicollinearity was not an issue. For each of the regression analyses, Step 1 was always the variable of focus.

As presented in Table 3, gender and political orientation were the statistically significant predictors of attitudes towards feminism in the regression analysis, partially supporting the hypothesis for younger adults. More positive attitudes toward feminism were associated with being female and having a more liberal political orientation. All variables entered in the regression equation accounted for 17% of the variance; F(3, 240) = 16.2, p < .01. For the older adults (Table 3), religiosity and political orientation were the statistically significant predictors of attitudes towards feminism in the regression analysis, thus partially supporting the hypothesis for older adults. More favorable attitudes toward feminism were associated with lower levels of religiosity endorsement and having a more liberal political orientation. All variables entered in the regression equation accounted for 21% of the variance; F(4, 223) = 14.7, p < .01.

Table 3 Hierarchical regression prediction of attitudes towards feminism in younger and older adults
Full size table

When the same hierarchical linear regressions were completed separately for each age group and gender, there was no difference from the results presented above.

Hypothesis 3: Does Adding Older Family Members’ Responses of Attitudes Towards Feminism, Religiosity, Political Orientation and Their Gender Add Additional Prediction for Young Adults’ Attitudes Towards Feminism?

Hierarchical linear regression was used to investigate whether within a family dyad certain older adult relatives’ variables (attitudes towards feminism, religiosity, political orientation, gender, education, relationship) would predict above and beyond the younger adults’ variables (religiosity, political orientation, gender) for their attitudes toward feminism. As noted in Table 4, adding the older adult relatives’ variables did not increase prediction; thus, Hypothesis 3 was not supported. Instead, similar results as Hypothesis 2 were found; more positive attitudes toward feminism were associated with being female and having a more liberal political orientation. All variables entered in the regression equation accounted for 17% of the variance; F(9, 152) = 3.49, p < .01.

Table 4 Hierarchical regression prediction of attitudes towards feminism in younger adults, including older adult variables
Full size table

Hypothesis 4: Are Parents’ Variables Better Predictors than Grandparents?

Two hierarchical linear regressions were conducted to investigate Hypothesis 4. First, the contribution of parental influence was explored by determining whether parental variables (attitudes towards feminism, religiosity, political orientation, gender, education, relationship, gender) predicted young adult attitudes towards feminism above and beyond their own variables (religiosity, political orientation, gender). As shown by Table 5, younger adult gender and political orientation were the statistically significant predictors of attitudes towards feminism in the regression analysis. Adding the parental variables did not increase prediction; rather, it decreased predictive ability. Thus, this part of hypothesis 4 was not supported. Again, the same results as Hypothesis 2 were found; more favorable attitudes toward feminism were associated with being female and having a more liberal political orientation. All variables entered in the regression equation accounted for 20% of the variance; F(9, 88) = 3.16, p < .01.

Table 5 Hierarchical regression prediction of attitudes towards feminism in younger adults, including parental variables
Full size table

An analysis was conducted to investigate whether mothers, in particular, predicted young adult attitudes towards feminism above and beyond their own variables. No variables were found to be significant predictors in the final step (Table 6). All variables entered in the regression equation accounted for 25% of the variance; F(7, 55) = 2.67, p < .05.

Table 6 Hierarchical regression prediction of attitudes towards feminism in younger adults, including mother variables
Full size table

Secondly, to explore whether grandparents’ variables contributed to predictive ability above and beyond the young adults’ variables, another hierarchical regression was conducted. The same variables were entered into the model; for grandparents these were attitudes towards feminism, religiosity, political orientation, gender, education, relationship and gender. The young adult variables were religiosity, political orientation and gender. As reflected in Table 7, younger adult gender and political orientation were the statistically significant predictors of attitudes towards feminism in the regression analysis. Adding the grandparents’ variables did not increase prediction; this part of hypothesis 4 was also not supported. The same results as Hypothesis 2 were found; more favorable attitudes toward feminism were associated with being female and having a more liberal political orientation. All variables entered in the regression equation accounted for 16% of the variance; F(8, 118) = 2.89, p < .01.

Table 7 Hierarchical regression prediction of attitudes towards feminism in younger adults, including grandparent variables
Full size table

Again, to explore further, an analysis was conducted to investigate whether grandmothers, in particular, predicted young adult attitudes towards feminism above and beyond their own variables. Results indicated that only the young adult’s gender was found to be a significant predictor in the final step (Table 8). All variables entered in the regression equation accounted for 21% of the variance; F(7, 95) = 3.58, p < .01.

Table 8 Hierarchical regression prediction of attitudes towards feminism in younger adults, including grandmother variables
Full size table

Given that an upwards transmission of values from children to older adults has been seen in the past, an exploratory analysis was conducted to determine if young adults’ information would add predictive influence to older adults in the present study. First, a hierarchical regression was conducted to investigate whether young adults’ variables contributed to the ability to predict parent’s attitudes towards feminism above and beyond the older adults’ variables. The young adult variables entered into the model were religiosity, political orientation, gender, and attitudes towards feminism. For parents, the variables entered were religiosity, political orientation, gender, and education. As shown in Table 9, parents’ religiosity, political orientation, and education were significant predictors. Younger adults’ political orientation was also a statistically significant predictor of parental attitudes towards feminism in the regression analysis. Specifically, higher levels of parental attitudes toward feminism were associated with lower levels of parental religiosity endorsement, a more liberal political orientation, higher education, and having a child with more liberal political orientation. All variables entered in the regression equation accounted for 56% of the variance; F(8, 89) = 14.13, p < .01.

Table 9 Hierarchical regression prediction of attitudes towards feminism in parents, including younger adult variables
Full size table

Secondly, a hierarchical regression was conducted to investigate whether young adults’ variables contributed to the ability to predict grandparent’s attitudes towards feminism above and beyond the older adults’ variables. The young adult variables entered into the model were religiosity, political orientation, gender, and attitudes towards feminism. For grandparents, the variables entered were religiosity, political orientation, gender, and education. As presented in Table 10, grandparents’ political orientation and younger adults’ political orientation were both statistically significant predictors of grandparental attitudes towards feminism in the regression analysis. Specifically, higher levels of grandparental attitudes toward feminism were associated with a more liberal political orientation in both the grandparent and the grandchild. All variables entered in the regression equation accounted for 17% of the variance; F(8, 118) = 2.97, p < .01.

Table 10 Hierarchical regression prediction of attitudes towards feminism in grandparents, including younger adult variables
Full size table

Discussion

Consistent with previous research (Huddy et al. 2000; Nelson 1988; Spence and Helmreich 1972; Thornton et al. 1983; Truett 1992), results demonstrated that when compared to their older adult relatives, the U.S. college students in the present study were more liberal in their political orientation, religiosity, and attitudes towards women. It appears that regardless of gender, the younger adults were more willing to agree with positive responses about feminism and the women’s movement.

Given that generations likely internalize the women’s movement differently based on their developmental stage when events occur (Zucker and Stewart 2007), it makes sense that younger adults might be socialized to align with statements that promote gender equality. Although current young adults in the U.S. did not experience the women’s movement as did their older relatives, they have grown up in a country that is likely more accepting of women’s rights than ever before. In addition, based on their current age, older adults likely experienced varying aspects of the women’s movement if they were closer to age 50, or 87 years of age.

Similar to previous research (e.g., Dambrot et al. 1984; Thornton et al. 1983), young women in the present study were more liberal in their attitudes towards feminism than the older women. No such generational gap was found between younger and older men. A gender gap was only found between younger adult men and younger adult women; younger women in this study endorsed significantly more positive views towards feminism and the women’s movement. This is in line with the literature that finds men consistently endorse less liberal beliefs about attitudes towards women than their female counterparts (Loo and Thorpe 1998; Nelson 1988); yet it was only found in younger men in this study. It may be that because younger women are more liberal than any other group, the gender difference is only found in this young cohort and between women. This may be due to the opportunity (e.g., education, occupational prospects) offered U.S. younger women compared to previous cohorts. Previous studies that did not tease apart age might have missed generational differences, particularly between older and younger women.

Younger and older women cohorts consistently differed in regard to attitudes toward feminism, religiosity, and political orientation; the two generations of men differed only on religiosity. As presented in Fig. 1, the younger female cohort is much different from the older one, especially when compared to the generational differences between men.

Fig. 1
figure 1

Differences between the feminism, religiosity, and political orientation measures in older and younger adult men and women (The arrow indicates increasing liberalism)

Full size image

One explanation for the visible differences may be the extent to which society has changed in the U.S. over the past few decades for women. For example, in 1950 only 36% of women 25 years and older had a high school degree or higher; in 2000, this number jumped to 81% (United States Census Bureau (2006). Similarly, women in the U.S. also participated more in the workforce. In 1950 40% of women (aged 16 and up) worked outside of the home; this number increased to 60% in 1998 (Bureau of Labor Statistics 2000).

Family Influence

When exploring the influence between the family generations, older adults’ attitudes and demographic information was not associated with younger adults’ attitudes towards feminism and the women’s movement. In contrast, young adults’ political views were associated with older adults’ attitudes towards feminism; the younger adults’ gender and political orientation were consistently the strongest indicators of whether one would endorse positive views towards feminism and the women’s movement.

Although previous research (e.g., Eisenberg 1988; Mookherjee 1995; Roberto and Stroes 1992) has found mothers and grandmothers to be particularly influential on their children and grandchildren in the U.S., this was not found to be the case in the present study. These results may suggest that family does not impact young adults’ attitudes towards feminism and the women’s movement. Or perhaps the nature of college life offers an insular environment in which students begin to question and adopt new views from those in their family upbringing. A vast amount of research has been conducted on the liberalizing effect of college courses and women’s studies classes in particular (e.g., Aronson 2003; Bryant 2003; Stake 2007). Using the life course perspective as a model to understand this dynamic, perhaps it is this intersection between unique individual experiences (e.g., college experiences, gender, peers) and broader societal events (e.g., political climate, women in the news) that contributes to the lack of family influence seen. Perhaps there is a more salient or different type of influence when experiences draw more from the family (e.g., such as in high school, or if the student is living at home while attending college) than from the college environment.

Are Younger Adults’ Views Associated with Older Adults’ Views?

Most developmental theories assume a one-direction model of child socialization from parent to child. A small body of literature highlights the importance of considering family influence from a bidirectional point of view. Kuczynski et al. (1997) outlined a bidirectional model that suggested that children are active agents who have a considerable amount of power to influence their parents’ internalization of attitudes and values. More specifically, Glass et al. (1986) found “upward transmission” of gender ideology from children to parents. Likewise, similar findings on child influence have been discovered with attitudes towards cohabitation (Axinn and Thornton 1993), religion, and technology (Pinquart and Silbereisen 2004).

In the present study, adding older adults’ information did not help predict younger adults’ attitudes towards feminism. In fact, contrary to expectations, the reverse was found. Specifically, higher levels of parental attitudes toward feminism were predicted by parental attitudes (i.e., lower levels of religiosity endorsement, more liberal political orientation, higher education) and having a child with more liberal political orientation. A similar finding was also found for grandparents, as higher levels of grandparental attitudes toward feminism were associated with a more liberal political orientation in both the grandparent and the grandchild.

Thus, results suggested that the presence of a liberal young adult in the family contributes to more positive attitudes towards feminism in older adults. Because this is a cross-sectional study, it cannot be determined whether the young adults influence the older adults’ attitudes; but this does suggest that liberal young adults are associated with more liberal attitudes in parents and grandparents.

From a life course perspective, perhaps older adults’ unique experiences, such as having a politically liberal child, might be intersecting in a powerful way with current social events, such as having a woman campaign seriously for president. Pinquart and Silbereisen (2004) noted the transmission of values is affected by the salience of topics and the motivation to discuss issues. It should be noted that data collection for the present study began in February of 2007, a time when presidential campaigning in the U.S. was becoming increasingly relevant. Although there is no way to be sure given the design of the current study, one explanation might be that politics and women’s issues became more prominent, perhaps creating more family discussions during the time frame of this study. In 1997, the fairly radical possibility of a woman or African-American being elected as the President of the United States could have conceivably led to discussions in families where more liberal young adults shared their views with parents and grandparents.

It should also be noted that U.S. society has become steadily more liberal over the past few decades. Thus, the young adults in the present study were living during a time in U.S. history more liberal than the time period their parents or grandparents were born, potentially allowing more room for older adults to move in a liberal direction (and be influenced by their younger relatives) than there is for younger adults to be influenced by older adults. In other words, there might be less difference between younger adults’ attitudes towards feminism and “societies” attitudes towards feminism, which would have allowed for less predictive ability to be seen from older relatives.

Limitations

Because of the cross-sectional design, it is impossible to determine whether familial attitudes and beliefs influenced others; rather this methodology allowed for important predictors to be examined without causal information. In addition, no data was collected regarding the reason a young adult chose to send the survey to a particular family member; thus, there may be something unique about those family members who were chosen and/or returned the surveys (e.g., reliability, knowledge of college course requirements, salience of relationship to student). Likewise, the college students who chose to complete this study about feminism may be distinct from the general student population.

Another limitation of the present study included offering only three levels for the political affiliation. Offering additional options or utilizing a 7-point Likert scale might have allowed for more accurate representation of participant political affiliation.

Future Research

The present study provided a snapshot of younger and older U.S. family members’ attitudes towards feminism and the women’s movement. Future research may capitalize on longitudinal designs to better determine potential influence of family members. Obtaining additional information such as level of emotional closeness, amount of time spent together, or discussion topics could further elucidate the concept of family influence. Given that older adult variables did not add predictive value of the views toward feminism in the college-aged participants in the study, it would be interesting to conduct a similar study with similar aged young adults who are not attending college, or with younger participants, such as high school students. This may assist in continuing to identify the influence of family, compared to a college environment, on a young adult’s view toward feminism.

The notion of religiosity could be expanded upon in future studies to include a spirituality component. Both younger and older adults may resonate more with this term or concept, which could potentially lead to additional information regarding the impact of faith on attitudes towards feminism. Future research should also attempt to include participants with a broader range of educational attainment, as this might allow for the association between education and attitudes towards feminism that have been found in previous research.

In addition, many of the above recommendations should be explored internationally, to better understand worldwide generational differences of views regarding the feminist movement. Taking a closer look at attitudes across generational cohorts, with an eye towards the significant historical events and social context within individual countries, may contribute to the overall understanding of attitudes towards feminism and the women’s movement, and the influence of religiosity, political affiliation, gender, education, and familial relationship.

Lastly, a qualitative component in future research about attitudes toward feminism would allow for participants to speak to what the concept means to them in their current lives and country of residence. This would allow for the idea of global feminism and the women’s movement to be continually re-defined for a contemporary understanding.