Gender and Homosexuality Attitudes across Religious Groups from the 1970s to 2014: Similarity, Distinction, and Adaptation

Landon Schnabel1 Indiana University, Bloomington, USA

This study uses General Social Survey data to compare gender and homosexuality across American religious groups from the 1970s to 2014, examining three possible patterns for how evangelical attitudes relate to those of other groups: (1) they are similar; (2) they are different, but move together over time; (3) they are different and converge or diverge over time. Evangelical gender attitudes regarding work and family issues are more conservative than those of all other groups, but are adaptive to broad trends, changing at a rate similar to those of other groups. Evangelical attitudes toward the morality of homosexuality and same-sex marriage are more conservative than those of all other religious groups, and their rate of change is slower over time. Separate trends on the two issues suggest that gender and sexuality attitude change is decoupled, especially among evangelicals who are adapting more on gender while increasingly distinguishing themselves on same-sex relationships. A three-stage process of religious tension appears to characterize evangelical identity-building: (1) similarity, (2) distinction, and (3) adaptation.

Key Words: Public Opinion; Attitudes; Religion; Gender Roles; Same-Sex Marriage; Morality of Homosexuality; Symbolic Boundaries; Culture; Identity

This study has been published as: Schnabel, Landon. 2016. “Gender and Homosexuality Attitudes across Religious Groups from the 1970s to 2014: Similarity, Distinction, and Adaptation.” Social Science Research 55(1): 31-47. doi:10.1016/j.ssresearch.2015.09.012

1 The author owes special thanks to Clem Brooks, Youngjoo Cha, and especially Brian Powell for frequent support on this project. He would also like to thank Cate Taylor, Lisa Miller, Emily Wurgler, Ariel Sincoff-Yedid, fellow participants in the summer cohort writing group, fellow participants in the S700 writing workshop, and the anonymous reviewers for insightful comments on earlier drafts. This paper was presented at the 2014 annual meetings of Sociologists for Women in Society and the Society for the Scientific Study of Religion. Direct correspondence to Landon Schnabel, Department of Sociology, Indiana University, 744 Ballantine Hall, 1020 E. Kirkwood Ave., Bloomington, IN 47405. Email: [email protected].

Should women be supported in the workplace that evangelical Protestant gender attitudes are and hold civil leadership positions? Should more conservative relative to other groups, but gay men and lesbians have the right to get have changed at rates similar to other married? Gender and sexuality are contentious groups—in fact, there has been no clear topics in both private and public debate, differentiation in the rate of gender attitude sparking conflict in American families, change across religious groups since 1994. religious communities, and national politics. Likewise, evangelical Protestant attitudes Examining beliefs about gender and sexuality toward same-sex relationships are more across religious groups provides an conservative relative to other groups, but have opportunity to consider not only attitude diverged to become even more distinct over change, but also how social groups distinguish time. themselves from one another (Edgell 2012, Lamont and Molnár 2002). Gender and BACKGROUND sexuality are both central for boundary drawing among religious groups and often Symbolic Boundaries and American seem to operate in tandem (Tranby and Evangelicalism Zulkowski 2012), but do American religious In the United States, religion is a particularly groups reveal similar patterns for attitude strong cultural power for shaping beliefs about change on both of these issues? Is it possible gender and sexuality (Tranby and Zulkowski that those in conservative Protestant 2012). Culture influences attitudes and action denominations are adapting more on one issue through ideas, symbols, and metaphors that while distinguishing themselves more on the function as tools in a “cultural toolkit” other? (Swidler 1986). People draw upon these To investigate the questions at hand, I cultural tools to construct boundaries between use General Social Survey (GSS) data and a themselves and others that demarcate their work and family focused gender attitude scale membership in a worthy and moral group (Cotter et al. 2011) to first compare overall (Edgell, Gerteis and Hartmann 2006, Swidler religious group differences and then compare 1986). In this construction of boundaries, they rates of attitude change over time. Following draw distinctions between the worthy in-group the examination of gender attitudes, I compare and the out-group of “others” (Edgell, Gerteis them to attitudes toward the morality of and Hartmann 2006, Edgell 2012, Tranby and homosexuality and same-sex marriage. Zulkowski 2012). These boundaries are moral Though I examine and discuss all major and symbolic rather than ethnic or material, American religious groups, I focus on and thus are labeled symbolic boundaries evangelicals because American evangelicalism (Edgell, Gerteis and Hartmann 2006, Lamont exemplifies the drawing of symbolic and Molnár 2002, Tranby and Zulkowski boundaries—they use moral stances to 2012). Lamont and Molnar (2002) defined distinguish the people in their group from symbolic boundaries as conceptual distinctions those outside their group (Smith et al. 1998, made to categorize objects, practices, people— Tranby and Zulkowski 2012). These and even time and space—by which examinations will reveal whether conservative individuals and groups grapple with and religion slows movement toward equality in eventually agree upon shared definitions of the realms of gender and sexuality, and reality. Symbolic boundaries come to take on a whether the liberalization of gender attitudes “real” character, generating collective and sexuality attitudes follow similar or identities and separating people into distinct different patterns. The findings demonstrate groups. 1

Sociological research on evangelicals light of this slowing of gender attitude highlights how they seek to be both in the liberalization and lack of change in other world (adaptation), but not of the world measures of gender equality—such as (distinction) (Bartkowski and Read 2003, occupational segregation (Charles and Grusky Gallagher and Smith 1999, Heath 2003, 2005)—sociologists have suggested that Tranby and Zulkowski 2012). Symbolic movement toward gender equality is uneven, boundary drawing is central to evangelicals’ has slowed, and may even have stalled (Cotter, strength because they thrive with a sense of Hermsen and Vanneman 2011, England 2010, embattlement with secular society even as they Gerson 2010). adapt to it (Smith et al. 1998). This strength is Cotter, Hermsen, and Vanneman the type that arises in strict churches that are in (2011) argued that a neo-traditional frame of tension with the outside world and thus reduce “egalitarian essentialism” had arisen. Such a free-riding and foster commitment and vitality frame is new, according to Cotter et al. (2011), (Iannaccone 1994, Stark and Finke 2000). because it combines some aspects of feminist Gender and sexuality attitudes are particularly equality—which distinguishes it from the important areas for the drawing of symbolic older “separate spheres” frame—with boundaries, especially for evangelicals traditional familism. Egalitarian essentialism (Bartkowski 2001, Edgell 2006, Gallagher asserts that women and men have equal but 2003 2004b, Heath 2003, Hoffman and essentially different natures that make them Bartkowski 2008, Tranby and Zulkowski more or less skilled at different roles. In this 2012). frame—which is highly consistent with ideas Though evangelicals have constructed surrounding intensive mothering and the marked symbolic boundaries between “opting out” narrative1 (Stone 2007)—women themselves and other groups—drawing on are viewed as natural caretakers who are gender and sexuality as important sites to inherently skilled at nurturing. emphasize cultural distinctions to show that Correspondingly, men are viewed as they are not like the rest of the world— particularly able to produce in the workforce evangelical attitudes do seem to be adaptive to and provide for their families. In sum, this social change and broad cultural trends (Edgell frame’s justification of gender roles moves 2006, Wilcox 2004). Attitude change hints at away from explicit subordination to potential shifts in the drawing of symbolic assumptions about different propensities and boundaries. The present study’s comparative skills, and mothering is spoken about as examination of rates of change in gender and having value equal to, or even greater than, sexuality attitudes across religious groups will paid labor. Cotter and colleagues (2011) assert help us to better understand not just attitude that this frame—by combining conflicting change on these relevant social issues, but elements of feminism and traditional whether and how evangelical symbolic familism—facilitates a return to traditional boundaries are adapting to a changing world. gender roles while denying implications of a lower valuing of women. Gender Attitudes Such frames could be most prominent among religious groups that promote American gender attitudes surrounding issues complementarian theology. Complementarian of work and family changed rapidly through theology argues that, having been created to the early-1990s, but appeared to stall in the complement one another, women and men are mid-1990s, and were no more egalitarian in essentially different and suited to different 2008 than they were in in the early 1990s roles. Evangelical religious leaders developed (Cotter, Hermsen and Vanneman 2011). In 2

the current version of complementarianism— know that religion influences gender attitudes, God assigned women and men different, but but we do not know how that influence may equal, roles2 that complement one another—in change over time (but see Petersen and the 1970s in direct response to second-wave Donnenwerth 1998 for an older study on feminism (Schnabel 2015). By combining attitude change). some elements of feminism—women and men are of equal worth—with complementary (i.e., Attitudes toward Same-Sex Relationships traditional) gender roles, this religious Previous research shows that sexuality, like response to feminism sought to maintain gender, is an important area in which religious men’s leadership role and women’s people draw symbolic boundaries (Olson, submissive role while denying claims that this Cadge and Harrison 2006, Tranby and means that women and men are unequal. If Zulkowski 2012). Rather than examine all evangelical religious leaders were successful available sexuality attitudes, this paper in imparting this perspective to everyday examines particularly relevant attitudes— evangelicals, there should be different patterns beliefs about homosexuality and same-sex of attitude change among evangelicals than marriage. In America, religion is central to among other religious groups. Research on public debate surrounding same-sex marriage gender and American religion, however, and is particularly salient in the shaping of shows that the rhetoric of religious leaders and attitudes toward homosexuality (Adamczyk the everyday beliefs and practices of religious and Pitt 2009). Though attitudes toward the adherents are often only loosely coupled morality of homosexuality and same-sex (Denton 2004, Edgell 2006, Gallagher 2003). marriage did not start changing as early as did It is possible, therefore, that everyday gender attitudes, they now appear to be evangelicals have not followed the traditional consistently liberalizing to the apparent elite discourse, and have liberalized along with dismay of many conservative religious leaders other Americans. and pundits (Baunach 2011, Loftus 2001, Previous Research on Religion and American Powell et al. 2010, Sherkat et al. 2011). Gender Attitudes Petersen and Donnenwerth (1998) compared gender attitudes and support for gay Survey research on the relationship between rights from the early 1970s to the early 1990s, religion and gender attitudes in America finds which is just prior to the stall in gender that conservative religion is linked with attitude change noted by Cotter et al. (2011). conservative gender attitudes and that Using the General Social Survey, they found nonreligion is associated with particularly that all Christian groups liberalized on gay egalitarian attitudes (Brinkerhoff and MacKie rights and gender attitudes from the 1970s to 1984, Brinkerhoff and MacKie 1985, Burn and 1993 (Petersen and Donnenwerth 1998). Busso 2005, Moore and Vanneman 2003, During this same period—in which all Peek, Lowe and Williams 1991). Not only are Christian groups liberalized on limited gay conservative Protestants less gender egalitarian rights—American beliefs on the morality of than the nonreligious, but qualitative research homosexuality did not liberalize (Loftus on a Catholic, Black Protestant, and LGBTQ 2001). Analyzing the General Social Survey inclusive mainline Protestant congregations through 2002, Linneman (2004) found that has found surprisingly high levels of conservative Christians’ views on the morality affirmation for traditional gender roles in these of homosexuality changed more slowly than more progressive congregations (Edgell and people who were not conservative Christians, Docka 2007). Based on previous research, we but conservative Christians’ attitudes toward 3

basic gay rights were changing at a similar rate 2012). Similarly, Bean and Martinez (2014) to other Americans. He argued that found that while many active evangelicals still conservative Christians perceived tension and view gay men and lesbians as immoral, some hostility from lesbians, gay men, and of them are open to the idea of civil unions. American society in general over LBGTQ Nevertheless, conservative religion is still one issues, believed that their conservative of the most important predictors for opposition Christian values were marginalized, and to same-sex marriage in America (Powell et al. therefore changed their views toward the 2010). morality of homosexuality more slowly. In a study focused on California, Lewis and The Relationship between Gender and Gossett (2008) used news polls to examine Same-Sex Relationship Attitudes attitude change on same-sex marriage. They Much of the literature tends to suggest that found that Protestant’s attitudes changed more beliefs and attitudes about gender roles and slowly than those of other religious groups. same-sex relationships map closely onto one In more recent years, the General another (McVeigh and Diaz 2009, Tranby and Social Survey has included a measure of Zulkowski 2012), are inseparable parts of support for same-sex marriage. Not only is polarized worldviews (Luker 2006), and religion an important factor for understanding operate in tandem (Kenneavy 2012). The American attitudes toward same-sex marriage, relationship, however, may not be as but Olson et al. (2006) found that it has a straightforward as it seems. stronger influence than sociodemographic In some cases, gender and sexuality characteristics and Perry (Forthcoming) found attitudes appear to shift differently over time. that conservative religious identity predicts Whereas a large part of gender attitude change stronger opposition to same-sex marriage than can be explained by generational change belief in inerrancy, belief in creationism, and (Brooks and Bolzendahl 2004, Pampel 2011, frequency of bible reading. Sherkat and Schnittker, Freese and Powell 2003), Baunach colleagues (Sherkat, de Vries and Creek 2010, (2011) showed that much of the recent rapid Sherkat et al. 2011) have used the same-sex rise in support for same-sex marriage is due to marriage measure to show that religion is a intracohort change—or people changing their strong predictor for negative views of views over time—rather than younger homosexuality and opposition to same-sex 3 generations replacing older generations. Other marriage, and that the influence of religion studies have discussed how gender and may be increasing over time. Similarly, sexuality attitudes are heterogeneous, Baunach (2011, 2012) found that opposition to suggesting that they are related, but in same-sex marriage was higher among all complex ways (Bolzendahl and Myers 2004, groups in 1988, and that groups—such as Schnittker, Freese and Powell 2003). One evangelicals, Republicans, Southerners, and recent study explicitly argued that value issues Blacks—began to become more distinct from are not always aligned, using abortion and gay other Americans over time. rights attitudes to highlight attitude Other research, however, indicates that asynchrony (Dillon 2014). Finally, some evangelical attitudes are changing more research has suggested that future studies quickly than suggested by the aforementioned might be able to uncover a decoupling of rates literature. Conservative religious groups are of change in gender and sexuality attitudes now more willing to extend limited civil rights across religious groups (Powell et al. 2010). In (e.g., freedom of speech) to gay men and other words, one set of beliefs could change lesbians than they were in the past (Kenneavy more quickly than the other even in the 4

presence of a continued correlation between DATA, MEASURES, AND METHODS the attitudes, and the relative rates of change could differ across religious groups. Data

Expectations I use General Social Survey (GSS) data to analyze gender attitude and sexuality attitude This study uses the General Social Survey to change across religious categories. The NORC examine change in gender and sexuality General Social Survey is a probability-based attitudes over time across religious groups to face-to-face interview sampling of non- assess whether they reveal similar or divergent institutionalized U.S. adults 18 and older, trends. Three competing patterns of attitude fielded from 1972 through 2014. Following change are possible and will be considered, recent work on gender attitude change (Cotter, first for gender attitudes and then for sexuality Hermsen and Vanneman 2011), I examine four attitudes: key gender attitude measures4 that have been consistently asked from 1977 to 2014 (1) Evangelical attitudes could be similar to (N=21,435).5 Separate samples are used for those of other groups. In this case, corresponding analyses of attitudes toward the evangelical attitudes would be just as morality of homosexuality (asked from 1973 egalitarian as those of other groups, and to 2014; N=29,983) and support for same-sex their attitudes would move along with marriage measures (asked in 1988 and then other groups to maintain this similarity. again from 2004 to 2014; N=8,785). (2) Evangelicals could have more conservative attitudes overall, but move along with Dependent Variables everyone else. In this case, their attitudes would be distinctive at individual points in Gender Attitudes Scale time, but adaptive to broad social forces This study uses the same standardized gender over time, changing just as rapidly as those attitudes scale—constructed by standardizing of other religious groups. and then summing scores for each of four (3) Evangelical attitudes could be distinctive gender attitude measures (Cronbach’s α = and change at a different rate than that of 0.70)—used by Cotter et al. (2011). Using the other groups. If evangelical attitudes same scale allows for comparison with, and change more slowly than other groups, extension of, Cotter and colleagues’ findings. they would become increasingly different Three items ask the respondent whether and over time. how strongly s/he agrees (strongly disagree=1 to strongly agree=4) with the following Gender and sexuality attitudes could fit statements: the same pattern, or they could follow different  “A working mother can establish just as patterns. If evangelicals have adapted their warm and secure a relationship with her attitudes more in one area and come to children as a mother who does not distinguish themselves more in the other, it work.” would not only suggest a decoupling of change  “A preschool child is likely to suffer if in gender and sexuality attitudes, but also a his or her mother works.” remapping of symbolic boundaries—or how  “It is much better for everyone involved evangelicals are distinguishing themselves if the man is the achiever outside the from the broader culture. home and the woman takes care of the home and family.”

5

explore trends in distinction and adaptation. These are coded so that higher values are more The scheme uses self-reported denominational egalitarian. The other measure is dichotomous, affiliation to place respondents into seven asking the respondents whether they agree categories based on the historical development (disagree=1) with the following statement: of religious traditions in the United States: evangelical Protestant, Black Protestant,  “Most men are better suited emotionally mainline Protestant, Catholic, no affiliation, for politics than are most women.” Jewish, and “other faith.”

Taken together, these four variables provide a Year good measure—the best available for To compare gender attitude change over time, comparing attitude change from the 1970s to I use a best-fitting spline function with “knots” the present—of a respondent’s level of gender (or hinges) at the same locations—1994 and essentialism and her/his attitudes about gender 2000—used by Cotter et al. (2011). A spline roles, especially in regard to work and family function splices lines—specifically, it joins issues. together two or more lines with different slopes. The slopes, “hinged” at statistically Morality of Homosexuality and Support for driven locations, are then estimated by Same-Sex Marriage transforming year so that each period between To compare gender attitudes with attitude the hinges can be interpreted as a linear term change toward gay and lesbian Americans, I for that time range.7 As Cotter et al. (2011) use a measure of attitudes toward the morality showed, American gender attitudes “hinge” at of homosexuality and a measure of support for 1994 and 2000. As presented in the models, same-sex marriage.6 Fielded consistently since the period coefficients measure the linear 1973, the morality of homosexuality question change during the indicated interval of time— asks respondents: “What about sexual relations Year Spline (1977 to 1994) measures attitude between two adults of the same sex—do you change from 1977 to 1994. Religious groups think it is (1) always wrong, (2) almost always are interacted with period terms to compare wrong, (3) wrong only sometimes, or (4) not rates of change within intervals of time. wrong at all?” Fielded in 1988 and then again The same-sex relationship attitudes do from 2004 to 2014, the support for same-sex not follow the same rise, decline, and marriage question asks whether respondents subsequent rise pattern of the gender attitudes. (1) strongly disagree to (5) strongly agree with For these attitudes, Bayesian Information the following statement: “Homosexual couples Criterion (BIC) scores prefer the simpler should have the right to marry one another.” model with a linear interaction term for year and religious groups rather than interaction [Table 1 here] terms for spline intervals and religious groups (Raftery 1995). Whereas the pattern of gender Key Independent Variables attitude change required a spline function, a straightforward linear term for year is Religious Affiliation appropriate for sexuality attitude change, Religious affiliation is measured using the already suggesting that the general population Steensland et al. (2000) religious follows a different attitude change pattern on 8 categorization scheme (RELTRAD), which these attitudes. Religious groups are allows for meaningful comparisons between interacted with year to compare change from evangelicals and other religious groups to 6

1973 to 2014 for morality of homosexuality of change during three periods of time. and from 1988 to 2014 for same-sex marriage. Subsequently, I present the parallel analyses of attitudes toward the morality of homosexuality Demographic Controls and support for same-sex marriage. Finally, I present a linear analysis of gender attitudes for This study takes into account key demographic the sake of easier comparison with the controls that may affect gender and sexuality sexuality attitude analysis, and to examine attitudes. These items were selected based on whether there has been a consistent gender theory and what is standard in the literature attitude liberalization trend since the start of (Bartkowski and Hempel 2009, Cotter, the “stall” noted by Cotter et al. (2011). Hermsen and Vanneman 2011). These measures were then tested using BIC scores to RESULTS determine whether they should be included (Raftery 1995). Controls include sex Gender Attitudes (female=1), race (white=1, Black=2, other=3),9 age (in years), parental status Gender Attitudes across Religious Groups (parent=1), marital status (married=1), female Table 2 presents religious group affiliation work status (coding explained below), region 13 (South=1), education (dummies for highest effects with just controls for survey year degree attained), and family income (inflation- (Model 1) and then affiliation effects with adjusted to 2000 and expressed in tens-of- demographic controls added (Model 2). thousands of dollars). Female work status is Evangelical is the reference category. In coded 1 if a female respondent works full time Model 1, all religious categories are more and if a married male respondent indicates that egalitarian than evangelicals (p < .001). For his spouse works full time.10 example, the unaffiliated (b = .352, p < .001) are half a standard deviation of the gender Analytic Strategy attitude scale more egalitarian than evangelicals. Evangelical attitudes are clearly Following Cotter et al. (2011), this study uses distinctive, being more conservative than those ordinary least squares (OLS) regression for the of all other groups. 11 gender attitudes scale. Ordinal logistic regression is used for the morality of [Table 2 here] homosexuality and support for same-sex marriage measures. All outcomes are In the spline function, the 1977 to 1994 estimated using Huber/White sandwich robust interval measures attitude change during this 12 standard errors. All analyses are unweighted. time period. Similarly, 1994 to 2000 and 2000 Sensitivity analyses including weights yielded to 2014 measure change during these time substantively equivalent results. periods. Gender attitudes have become more I first compare religious group main egalitarian since the 1970s except for a period effects on gender attitudes. Second, I present from 1994 to 2000 when they became less figures that show gender attitude change egalitarian (b = -.025, p < .001). across groups based on separate models for Model 2, which includes demographic each year with centered covariates. I then controls, shows that, with one exception, the include all years in the same model (the religious affiliation differences are not primary method as described above) and attributable to sociodemographic differences. interact religious groups with best-fitting All but the “other faith” category are still more spline function interval terms to compare rates egalitarian (p < .001) than evangelicals. 7

Consistent with the literature, women are more Gender Attitudes across Religious Groups egalitarian than men (b = .200, p < .001). The over Time gender attitudes of Black Americans do not Figure 1 shows gender attitudes from 1977 to differ significantly from whites.14 Those in the 2014 for all Americans and then for Protestant “other race” category are less egalitarian than religious groups, controlling for demographics whites (b = -.186, p < .001). People who are with centered covariates. These plotted values older, married, or from the South are less are based on separate OLS regression models egalitarian (p < .001), whereas parents, (not shown) for each year with centered working women or men married to working controls (the same sociodemographic measures women, people with more education, and those included in Tables 2 through 4).16 We see in with a higher income are more egalitarian (p < Figure 1a that the overall gender attitude trend .001).15 The 2000 to 2014 year interval shows was rapid change in the 1970s and 1980s with that attitudes are currently liberalizing and that decline in the 1990s. From 2000 onward, this change is not attributable to attitudes have become increasingly egalitarian. sociodemographic shifts (b = .020, p < .001). As noted by Cotter et al. (2011), the attitudes Affiliation effects decrease with the of all Americans in 2008 had not risen much inclusion of demographic controls, but they above the level in the early 1990s. Challenging persist and can be compared to demographic the idea that American gender attitudes have characteristics. Being an evangelical, when stagnated, however, in 2010, in 2012, and compared to those with no religious affiliation again in 2014 Americans’ attitudes became (b = .249, p < .001) or Jews (b = .313, p < more egalitarian than ever before. .001), has a larger effect than gender (b = .200, p < .001). In fact, being evangelical, in [Figure 1 here] comparison to the most liberal groups, has a larger effect than all demographics except Figure 1b shows that evangelical having a bachelor’s degree (b = .272, p < attitudes have followed the same general trend .001)—about the same magnitude as being of all Americans, liberalizing until the mid- evangelical compared to the unaffiliated and 1990s, declining through 2000, and then rising Jews—or having a graduate degree (b = .413, again since 2000. Evangelicals started out p < .001). closer to the average in the late 1970s, In sum, evangelicals are the least diverged in the late 1980s, and have generally egalitarian religious group and religion is a paralleled average American attitude change stronger predictor of gender attitudes than all since then. Mainline Protestants have also measures but education, suggesting that gender followed the same general pattern, as revealed attitudes are an important field for drawing in Figure 1c. Mainline Protestants, however, symbolic boundaries and expressing have remained much closer to the average, evangelical collective identity. This still leaves diverging from evangelicals in the late 1980s. open the question of whether the first or third Black Protestants also followed a similar possible expectation fits the data: are pattern, though with more year-to-year evangelical attitudes changing more slowly volatility because they make up a smaller than other groups (expectation 1), or are they portion of the sample.17 All Protestant groups adapting to broad social change at the same followed the same trend, paralleling the rate as other groups (expectation 3)? average for all Americans. They liberalized quickly from the late 1970s to the mid-1990s, when they became more traditional. Since

8

2000, their attitudes have become more time period in Model 2 show that from 1977 to egalitarian. 1994 mainline Protestant (b = .009, p < .05) Figure 2 shows the gender attitudes by and Catholic (b = .010, p < .01) attitudes year for Catholics, the unaffiliated, Jews, and liberalized more rapidly than those of those in the “other faith” category. Catholic evangelicals. The interaction of religious attitudes, like those of mainline Protestants, group and the 1994 to 2000 time period are similar to the average and move in demonstrates that evangelical decline in conjunction with the general population. The egalitarianism in this period did not differ unaffiliated are generally more egalitarian than from that of other Christian groups or the the American average, but again follow the unaffiliated. The “other faith” (b = .042, p < same trend. The Jewish and “other faith” .01) category operated differently, but that is categories are very small, and thus unstable because, as shown in Figure 2, they did not and not particularly meaningful.18 That said, follow the same pattern as other groups. The they do not seem follow the same gender interaction effects for the interval from 2000 to attitude change trend as other Americans. Jews 2014 show that evangelicals have liberalized start out quite egalitarian and are typically just as quickly as other Christian groups and more egalitarian than other groups, without a the unaffiliated since 2000. In fact, before clear pattern of change over time. The “other sociodemographic controls are taken into faith” category contains disparate groups and account, evangelical attitudes actually changed thus their trends cannot be meaningfully more quickly than Catholic attitudes (b = .009, interpreted. In sum, Figures 1 and 2 show that p < .05). When controls are included in the most religious groups followed a similar model, just the disparate “other faith” category upward, downward, and then upward trend. is different from evangelicals (b = -.021, p < [Figure 2 about here] .05). By including all years in the same Although one might expect evangelical models presented in Table 2 and interacting attitude change to be slower than that of other religious groups with time periods, we can groups based on the rhetoric of evangelical more explicitly test rates of change during the public figures, Table 3 showed that move toward egalitarianism from 1977 to evangelical attitudes changed just as quickly— 1994, during the decline from 1994 to 2000, upward from 1977 to 1994, downward from and during the resurgence of gender attitude 1994 to 2000, and upward again from 2000 change from 2000 to 2014. Table 3 presents onward—as those of other Christians and the coefficients for the effect of year and period on unaffiliated. The only noteworthy difference gender attitudes with evangelical as the was that mainline Protestants and Catholics reference category. These values represent the liberalized more quickly from 1977 to 1994. overall rate of group change during the move Since then, evangelical attitudes have toward egalitarianism from 1977 to 1994, the paralleled those of other Christian groups and decline from 1994 to 2000, and the resurgence the unaffiliated. In sum, gender attitude change of attitude change from 2000 onward. has not stagnated and evangelical attitudes are changing at a rate similar to that of other [Table 3 here] groups. Gender attitudes follow the second possible pattern set forth in the expectations: Model 1 (without controls) and Model similar rates of change but continued 2 (with controls) of Table 3 reveal similar rates difference; or, in other words, adaptation over of attitude change over time after 1994. time with distinction at any given point in Interaction effects for religious groups and time. 9

Attitudes toward Same-Sex Relationships [Figure 3 here] The gender attitudes of all Christian groups [Figure 4 here] and the unaffiliated moved in parallel to one another. But did attitudes toward gay men and Table 4 shows overall religious group lesbians follow a similar pattern? Figures 3 differences and then differences in rates of and 4 present the yearly means for all change over time on attitudes toward Americans, the unaffiliated, and evangelicals homosexuality and same-sex marriage. on attitudes toward the morality of Evangelical is the reference category. With homosexuality and same-sex marriage. On only a control for year in Model 1, belief in morality, we see that attitudes are evangelicals are more conservative (p < .001) relatively stagnant from the early 1970s to the than all but Black Protestants on the morality early 1990s. In the mid-1990s, a continuing of homosexuality. When all controls are upward trend toward accepting same-sex included in Model 2, evangelicals are more relationships as moral starts among all conservative than all other religious groups (p Americans, the same period in which gender < .001), including Black Protestants. On attitudes declined. Evangelicals and the support for same-sex marriage, evangelicals unaffiliated follow a similar trend: stable are more conservative than all religious groups attitudes through the mid-1990s and then an (p < .001), both without (Model 4) and with upward trend. The evangelical shift, however, controls (Model 5). Similar patterns show up appears to be slower than that of the across both outcome measures, with the religiously unaffiliated and the general unaffiliated and Jewish categories revealing population. Evangelicals start out closer to the the largest difference from evangelicals. Year American average when most people were is significant (p < .001) for all models, opposed to same-sex relationships, but as revealing a liberalizing trend over time. society liberalized evangelical attitudes Evangelical attitudes toward same-sex diverge. Mainline Protestants and Catholics relationships—like gender attitudes—are (not shown) closely follow the American 19 distinctive from those of other groups, but are average. they as adaptive as gender attitudes to the Although fewer years are available, we overall liberalizing trend? see a similar pattern on support for same-sex marriage: slower change early on and more [Table 4 here] rapid change in recent years. Evangelical attitudes again start out closer to the average Interaction terms for provide an and then appear to be changing more slowly effective way to test group rates of attitude than those of others. This trend already change. Models 3 and 6 on Table 4 show that suggests both that gender and sexuality both evangelicals and Black Protestants reveal followed clearly different patterns overall, and slower change than other groups on the that attitudes toward same-sex relationships morality of homosexuality and support for may be becoming more important for same-sex marriage. On the morality of expressing a distinctive evangelical collective homosexuality, mainline Protestants (b = .012, identity in opposition to secular society. The p < .001), Catholics (b = .019, p < .001), the uptick for evangelical support for same-sex unaffiliated (b = .013, p < .001), and those in marriage in 2014, however, may indicate that the “other faith” category (b = .022, p < .001) what has happened in the past could shift in liberalized more quickly than evangelicals. the future. Catholics (b = .021, p < .01) and the 10

unaffiliated (b = .040, p < .001) have moved to groups. Table 5 shows the effect of religious accept same-sex marriage more quickly than affiliation on gender attitudes over time with a have evangelicals. In sensitivity analyses linear term for year—instead of the spline examining just 2004 to 2014, the coefficients function shown in Table 3—to make it more for the religious group and year are similar on easily comparable to the same-sex relationship views of the morality of homosexuality, but attitudes analyses. Observations from 1994 to smaller on same-sex marriage. In other words, 2014, the period of time that evangelical evangelicals’ views are continuing to lag on attitudes changed at the same rate as other the morality of homosexuality at a similar rate groups in the spline function, are included. as in the past, but in recent years their attitudes The table shows evangelical attitudes changed toward same-sex marriage are trending toward just as quickly as those of other groups from changing at the same rate as other groups. 1994 to 2014, confirming that gender and Overall, these sexuality attitudes sexuality attitudes follow a different pattern of followed the first possible pattern discussed in change across religious groups.20 The simple the expectations: distinctive differences and year term also allows us to see that American slower attitude change among evangelicals. gender attitudes have not stagnated. In spite of Gender attitudes and sexuality attitudes follow the decline in attitudes from 1994 to 2000 different patterns of change across American shown in the spline function, American religious groups, suggesting that change in attitudes have liberalized significantly since these attitudes is decoupled for some groups. 1994.21 Evangelical attitudes toward same-sex relationships are becoming increasingly [Table 5 here] distinctive, whereas their gender attitudes are more adaptive, paralleling the national trend. DISCUSSION AND CONCLUSION Therefore, attitudes toward same-sex relationships seem to be becoming more This study used General Social Survey data to important for drawing symbolic boundaries compare attitudes toward gender roles and and the expression of collective evangelical same-sex relationships across American identity. The 2014 evangelical uptick in religious groups from the 1970s to 2014, support for same-sex marriage, however, may considering three possible patterns for how indicate that this trend could shift in the future. evangelical attitudes relate to those of other groups: (1) they are similar; (2) they are Further Analysis Confirming Different different, but move together over time; (3) Change Patterns they are different and converge or diverge over time. The results show that gender attitudes fit Figures 1 through 4 showed that the gender the second pattern, and same-sex relationship and sexuality attitudes of the general attitudes the third. Evangelical gender attitudes population have followed different change regarding work and family issues are more patterns. Illustratively, gender attitudes conservative than those of all other groups, but changed more quickly than sexuality attitudes are adaptive to broad trends, changing at a rate until 1994. From 1994 to 2000, sexuality similar to those of other groups since 1994. attitudes liberalized while gender attitudes Evangelical attitudes toward the morality of became more conservative. Since 2000, both homosexuality and same-sex marriage are also types of attitudes have been changing quickly. more conservative than those of all other Further regression analyses confirm religious groups. Their rate of change, that gender and sexuality attitudes have however, is slower than other religious groups, followed different patterns across religious 11

meaning that they are diverging over time. the third possible pattern of more conservative Separate patterns on the two issues suggest attitudes and a distinctly slower rate of change. that gender and sexuality attitude change is Gender and sexuality attitude change decoupled, especially among evangelicals who fit different patterns. Attitudes toward same- are adapting more on gender while sex relationships began to liberalize in the increasingly distinguishing themselves on mid-1990s, the same time that gender attitudes same-sex relationships. declined, indicating a decoupling of change— Gender attitudes differ across religious this is not to say that they are not still related, groups at any point in time, but all Christian just that they are changing differently.22 groups and the unaffiliated have followed the Beyond the decoupling of change, groups same change pattern: liberalization in the late- differ in their location on the conservative- 1970s to mid-1990s, a shift toward liberal spectrum on gender and sexuality traditionalism from the mid-1990s to early attitudes. For example, Black Protestants were 2000s, and then an upward trend until 2014. just as gender egalitarian as mainline Mainline Protestant and Catholic gender Protestants and Catholics, but their attitudes attitudes did liberalize more quickly than toward same-sex relationships were more evangelicals from 1977 to 1994, which conservative. The distinctive patterns in rates suggests that gender roles used to be a more of attitude change suggest an increasing important site for distinction. But evangelicals importance of sexuality attitudes for the did not differ from other Christian groups or drawing of symbolic boundaries. Evangelicals the unaffiliated in their rate of attitude decline now seem to be defining and distinguishing from 1994 to 2000 or in their rate of attitude themselves more on their views toward same- resurgence from 2000 to 2014. Religious sex relationships than gender roles. affiliation does not seem to affect gender This study shows that religious attitude change in more recent years—though affiliation cannot be used to explain the gender evangelicals are still less gender egalitarian attitude decline in the mid-1990s noted by than other groups. Therefore, on gender Cotter et al. (2011) and others. Nevertheless, it attitudes, evangelicals followed the second is possible that complementarian religious possible pattern with continued distinction, but ideas diffused across other groups. Figure 1 rates of change similar to those of other showed that evangelical and mainline groups. Protestants diverged from one another in the The examination of attitudes toward late 1980s, which hints at the importance of same-sex relationships revealed different the rise of the Christian Right and especially change patterns. Attitudes on the morality of the prominence of the Moral Majority at that sexuality began liberalizing quickly at the time (Schnabel 2013). In the mid-1990s same time that gender attitudes moved back through 2000, both the non-denominational toward traditionalism (based on the trends Promise Keeper’s Christian men’s movement across available years, support for same-sex and the Christian Coalition were at their peak. marriage likely followed a similar pattern). 2000 was a particularly important turning Whereas evangelical gender attitudes moved point in the decline of both movements for a in parallel to those of other Christian groups variety of sociopolitical reasons (Bartkowski and the unaffiliated, their attitudes on the 2004; Schnabel 2013). Although these groups morality of homosexuality and support for were most popular among evangelicals, they same-sex marriage—along with those of Black appealed to social conservatives across Protestants—changed more slowly. Therefore, religious divides. The popularity of these on sexuality attitudes, evangelicals followed groups at the same time that gender attitudes 12

became more traditional suggests a gender, work, and family. When women had connection; this popularity could also be started joining the workforce but most indicative of a general essentialist backlash in Americans were still opposed to same-sex conjunction with the rise of intensive relationships, evangelicals distinguished mothering, which combines a rhetoric of themselves more on conservative gender roles choice and equality with an idealization of and less on views toward same-sex stay-at-home mothering (Cotter, Hermsen and relationships. Today, however, evangelical Vanneman 2011, Faludi 1991, Stone 2007). leaders can support working women and Indicative of the rising popularity of these women politicians without facing much ideas, one of the foundational books for boundary policing from evangelicals in the attachment parenting and intensive mothering pews (consider evangelical support for Sarah came out just before the attitude decline (Sears Palin), but would face backlash from some and Sears 1993). segments of evangelicalism for coming out in The results also demonstrate that favor of same-sex marriage. Research suggests gender attitudes have not “stalled” as proposed that evangelical affiliation and evangelical by Cotter et al. (2011), whose data went politics, of which same-sex marriage is a key through 2008.23 Rather than stagnating, issue, are so closely tied that people will shift American gender attitudes have liberalized their religious identity because of their consistently both before and after a short political identity (Hout and Fischer 2002, Hout period of decline. Since 2008, we have seen and Fischer 2014). Slower change on views of the highest levels of egalitarianism yet. same-sex relationships, therefore, could come Current liberalization, similar to past change, both from evangelicals changing their views is happening at similar rates across religious more slowly and the disaffiliation of some groups. Rather than trying to understand why evangelicals with more liberal views on same- gender attitude change has slowed or stalled, sex relationships. the question future research may want to Things may be changing, however. address is what caused gender attitudes to Following the nationwide legalization of become more traditional while attitudes same-sex marriage, the tension between toward same-sex relationships continued to traditional evangelical views on same-sex liberalize. relationships and cultural norms of equality, Why are evangelical attitudes toward especially among younger people, may same-sex relationships changing more slowly, become too strong for evangelical identity to and what might happen in the future? There continue to be defined by opposition to same- are a few possibilities, but one likely sex relationships. The uptick in support for explanation is that theology, history, and same-sex marriage among evangelicals in politics have linked conservative Christian 2014 and recent news about key evangelical identity and opposition to same-sex marriage. leaders, such as Tony Campolo and David Evangelical “traditional values” are currently Neff, coming out in favor of same-sex in more tension with the larger society on marriage suggest that such a transition may same-sex relationships than on work and happen sooner rather than later. Just as gender family roles for women and men. There is a roles became a less important factor for clear and widely shared evangelical view on expressing and policing evangelical identity, same-sex relationships rooted in evangelical opposition to same-sex marriage could also politics and literal interpretations of the Bible, become a less important factor and eventually but it would now be harder to determine and be replaced in evangelical political rhetoric by define any one shared evangelical position on other issues. For now, however, sexuality 13

remains a key defining issue for American affiliation has a stronger or weaker effect on religions as demonstrated by public responses issues less commonly debated in religious to same-sex marriage legalization and related terms—such as marijuana and vaccination. debates about religious liberty. Religious affiliation remains important The results of this study suggest a for understanding Americans’ gender and three-stage process of religious tension and sexuality attitudes, but does not seem evangelical identity building that should be particularly important for understanding rates further tested in future research: (1) similarity, of gender attitude change. Religion, however, (2) distinction, and (3) adaptation. When all of continues to be one of the most important society agrees that something is wrong, factors keeping some Americans from evangelicals are less likely to use that issue to supporting the recent nationwide legalization distinguish themselves—for example, of same-sex marriage. Through social change, evangelical attitudes toward gender evangelicalism continues its pursuit of egalitarianism in 1977 and their attitudes distinction—to be in but not of the world toward same-sex relationships through the (Smith et al. 1998, Stark and Finke 2000)—by early 1990s were similar to the rest of recreating a changing but still distinctive American society because most Americans collective identity. Just as there was a shift agreed that these were unacceptable. When an from gender roles to same-sex relationships as issue becomes contested, however, key sites for symbolic boundary drawing, the evangelicals tend to hold the “traditional” future could reveal a shift from same-sex position and use it to distinguish themselves. relationships to other sites of religious tension When the vast majority of society and identity building. subsequently shifts and the issue is no longer widely contested—as happened with slavery, racial segregation, interracial marriage, women ENDOTES joining the workforce, and now same-sex marriage (Schnabel 2015)—the tension may 1 The opting out narrative suggests that women become too great and American evangelicals should and do leave professional careers when then shift along with the broader society and they have children because they feel naturally distinguish themselves on newly contested driven to be full-time caregivers. issues. 2 The theological literature promoting this In light of the decoupling of rates of perspective uses the term “role” to make this change in gender and sexuality attitudes point. Furthermore, much of the gender among evangelicals and Black Protestants, attitude literature also uses the term (Cotter, future work could consider the relationship Hermsen and Vanneman 2011). Therefore, I between affiliation and attitudes on other use it at times for convenience and issues of sexuality (i.e., sexual education, comparability while recognizing that it has pornography, pre- and extra-marital sex, been problematized as presenting an overly cohabitation, and abortion) and personhood static view of gender. (i.e., birth control, abortion, physician-assisted 3 Generational change does still account for a suicide, and stem cell research—(Schnabel and significant portion of the change in same-sex Breitwieser 2015; Schnabel 2014)) that are marriage attitudes (Sherkat et al. 2011). commonly debated on religious grounds in the 4 There are other repeated questions in the public sphere. Finally, future research could GSS, but the survey stopped asking them in compare the effect of religious affiliation on 1998. other changing attitudes to examine whether 14

5 I do not include the 1982 and 1987 Black knot is reached (in Stata, this is done by using oversamples, and restrict my analyses to the margin option with the mkspline respondents for whom complete data are command). Rather than providing the linear available, losing 10.2% of the total 23,874 for term for year within a period of time, their missing controls, primarily due to the income approach tested the rate of change against the measure. As a sensitivity test, I used a multiple immediately previous interval, limiting their imputation method—Gaussian normal ability to compare the rate of change after regression imputation—to replace missing 2000 to the rate of change before 1994. values and arrived at substantively equivalent 8 I conducted a sensitivity analysis with a results across models. spline function for the morality measure with a 6 There are three other questions in the GSS— knot at 1994 to parallel the gender attitudes as part of a sequence of questions on speech analyses and another analysis with a knot at rights for different groups, such as atheists, 1990 when attitudes toward the morality of Muslims, and communists—that ask about homosexuality began to steadily rise. The whether a gay man should be allowed to speak patterns for group change were substantively in a community, whether he should be able to similar to the linear model. hold a position at a college, and whether a gay- 9 Although the Black Protestant religious friendly book should be allowed to stay at the affiliation category and Black race category public library. These questions, while are correlated at .7, I control for race to ensure interesting and important, are less about that I measure the impact of Black sexuality (and more about free speech) than Protestantism rather than just racial attitude the two questions this paper examines—which differences. Scholars have been concerned are particularly relevant to current political about collinearity when including both Black debate. Furthermore, the rights questions have Protestantism and race in the same model, but limited variability, especially among more a follow-up to the Steensland et al. (2000) liberal religious groups, such as Jews and the religious categorization scheme argues that unaffiliated, who already revealed very high collinearity should not be a problem when support for these minimal rights in the 1970s. using large surveys such as the GSS 7 I used the default mkspline command in Stata (Woodberry et al. 2012). Sensitivity analyses 13 for the spline function with hinges at 1994 not controlling for race reveal substantively and 2000. The first interval in the spline equivalent findings for rates of attitude change function, 1977-1994, is coded as 0 for 1977 over time across religious categories. and increases by one for each year until 1994, 10 I use the same female work status coding which is 17. 17 is then the value for 1994 and method used by Bartkowski and Hempel all following years. The second interval, 1994- (2009). Though this control is not as standard 2000, is 0 until 1994 and then increases by one to the literature as the others, its inclusion is for each year until 2000, which is 6, and 6 is supported by BIC tests. Sensitivity analyses then the value for 2000 and all following without this measure provided substantively years. The third interval, 2000-2014, is coded equivalent results. as 0 for all years until 2000 and then increases 11 I conducted sensitivity analyses using by one for each year until the final year in the ordinal logistic regression with the same sample 2014, where it has a value of 14. Cotter models and the results were substantively et al. (2011) coded their spline function equivalent. variables so that the year term does not stop 12 Winship and Radbill (1994) showed that increasing by one for every year once the next unweighted regression estimates tend to be 15

unbiased and consistent when using large 18 Despite their year-to-year volatility, I present surveys with weights that are solely a function these figures for Jewish and “other faith” of independent—rather than dependent— attitudes to provide context for the interaction variables. effects between year and these groups in Table 13 Sensitivity analyses using BIC tests show 3. that dummy variables for year are very 19 Groups not shown for clarity and parsimony strongly preferred for fit over just a linear term generally parallel the attitudes of one of these for year. However, including the best-fit spline three groups on both the morality of with knots at 1994 and 2000, the method homosexuality and support for same-sex presented, is very strongly preferred by BIC marriage. Jewish attitudes are similar to those tests over year dummy variables. of the unaffiliated, mainline Protestant and 14 A model (not shown) including the Catholic attitudes are similar to the average, demographics, but not the independent and Black Protestant attitudes are similar to variable (religious affiliation), shows that those of evangelicals. Blacks are significantly more egalitarian than 20 When including all years, the faster whites. This effect is no longer significant with liberalization of mainline Protestants and the inclusion of religious categories. Catholics from 1977 to 1994 shows up as a 15 In sensitivity analyses, I included measures significant effect. In a sensitivity analysis, I of religious belief (biblical literalism, which is examined the morality of homosexuality not available for all years) and practice measure including only those cases from 1994 (attendance). More literal views of the Bible to 2014 to ensure that the linear term for and more frequent religious service attendance gender attitudes presented in 5 is comparable are both associated with less egalitarian gender to what is presented in Table 4. In this attitudes. Including these measures in the analysis, evangelical attitudes on the morality models decreases overall group difference, but of homosexuality still changed more slowly leads to substantively equivalent findings for than those of other groups. rates of change over time. 21 A sensitivity analysis including both 16 When using OLS regression to compare attitudes in the same model further confirms groups with centered covariates, the intercept the differing change patterns. Though only is the value of the excluded group and the about half the people were asked both coefficients for other groups can be added to questions due to the years the questions were the intercept for their predicted values. I used asked and subsampling within years, an separate models for each year with religious analysis of the sexuality attitudes examined in groups to get the values for each group and table 4 with controls for gender attitudes— then a model without religious groups and gender attitudes, year binaries, and interaction binary variables for each year to ascertain the terms for these—does not alter the change yearly values for all Americans. patterns across religious groups—evangelical 17 Sensitivity analyses of the figures without sexuality attitudes still change more slowly controls reveal substantively equivalent than those of the other groups. patterns of change over time across groups. 22 Polychoric correlations show that people Without controls, including the one for race, with conservative gender attitudes continue to Black Protestants appear slightly more be more likely to have conservative attitudes egalitarian than the average, but follow the toward same-sex relationships than people same pattern over time. with more liberal gender attitudes.

16

23 This is not to say that Cotter et al. (2011) gender attitudes as stalled. Furthermore, they were wrong. They made it clear that the were careful to only speculate tentatively question mark at the end of their title—“The about attitude change following 2008 (the last End of the Gender Revolution?”— was year in their sample). important even though they did talk about

17

REFERENCES

Adamczyk, Amy and Cassady Pitt. 2009. "Shaping Attitudes About Homosexuality: The Role of Religion and Cultural Context." Social Science Research 38(2):338-51. Bartkowski, John. 2001. Remaking the Godly Marriage: Gender Negotiation in Evangelical Families. New Brunswick, NJ: Rutgers University Press. Bartkowski, John and Jen'nan Ghazal Read. 2003. "Veiled Submission: Gender, Power, and Identity among Evangelical and Muslim Women in the United States." Qualitative Sociology 26(1):71-92. Bartkowski, John. 2004. The Promise Keepers: Servants, Soldiers, and Godly Men. New Brunswick, NJ: Rutgers University Press. Bartkowski, John and Lynn M. Hempel. 2009. "Sex and Gender Traditionalism among Conservative Protestants: Does the Difference Make a Difference?" Journal for the Scientific Study of Religion 48(4):805-16. Baunach, Dawn Michelle. 2011. "Decomposing Trends in Attitudes toward Gay Marriage, 1988- 2006." Social Science Quarterly 92(2):346-63. Baunach, Dawn Michelle. 2012. "Changing Same-Sex Marriage Attitudes in American from 1988 through 2010." Public Opinion Quarterly 76(2):364-78. Bean, Lydia and Brandon C. Martinez. 2014. "Evangelical Ambivalence toward Gays and Lesbians." Sociology of Religion 75(3):395-417. Bolzendahl, Catherine and Daniel J. Myers. 2004. "Feminist Attitudes and Support for Gender Equality: Opinion Change in Women and Men, 1974-1998." Social Forces 83(2):759-90. Brinkerhoff, Merlin B. and Marlene MacKie. 1984. "Religious Denominations' Impact Upon Gender Attitudes: Some Methodological Implications." Review of Religious Research 25(4):365-78. Brinkerhoff, Merlin B. and Marlene MacKie. 1985. "Religion and Gender: A Comparison of Canadian and American Student Attitudes." Journal of Marriage and Family 47(2):415- 29. Brooks, Clem and Catherine Bolzendahl. 2004. "The Transformation of U.S. Gender Role Attitudes: Cohort Replacement, Social-Structural Change, and Ideological Learning." Social Science Research 33(1):106-33. Burn, Shawn Meghan and Julia Busso. 2005. "Ambivalent Sexism, Scriptural Literalism, and Religiosity." Psychology of Women Quarterly 29(4):412-18. Charles, Maria and David B. Grusky. 2005. Occupational Ghettos: The Worldwide Segregation of Women and Men. Standford, CA: Standford University Press. Cotter, David, Joan M. Hermsen and Reeve Vanneman. 2011. "The End of the Gender Revolution? Gender Role Attitudes from 1977 to 2008." American Journal of Sociology 117(1):259-89. Denton, Melinda Lundquist. 2004. "Gender and Marital Decision Making: Negotiating Religious Ideology and Practice." Social Forces 82(3):1151-80. Dillon, Michele. 2014. "Asynchrony in Attitudes toward Abortion and Gay Rights: The Challenge to Values Alignment." Journal for the Scientific Study of Religion 53(1):1-16. Edgell, Penny. 2006. Religion and Family in a Changing Society. Princeton, NJ: Princeton University Press.

18

Edgell, Penny, Joseph Gerteis and Douglas Hartmann. 2006. "Atheists as “Other”: Moral Boundaries and Cultural Membership in American Society." American Sociological Review 71(2):211-34. Edgell, Penny and Danielle Docka. 2007. "Beyond the Nuclear Family? Familism and Gender Ideology in Diverse Religious Communities." Sociological Forum 22(1):25-50. Edgell, Penny. 2012. "A Cultural Sociology of Religion: New Directions." Annual Review of Sociology 38(1):247-65. England, Paula. 2010. "The Gender Revolution: Uneven and Stalled." Gender & Society 24(2):149-66. Faludi, Susan. 1991. Backlash: The Undeclared War against American Women. New York: Crown. Gallagher, Sally and Christian Smith. 1999. "Symbolic Traditionalism and Pragmatic Egalitarianism." Gender & Society 13(2):211-33. Gallagher, Sally K. 2003. Evangelical Identity & Gendered Family Life. New Brunswick, NJ: Rutgers University Press. Gallagher, Sally K. 2004. "Where Are the Antifeminist Evangelicals? Evangelical Identity, Subcultural Location, and Attitudes toward Feminism." Gender & Society 18(4):451-72. Gerson, Kathleen. 2010. The Unfinished Revolution: How a New Generation Is Reshaping Family, Work, and Gender in America. New York: Oxford University Press. Heath, Melanie. 2003. "Soft-Boiled Masculinity: Renegotiating Gender and Racial Ideologies in the Promise Keepers Movement." Gender & Society 17(3):423-44. Hoffman, John P. and John Bartkowski. 2008. "Gender, Religious Tradition and Biblical Literalism." Social Forces 86(3):1245-72. Hout, Michael and Claude S. Fischer. 2002. "Why More Americans Have No Religious Preference: Politics and Generations." American Sociological Review 67(2):165-90. Hout, Michael and Claude S. Fischer. 2014. "Explaining Why More Americans Have No Religious Preference: Political Backlash and Generational Succession, 1987-2012." Sociological Science 1(October):423-47. Iannaccone, Laurence R. 1994. "Why Strict Churches Are Strong." American Journal of Sociology 99(5):1180-211. Kenneavy, Kristin. 2012. "Support for Homosexuals' Civil Liberties: The Influence of Familial Gender Role Attitudes across Religious Denominations." Social Forces 90(4):1347-75. Lamont, Michèle and Virág Molnár. 2002. "The Study of Boundaries in the Social Sciences." Annual Review of Sociology 28:167-95. Lewis, Gregory B. and Charles W. Gossett. 2008. "Changing Public Opinion on Same-Sex Marriage: The Case of California." Politics & Policy 36(1):4-30. Linneman, Thomas J. 2004. "Homophobia and Hostility: Christian Conservative Reactions to the Political and Cultural Progress of Lesbians and Gay Men." Sexuality Research & Social Policy 1(2):56-76. Loftus, Jeni. 2001. "America's Liberalization in Attitudes toward Homosexuality, 1973-1998." American Sociological Review 66(5):762-82. Luker, Kristin. 2006. When Sex Goes to School. New York: Norton. McVeigh, Rory and Maria-Elena D. Diaz. 2009. "Voting to Ban Same-Sex Marriage: Interests, Values, and Communities." American Sociological Review 74(6):891-915. Moore, Laura M. and Reeve Vanneman. 2003. "Context Matters: Effects of the Proportion of Fundamentalists on Gender Attitudes." Social Forces 82(1):115-39. 19

Olson, Laura R., Wendy Cadge and James T. Harrison. 2006. "Religion and Public Opinion About Same-Sex Marriage." Social Science Quarterly 87(2):340-60. Pampel, Fred. 2011. "Cohort Changes in the Socio-Demographic Determinants of Gender Egalitarianism." Social Forces 89(3):961-82. Peek, Charles W., George D. Lowe and L. Susan Williams. 1991. "Gender and God's Word: Another Look at Religious Fundamentalism and Sexism." Social Forces 69:1205-21. Perry, Samuel L. Forthcoming. "Bible Beliefs, Conservative Religious Identity, and Same-Sex Marriage Support: Examining Main and Moderating Effects." Journal for the Scientific Study of Religion. Petersen, Larry R. and Gregory V. Donnenwerth. 1998. "Religion and Declining Support for Traditional Beliefs About Gender Roles and Homosexual Rights." Sociology of Religion 59(4):353-71. Powell, Brian, Catherine Bolzendahl, Claudia Geist and Lala Carr Steelman. 2010. Counted Out: Same-Sex Relations and Americans' Definitions of Family. New York: Russell Sage Foundation. Raftery, Adrian E. 1995. "Bayesian Model Selection in Social Research." Sociological Methodology 25(1):111-63. Schnittker, Jason, Jeremy Freese and Brian Powell. 2003. "Who Are Feminists and What Do They Believe? The Role of Generations." American Sociological Review 68(4):607-22. Sears, William and Martha Sears. 1993. The Baby Book: Everything You Need to Know About Your Baby—from Birth to Age Two. Boston, MA: Little, Brown, and Company. Schnabel, Landon. 2013.”When Fringe Goes Mainstream: A Sociohistorical Content Analysis of the Christian Coalition's Contract with the American Family and the Republican Party Platform.” Politics, Religion & Ideology 14 (1):94-113. Schnabel, Landon. 2014. “The Question of Subjectivity in Three Emerging Feminist Science Studies Frameworks: feminist postcolonial science studies, new feminist materialisms, and queer ecologies.” Women’s Studies International Forum 44 (3):10-16. Schnabel, Landon. 2015. “Toward a Standpoint Hermeneutic: The Case of the Evangelical Gender Subordination Debate.” Claremont Journal of Religion 4 (1):86-112. Schnabel, Landon, Breitwieser, Lindsey. 2015. “Recent Advances in Feminist Science and Technology Studies: Reconceptualizing Subjectivity and Knowledge.” Advances in Gender Research 20(1):43-63. Sherkat, Darren E., Kylan Mattias de Vries and Stacia Creek. 2010. "Race, Religion, and Opposition to Same-Sex Marriage." Social Science Quarterly 91(1):80-98. Sherkat, Darren E., Melissa Powell-Williams, Gregory Maddox and Kylan Mattias de Vries. 2011. "Religion, Politics, and Support for Same-Sex Marriage in the United States, 1988– 2008." Social Science Research 40(1):167-80. Smith, Christian, Michael Emerson, Sally Gallagher, Paul Kennedy and David Sikkink. 1998. American Evangelicalism: Embattled and Thriving. Chicago, IL: University of Chicago Press. Stark, Rodney and Roger Finke. 2000. Acts of Faith: Explaining the Human Side of Religion. Berkeley, CA: University of California Press. Steensland, Brian, Jerry Z. Park, Mark D. Regnerus, Lynn D. Robinson, W. Bradford Wilcox and Robert D. Woodbury. 2000. "The Measure of American Religion: Toward Improving the State of the Art." Social Forces 79(1):291-318.

20

Stone, Pamela. 2007. Opting Out? Why Women Really Quit Careers and Head Home. Berkeley, CA: University of California Press. Swidler, Ann. 1986. "Culture in Action: Symbols and Strategies." American Sociological Review 51(2):273-86. Tranby, Eric and Samanth E. Zulkowski. 2012. "Religion as Cultural Power: The Role of Religion in Influencing Americans' Symbolic Boundaries around Gender and Sexuality." Sociology Compass 6(11):870-82. Wilcox, W. Bradford. 2004. Soft Patriarchs, New Men: How Christianity Shapes Fathers and Husbands. Chicago: University of Chicago Press. Winship, Christopher and Larry Radbill. 1994. "Sampling Weights and Regression Analysis." Sociological Methods & Research 23(2):230-57. Woodberry, Robert D., Jerry Z. Park, Lyman A. Kellstedt, Mark D. Regnerus and Brian Steensland. 2012. "The Measure of American Religious Traditions: Theoretical and Measurement Considerations." Social Forces 91(1):65-73.

21

FIGURES AND TABLES

Figure 1: Gender Attitude Scale over Time: All Americans and Protestants

Figure 1a: All Americans Figure 1b: Evangelicals vs. All

0.6

0.4

0.2

0

-0.2 Egalitarian Egalitarian Attitudes -0.4

-0.6 1975 1980 1985 1990 1995 2000 2005 2010 2015 1975 1980 1985 1990 1995 2000 2005 2010 2015

Figure 1c: Mainline vs. All Figure 1d: Black Protesant vs. All

0.6

0.4

0.2

0

-0.2

Egalitarian Egalitarian Attitudes

-0.4

-0.6Source: General Social Survey, 1977-2012 1975 1980 1985 1990 1995 2000 2005 2010 2015 1975 1980 1985 1990 1995 2000 2005 2010 2015 Source: General Social Survey, 1977-2014 Note: Based on regression analyses with the following centered controls: female, race, age, parental status, marital status, female work status, region, education, and income.

22

Figure 2: Gender Attitude Scale over Time: Catholic, Unaffiliated, Other Faith, and Jewish

Figure 2a: Catholic vs. All Figure 2b: No Affiliation vs. All

0.6

0.4

0.2

0

-0.2 Egalitarian Attitudes Egalitarian -0.4

-0.6 1975 1980 1985 1990 1995 2000 2005 2010 2015 1975 1980 1985 1990 1995 2000 2005 2010 2015

Figure 2c: Other Faith vs. All Figure 2d: Jewish vs. All

0.6

0.4

0.2

0

-0.2 Egalitarian Attitudes Egalitarian -0.4

-0.6Source: General Social Survey, 1977-2012 1975 1980 1985 1990 1995 2000 2005 2010 2015 1975 1980 1985 1990 1995 2000 2005 2010 2015

Source: General Social Survey, 1977-2014 Note: Based on regression analyses with the following centered controls: female, race, age, parental status, marital status, female work status, region, education, and income.

23

Figure 3: Belief in Morality of Homosexuality

4

3.5

3

2.5

2

1.5 Always Wrong=1 to Not Wrong at All=4 at Wrong Not to Wrong=1 Always

1

1996 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1998 2000 2002 2004 2006 2008 2010 2012 2014 All Americans No Affiliation Evangelical

Source: General Social Survey, 1973-2014

24

Figure 4: Support for Same-Sex Marriage

5

4.5

4

3.5

3

2.5

2

1.5 Strongly Oppose=1 to Strongly Strongly to Oppose=1 Support=5 Strongly

1

1996 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1998 2000 2002 2004 2006 2008 2010 2012 2014 All Americans No Affiliation Evangelical

Source: General Social Survey, 1988-2014

25

Table 1: Descriptive Statistics Measures Metric Mean SD Range Dependent Variables Gender Attitudes Scale Standardized and Summated 0.013 0.726 -1.922–1.297 A working mother can establish just as warm and secure a relationship Strongly Disagree=1 to Strongly Agree=4 2.840 0.880 1–4 with her children as a mother who does not work. A preschool child is likely to suffer if his or her mother works. Strongly Agree=1 to Strongly Disagree=4 2.559 0.808 1–4 It is much better for everyone involved if the man is the achiever Strongly Agree=1 to Strongly Disagree=4 2.720 0.857 1–4 outside the home and the woman takes care of the home and family. Most Men are better suited emotionally for politics than are most Disagree=1 0.741 women. Morality of Homosexualitya Always Wrong=1, Almost Always Wrong=2, 1.920 1.294 1-4 Sometimes Wrong=3, Not Wrong at All=4 Support for Same-Sex Marriageb Strongly Disagree=1 to Strongly Agree=5 3.190 1.519 1-5 Religious Category Evangelical Evangelical=1 0.263 Black Protestant Black Protestant=1 0.080 Mainline Mainline=1 0.207 Catholic Catholic=1 0.248 No Affiliation No Affiliation=1 0.128 Jewish Jewish=1 0.020 Other Faith Other Faith=1 0.054 Survey Year Year Spline (1977 to 1994) 1977=0 to 1994=17; All Following Years=17 14.261 4.642 0–17 Year Spline (1994 to 2000) 1977-1994=0 to 2000=6; All Following Years=6 2.922 2.808 0-6 Year Spline (2000 to 2014) 1977-2000=0 to 2014=14 3.050 4.710 0-14 Controls Female Female=1 0.562 Race White White=1 0.811 26

Black Black=1 0.130 Other Other=1 0.059 Age in Decades In Decades 4.535 1.684 1.8–8.9 Parental Status Parent=1 0.725 Marital Status Married=1 0.513 Female Work Status If Female, Working Full Time=1; If Male, Spouse 0.347 Working Full Time=1 South South=1 0.356 Highest Degree Earned Less Than High School Less than High School Diploma=1 0.170 High School High School Diploma=1 0.521 Some College Some College=1 0.066 Bachelor Bachelor’s Degree=1 0.163 Graduate Graduate Degree=1 0.080 Family Income Inflation-Adjusted (2000) Tens-of-Thousands of Dollars 4.640 3.790 0.037–18.039 Source: General Social Survey 1977-2014 N=21,435 a N for Morality of Homosexuality is 29,983 b N for Support for Same-Sex Marriage is 8,785

27

Table 2: Regression of Group Membership Effects on Gender Attitudes Scale Model 1 Model 2 Religious Categories Evangelical — —

Black Protestant 0.219*** 0.149*** (0.019) (0.023) Mainline 0.182*** 0.161*** (0.014) (0.013) Catholic 0.170*** 0.126*** (0.014) (0.013) No Affiliation 0.352*** 0.249*** (0.016) (0.015) Jewish 0.437*** 0.313*** (0.035) (0.032) Other Faith 0.080*** 0.012 (0.024) (0.023) Survey Year Year Spline (1977 to 1994) 0.034*** 0.030*** (0.001) (0.001) Year Spline (1994 to 2000) -0.025*** -0.026*** (0.003) (0.003) Year Spline (2000 to 2014) 0.017*** 0.020*** (0.001) (0.001) Controls Female 0.200*** (0.009) White —

Black 0.030 (0.018) Other -0.186*** (0.019) Age in Decades -0.098*** (0.003) Parental Status 0.044*** (0.011) Marital Status -0.124*** (0.010) Female Work Status 0.203*** (0.010) South -0.078*** (0.010) Less Than High School —

High School 0.158*** (0.013) Some College 0.228*** (0.021) Bachelor 0.272*** (0.017) Graduate 0.413*** (0.020) Family Income 0.009*** (0.001) 28

Constant -0.609 -0.407 N 21435 21435 R2 .075 .224 BIC 45508.477 41880.434 Source: General Social Survey 1977-2014 Note: Standard errors in parentheses; Models estimated with robust standard errors * p < 0.05, ** p < 0.01, *** p < 0.001 (two-tailed)

29

Table 3: Interaction Effects of Membership and Year on Gender Attitudes Scalea Model 1 Model 2 Main Effects Evangelical — —

Black Protestant 0.154* 0.121 (0.072) (0.065) Mainline 0.105* 0.041 (0.048) (0.043) Catholic 0.045 -0.005 (0.050) (0.045) No Affiliation 0.485*** 0.351*** (0.071) (0.064) Jewish 0.546*** 0.443*** (0.117) (0.099) Other Faith 0.012 -0.055 (0.106) (0.101) Year Spline (1977 to 1994) 0.030*** 0.026*** (0.003) (0.003) Year Spline (1994 to 2000) -0.030*** -0.030*** (0.006) (0.005) Year Spline (2000 to 2014) 0.023*** 0.024*** (0.003) (0.003) Year Spline (1977 to 1994) Interaction Effects Evangelical*Year Spline (1977 to 1994) — —

Black Protestant*Year Spline (1977 to 1994) 0.003 0.002 (0.006) (0.005) Mainline*Year Spline (1977 to 1994) 0.005 0.009* (0.004) (0.003) Catholic*Year Spline (1977 to 1994) 0.011** 0.010** (0.004) (0.004) No Affiliation*Year Spline (1977 to 1994) -0.008 -0.006 (0.005) (0.005) Jewish*Year Spline (1977 to 1994) -0.011 -0.013 (0.009) (0.008) Other Faith*Year Spline (1977 to 1994) 0.001 0.001 (0.008) (0.008) Year Spline (1994 to 2000) Interaction Effects Evangelical*Year Spline (1994 to 2000) — —

Black Protestant*Year Spline (1994 to 2000) 0.009 0.001 (0.011) (0.010) Mainline*Year Spline (1994 to 2000) 0.004 -0.001 (0.009) (0.008) Catholic*Year Spline (1994 to 2000) -0.002 0.003 (0.008) (0.007) No Affiliation*Year Spline (1994 to 2000) 0.007 0.005 (0.010) (0.009) Jewish*Year Spline (1994 to 2000) 0.036 0.034 (0.020) (0.018) Other Faith*Year Spline (1994 to 2000) 0.044** 0.042** (0.014) (0.013) 30

Year Spline (2000 to 2014) Interaction Effects Evangelical*Year Spline (2000 to 2014) — —

Black Protestant*Year Spline (2000 to 2014) -0.003 -0.001 (0.006) (0.005) Mainline*Year Spline (2000 to 2014) -0.002 0.001 (0.005) (0.004) Catholic*Year Spline (2000 to 2014) -0.009* -0.006 (0.004) (0.004) No Affiliation*Year Spline (2000 to 2014) -0.008 -0.007 (0.004) (0.004) Jewish*Year Spline (2000 to 2014) -0.023* -0.015 (0.011) (0.010) Other Faith*Year Spline (2000 to 2014) -0.026*** -0.021** (0.007) (0.007) Controlsa No Yes Constant -0.558 -0.344 N 21435 21435 R2 .078 .226 BIC 45635.982 42004.611 Source: General Social Survey 1977-2014 Note: Standard errors in parentheses; Models estimated with robust standard errors a Controls are female, race, age, parental status, marital status, female work status, region, education, and family income. Control values are almost identical to those in Table 2. * p < 0.05, ** p < 0.01, *** p < 0.001

31

Table 4: Ordinal Regression of Religious Groups and Over-Time Interactions on Morality of Homosexuality and Support for Same-Sex Marriage Morality Marriage 1 2 3 4 5 6 Main Effects Evangelical — — — — — —

Black Protestant 0.036 0.518*** 0.547*** 0.374*** 0.680*** 0.732*** (0.060) (0.082) (0.156) (0.076) (0.101) (0.192) Mainline 0.786*** 0.675*** 0.405*** 0.741*** 0.718*** 0.450** (0.038) (0.041) (0.081) (0.062) (0.065) (0.152) Catholic 1.004*** 0.811*** 0.399*** 0.997*** 0.890*** 0.547*** (0.037) (0.040) (0.082) (0.055) (0.057) (0.144) No Affiliation 2.061*** 1.756*** 1.456*** 1.759*** 1.519*** 0.734*** (0.045) (0.047) (0.098) (0.064) (0.066) (0.204) Jewish 2.535*** 2.136*** 1.825*** 1.958*** 1.711*** 1.296*** (0.088) (0.094) (0.164) (0.155) (0.160) (0.368) Other Faith 1.074*** 0.803*** 0.322* 1.027*** 0.869*** 0.823*** (0.060) (0.062) (0.135) (0.100) (0.104) (0.220) Year 0.039*** 0.040*** 0.028*** 0.061*** 0.068*** 0.055*** (0.001) (0.001) (0.003) (0.002) (0.003) (0.006) Interaction Effects Evangelical*Year — —

Black Protestant*Year -0.002 -0.004 (0.006) (0.010) Mainline*Year 0.012*** 0.014 (0.004) (0.008) Catholic*Year 0.019*** 0.019* (0.003) (0.007) No Affiliation*Year 0.013*** 0.040*** (0.004) (0.010) Jewish*Year 0.015 0.023 (0.008) (0.020) Other Faith*Year 0.022*** 0.002 (0.006) (0.011) Controlsa No Yes Yes No Yes Yes Constant (Cut 1) 2.146 1.432 1.173 0.918 0.083 -0.160 Constant (Cut 2) 2.398 1.715 1.456 1.740 0.968 0.725 Constant (Cut 3) 2.812 2.173 1.915 2.352 1.632 1.391 Constant (Cut 4) 3.532 2.907 2.672 N 29983 29983 29983 8785 8785 8785 BIC 53250.456 50061.611 50076.383 26089.061 25236.232 25265.590 Source: General Social Survey (Morality 1973-2014; Marriage 1988, 2004-2014) Note: Standard errors in parentheses; Models estimated with robust standard errors a Analyses use the same controls as Table 2: female, race, age, parental status, marital status, female work status, region, education, and family income. * p < 0.05, ** p < 0.01, *** p < 0.001

32

Table 5: Interaction Effects of Membership and Linear Year Term on Gender Attitudes Scale (N=13,509)a Model 1 Model 2 Main Effects Evangelical — —

Black Protestant 0.225*** 0.157*** (0.042) (0.042) Mainline 0.184*** 0.171*** (0.032) (0.029) Catholic 0.228*** 0.179*** (0.030) (0.028) No Affiliation 0.367*** 0.287*** (0.035) (0.033) Jewish 0.486*** 0.336*** (0.070) (0.064) Other Faith 0.176*** 0.097* (0.051) (0.048) Year 0.009*** 0.010*** (0.002) (0.002) Year Interaction Effects Evangelical*Year — —

Black Protestant*Year 0.001 0.000 (0.004) (0.003) Mainline*Year 0.001 0.001 (0.003) (0.003) Catholic*Year -0.006* -0.003 (0.002) (0.002) No Affiliation*Year -0.003 -0.003 (0.003) (0.003) Jewish*Year -0.007 -0.004 (0.007) (0.006) Other Faith*Year -0.008 -0.004 (0.004) (0.004) Controlsa No Yes Constant -0.301 -0.167 N 13509 13509 R2 .035 .167 BIC 28069.815 26210.206 Source: General Social Survey 1994-2014 Note: Standard errors in parentheses; Models estimated with robust standard errors a Controls are female, race, age, parental status, marital status, female work status, region, education, and family income. Control values are similar to those in Table 2. * p < 0.05, ** p < 0.01, *** p < 0.001

33