Introduction
In 2024, annual conventions among many of America’s religious denominations were rife with gendered resolutions. Among the successful measures at the Southern Baptist Convention’s annual meeting were a resolution opposing in vitro fertilization and a resolution stating support for parental autonomy in raising children (ERLC staff 2024). A majority of voting attendees also supported a constitutional amendment that would have formally banned churches with women pastors from being a part of the Southern Baptist denomination (Adams Reference Adams2024). Although this measure failed to reach the two-thirds majority needed for amendment, women’s ordination remains a salient topic in Southern Baptist discussions. In the past two conventions, Southern Baptist Convention (SBC) leadership expelled multiple congregations for being led by women pastors (Smith Reference Smith2024), and groups such as Baptist Women in Ministry continue to provide organized opposition to male-only leadership policies (Tucker Reference Tucker2024). Meanwhile, at the Presbyterian Church (U.S.A.)’s annual convention, the assembly discussed adding “gender identity” and “sexual orientation” to their list of protected classes who were “guaranteed full participation and representation in the life of the church” (Berkes Reference Berkes2024). After considerable debate, the amendment passed with a vote of 297 to 130 (Berkes Reference Berkes2024). These discussions demonstrate the continued debate within religious congregations around the role of women in the church and in broader society.
While denominational policies regarding women’s ordination were formalized decades ago in many denominations (Chaves Reference Chaves1997), gender and sexuality continue to permeate discussions of religious leadership and political activism in America’s congregations.Footnote 1 Moreover, women remain vastly underrepresented in leadership positions in America’s places of worship—with only 13.8 percent of American congregations being led by a woman and only 8.1 percent of American worshippers attending a congregation with a woman as head pastor or priest (National Congregations Study 2021, 68). In a political context, similarly, women remain underrepresented in places of power. In the 115th Congress (January 2017 to January 2019), for example, women held 21 percent of seats in the United States Senate and 19 percent of seats in the United States House of Representatives (Dittmar Reference Dittmar2017). Despite (or perhaps, because of) their underrepresentation in places of political power, women officeholders tend to see themselves as representing the interests of women beyond their geographic constituency (Dittmar et al. Reference Dittmar, Sanbonmatsu and Carroll2018) and work to represent women’s interests. In a political context, women officeholders tend to prioritize so-called “women’s issues”—that is, policies related to children and families, issues of equality, women’s health, and education (e.g., Poggione Reference Poggione2004). This prioritization is reflected in women officeholders’ political communication (e.g., Evans and Clark Reference Evans and Clark2016; Mendez 2010) and political behavior (e.g., Swers Reference Swers2002, Reference Swers2013; Holman Reference Holman2014; Vega and Firestone Reference Vega and Firestone1995; Volden et al. Reference Volden, Wiseman and Wittmer2018). Put differently, in political contexts, descriptive gender representation often results in the substantive representation of women’s political interests.
In this study, I examine the relationship between descriptive gender representation and women’s substantive representation in America’s places of worship, applying theories of political representation to a religious context. In a religious context, as seen in a political context, is there a relationship between descriptive gender representation and the substantive representation of women’s interests? In particular, I explore the relationship between the presence of clergywomen in a congregation (dyadic descriptive representation) or gender inclusive leadership policies (i.e., policies allowing women to serve as the head pastor or priest; collective descriptive representation) and “women’s issues”: a congregation’s participation in political activism related to abortion, activism in support of LGBT rights, and a general count of the number of “women’s issues” a congregation pursues. In doing so, this study contributes to our understanding of the comparability of representation theories across political and religious contexts and the potential ramifications of the underrepresentation of women leaders in American religious congregations.
I hypothesize that congregations led by clergywomen or those with gender inclusive leadership policies will prioritize participation in political activities that substantively represent women’s interests. Statistical analysis reveals partial support for my hypotheses. Collective gender representation, as demonstrated through the presence of policies permitting women to lead within a congregation, is found to increase the likelihood that a congregation will lobby or march in support of LGBT rights and participate in a broader range of “women’s issues” activism.
This paper proceeds as follows. The section on “Literature” introduces literature on descriptive gender representation and the substantive representation of “women’s issues” as well as scholarship on clergywomen and political activity. The section on “Hypotheses” outlines hypotheses and their theoretical backing, the “Data and methods” section describes the data and methodology, the “Results” section presents the results of the analyses, and the “Discussion and conclusion” section discusses the findings and identifies avenues of future research.
Literature
Political representation
Political science representation literature typically delineates two types of political representation, descriptive and substantive representation.Footnote 2 Descriptive representation, as defined by Hannah Pitkin (Reference Pitkin1967), is “the extent to which representatives ‘resemble’ or share certain characteristics with the represented” (Dovi Reference Dovi2008, 150). Descriptive gender representation, for example, is demonstrated in the relationship between a woman legislator and a woman constituent. In an electoral context, descriptive representation can be further delineated into two types: dyadic and collective. Dyadic representation is the relationship between a constituent and the elected officials for whom she has the opportunity to vote; it is the relationship between “one legislator and one constituency” (Weissberg Reference Weissberg1978, 535). Collective representation characterizes “legislators as a group of people collectively representing the people as a whole” (Weissberg Reference Weissberg1978, 536). An individual is not only represented by her specific members of Congress—that is, those elected from her state or district—but rather by all legislators that share her descriptive characteristics. Thus, if an individual is not descriptively represented by her specific legislator, she may still be descriptively represented by a congresswoman or congresswomen elected from another district or state.
Descriptive gender representation—both dyadic and collective—often translates into the substantive representation of women’s political interests. Scholarship examines the ways in which women legislators differ from their men colleagues in their pursuit of so-called “women’s issues.” Although scholars have taken different approaches to conceptualizing “women’s issues” (see: Reingold and Swers Reference Reingold and Swers2011), in the following analysis, I will be using the phrase to represent policy areas that are “particularly salient to women—either because they primarily, more directly, or disproportionately concern or affect women in particular or because they reflect the more ‘traditional’ concerns (or interests) that women presumably have about others” (Reingold and Swers Reference Reingold and Swers2011, 431).Footnote 3 In this paper, I examine specific activism in the areas of abortion and LGBT rights and a general measure of “women’s issues” activism.
Historically, women legislators hold more liberal positions on “traditional areas of women’s interest” such as “education, health, and welfare” (Poggione Reference Poggione2004, 306; Clark Reference Clark1998; Diamond Reference Diamond1977; Vega and Firestone Reference Vega and Firestone1995). Studies such as Poggione (Reference Poggione2004) find that women legislators—of both parties—tend to hold more liberal positions on traditional women’s issues such as welfare policy (305). Women legislators are not more liberal than men legislators in all issue areas (Frederick Reference Frederick2011; Schwindt-Bayer and Corbetta Reference Schwindt-Bayer and Corbetta2004), and the rise of political polarization is increasing the gap in policy interests among Democratic and Republican legislators, women and men alike. Women officeholders of both parties do, however, tend to view themselves as representing not only the interests of their geographic constituency (Fenno 1978) but also the collective interests of women (Dittmar et al. Reference Dittmar, Sanbonmatsu and Carroll2018).
Traditionally, women officeholders are more supportive than men officeholders on “issues of concern to women” (Frederick Reference Frederick2011, 193). This support is found to affect legislator action in myriad areas of political behavior. A study by Swers (Reference Swers1998), for example, finds that congresswomen are more likely than congressmen to vote for “women’s issues” bills even after controlling for factors such as ideology, partisan identification, and seniority (see also: Swers Reference Swers2002; Reference Swers2013). Due to their underrepresentation in America’s legislative bodies, women officeholders find ways to effect change and demonstrate their policy priorities beyond voting. Berkman and O’Connor (Reference Berkman and O’Connor1993), for example, find that women in state legislatures work to promote their pro-choice legislative priorities by obtaining committee assignments that allow them to block pro-life legislation. Other studies find women officeholders to prioritize “women’s issues” through their bill sponsorship and cosponsorship (Gerrity, Osborn, and Mendez Reference Gerrity, Osborn and Morehouse Mendez2007; MacDonald and O’Brien Reference MacDonald and O’Brien2011; Swers Reference Swers2005; Thomas and Welch Reference Thomas and Welch1991; Vega and Firestone Reference Vega and Firestone1995; Volden et al. Reference Volden, Wiseman and Wittmer2018), in their communication on the floor of Congress (Diamond Reference Diamond1977; Osborn and Mendez Reference Osborn and Mendez2010; Pearson and Dancey Reference Pearson and Dancey2011), in congressional e-newsletters (Hellwege and Cormack Reference Hellwege and Cormack2022), and on social media (Evans Reference Evans2016; Evans and Clark Reference Evans and Clark2016).
While women officeholders find creative institutional avenues such as speech-giving to substantively represent women’s interests, their effectiveness in influencing policy outcomes can be contingent on institutional factors such as their tenure in office, leadership positions, partisan identification, and political ideology (Beckwith Reference Beckwith2007). Moreover, as the parties polarize, women officeholders’ policy priorities often align better with their men co-partisans than with women across the partisan aisle. Republican women, in particular, must be strategic in how to address policy issues such as welfare, women’s health, and women’s rights that more directly align with the Democratic Party agenda (e.g., Meeks and Domke Reference Meeks and Domke2015). Partisanship and gender, then, interact to shape legislator actions, and the role of gender in relation to political action may shift over time (Rolfes-Haase and Swers Reference Rolfes-Haase and Swers2022). Scholarship has also begun to analyze how multiple identities interact to promote the substantive representation of women’s interests (Minta and Brown Reference Minta and Brown2014). In this study, I extend this research agenda to a different context, examining women’s leadership and a congregation’s political behavior in American religious congregations.
Clergywomen, political action, and congregations
Clergywomen remain underrepresented in American congregations. From 2018 to 2019, only 13.8 percent of American congregations had a woman as their lead pastor or priest (National Congregations Study, “Appendix: Tables,” 68). Knoll and Bolin (Reference Knoll and Bolin2018) find fewer than 10 percent of American worshippers attend congregations led by a clergywoman, even though more than half attend a congregation with egalitarian policies. American women, then, can find better descriptive representation in the United States Congress than in their places of worship.
The literature on clergywomen has traditionally examined the career opportunities and behaviors of clergywomen, often comparing these actions to those of clergymen. Clergywomen, traditionally, make less money and receive fewer benefits than clergymen (Lehman Reference Lehman2002; Lummis and Nesbitt Reference Lummis and Nesbitt2000), although this pay discrepancy may be improving in recent years (Hoegeman Reference Hoegeman2017). There are also differences in the types of congregations women and men lead. Clergywomen are more likely than men to lead smaller congregations (Lee Reference Lee2024) in rural or urban (rather than suburban) areas (Hoegeman Reference Hoegeman2017; see also Lummis and Nesbitt Reference Lummis and Nesbitt2000). Clergywomen’s congregations tend to be theologically moderate or liberal (Hoegeman Reference Hoegeman2017) and are more likely to be racially diverse than those led by clergymen (Lee Reference Lee2024). Other studies focus on differences in ministerial style between clergywomen and clergymen, finding qualified and slight differences in ministerial style. Lehman’s (Reference Lehman1993) study, for example, finds that clergywomen are more likely than clergymen to value empowering laity but finds no difference in “regard to authority, status, preaching, interpersonal style, or dealing with social issues,” (Lehman Reference Lehman2002, 29).
Still other studies explore the political beliefs and practices of clergy. American clergy demonstrate a variety of approaches to political activism (Glazier Reference Glazier2018; Roso et al. Reference Roso, Chaves and Keshintürk2024). These activities can range from direct endorsements of candidates and distributing voter guides in church (Wilcox and Sigelman Reference Wilcox and Sigelman2001) to discussing stances on political and social issues (Boussalis et al. Reference Boussalis, Coan and Holman2021; Brown et al. Reference Brown, Brown and Wyatt2023; Guth and Smidt Reference Guth and Smidt2022), participating in protests (Beyerlein and Ryan Reference Beyerlein and Ryan2018; Munson Reference Munson2019), and holding Get Out the Vote events (Edwards and Oyakawa Reference Edwards and Oyakawa2022). Political speech is also common in religious congregations. In their examination of over 100,000 sermons, Boussalis et al. (Reference Boussalis, Coan and Holman2021) find over 70 percent of pastors in their sample mention a political topic.
There are, thus, many political activities that congregations choose to pursue both within the confines of their church and beyond. In this paper, I examine participation in political activism: lobbying an elected official or marching in support of a political cause. These actions may hold a higher barrier to entry for congregations, some of which may avoid political activity due to fear of putting their 501(c)3 status at risk.Footnote 4 From 2018 to 2019, 25 percent of American congregations reported lobbying or marching in support of at least one political cause (Chaves et al. Reference Chaves, Anderson, Eagle, Hawkins, Holleman and Roso2020). It is this group of congregations, those participating in political activism, that I analyze in this study.
Djupe and Olson (Reference Djupe and Olson2013) detail how clergywomen tend to be both more politically liberal and politically active than clergymen (see also: Deckman et al. Reference Deckman, Crawford, Olson and Green2003; Djupe and Gilbert Reference Djupe and Gilbert2008; Olson, Crawford, and Guth Reference Olson, Crawford and Guth2000). On the topic of abortion, in particular, clergywomen and clergymen have pronounced differences in political opinion (Deckman et al. Reference Deckman, Crawford, Olson and Green2003; see also Finlay Reference Finlay1996). Moreover, Deckman et al. (Reference Deckman, Crawford, Olson and Green2003) find clergywomen to be more liberal than men on “social welfare issues, more supportive of government action, and more critical of reforms that restrict or limit government action” (625). Deckman et al.’s study suggests that clergywomen hold similar ideological beliefs as women officeholders; both groups of leaders tend to prioritize so-called “women’s issues.”
Another subset of literature examines the relationship between women’s religious leadership, congregational decisions, and the attitudes and behaviors of congregants. Women’s leadership in congregations and inclusive gender policies can have an empowering effect for women in terms of their levels of religiosity, spirituality, and efficacy in their religious congregations (Knoll and Bolin Reference Knoll and Bolin2018). Stewart-Thomas’ (Reference Stewart-Thomas2010) analysis of 1998 National Congregations Study (NCS) data identifies a positive relationship between women’s religious leadership and participation in various social service programs, and Houston-Kolnik and Todd (Reference Houston-Kolnik and Todd2016) find that congregations that permit women’s leadership are more likely to have programs for survivors of sexual and dating violence. In the context of COVID-19, women’s religious leadership was associated with actions related to public health and safety. Respondents who attended congregations with women’s leadership were more likely to report canceled religious services due to COVID-19 and to feel that “their best health interests at heart” (Bolin and Rolfes-Haase 2022, 160). Scholarship, thus far, has demonstrated connections between women’s religious leadership and congregational decisions in at least three gendered policy areas: social services, public health, and violence against women. Importantly, these results are not exclusive to congregations currently led by clergywomen but extend to religious governing boards (Stewart-Thomas Reference Stewart-Thomas2010) and gender inclusive leadership policies (Houston-Kolnik and Todd Reference Houston-Kolnik and Todd2016). In this project, I extend this research agenda, exploring the relationship between women’s descriptive representation (as head pastor or priest or through gender inclusive leadership policies) and the pursuit of gendered policy issues.
Comparing political and religious leadership
Much of the research on clergywomen and politics assumes that clergy can be compared to elected politicians and tests the efficacy of this assumption. Djupe and Olson (Reference Djupe and Olson2013), for example, justify the comparability of religious leaders and political leaders and then explore the effects of clergywomen’s leadership on women congregants’ attitudes and behaviors (see also Djupe and Gilbert Reference Djupe and Gilbert2003; Djupe and Gilbert Reference Djupe and Gilbert2008; Knoll and Bolin Reference Knoll and Bolin2018). Among their explanations, they state that many clergy are selected, or elected, through a democratic process and must maintain a degree of responsiveness to their congregations (Djupe and Olson Reference Djupe and Olson2013, 332). Moreover, clergy, like political leaders, have agenda-setting power. Scholarship (e.g., Boussalis et al. Reference Boussalis, Coan and Holman2021) demonstrates that political speech is common from the pulpit, and clergy (to a greater extent than many elected officials) have a captive audience to hear their message.
Of course, the comparison between political and religious leadership contexts in the United States is not a perfect one. While women are underrepresented in leadership positions in both spheres, women are permitted to serve in elected political office throughout the United States and are barred from serving in some religious traditions. Although this difference in congregational policy provides for an imperfect comparison between political and religious contexts, it does allow for analysis of the potential influence of these policies—a tactic that I will be taking in this study.
Djupe and Olson (Reference Djupe and Olson2013) identify a connection between religious gender representation and congregants’ political participation, finding that “attending a church with a female clergyperson enhances women’s political interest and participation” (Djupe and Olson Reference Djupe and Olson2013, 343). In this way, they find support for their hypothesis that clergywomen, like women officeholders, can affect their constituents’ relationship to politics. In this study, I extend this research agenda to examine another potential outcome of descriptive representation: the substantive representation of “women’s interests.”
Theoretical contribution
In this project, I apply the political representation theories of descriptive and substantive representation to a religious context. In particular, I explore the relationship between two forms of descriptive representation (i.e., dyadic and collective) and the substantive representation of “women’s interests” in a religious context. In a religious context, dyadic descriptive representation can be characterized by a woman belonging to a congregation with a woman pastor or priest. It is the relationship between a congregant and her pastor or priest. Collective descriptive representation is exhibited through a woman belonging to a congregation with gender inclusive leadership policies (i.e., a congregation that allows women to serve as the primary religious leader of a congregation, regardless of the gender of the current head pastor or priest). While a woman may not be descriptively represented in her own congregation, if her denomination has inclusive leadership policies, she can be collectively represented by clergywomen in other congregations within her denomination.Footnote 5 Even if she is not currently represented by a woman pastor or priest, her congregation’s leadership policy suggests that she may have had a woman pastor or priest lead her congregation in the past and, perhaps more importantly, the opportunity remains for a woman to lead her congregation in the future.Footnote 6
Moreover, a congregation’s leadership policy may have symbolic benefits to women congregants. In his landmark study of women’s ordination, Chaves (Reference Chaves1997) describes the symbolic motivations for institutional rules permitting women’s ordination. Permitting women to serve as head pastor or priest, regardless of whether a woman is currently serving in that role, represents a congregation’s commitment to gender equality. In a political context, women’s leadership can have both symbolic and substantive outcomes. In a religious context, scholarship has established that gender inclusive policies serve as a symbol—in this study, I will assess the substantive outcomes of these policies.Footnote 7
Additionally, this study explores the theories of descriptive and substantive representation in a nationally representative religious context. Much of the related scholarship on clergywomen and religious leadership examines one or a few religious traditions.Footnote 8 While these foundational studies are valuable for capturing the attitudes and behaviors of clergy and congregants within those specific denominations, they cannot fully capture the practices of clergy and congregants across America’s places of worship. This study contributes to a growing research agenda utilizing data from the NCS (National Congregations Study 2018), a nationally representative sample of religious congregations in the United States (e.g., Hoegeman Reference Hoegeman2017; Lee Reference Lee2024). By including all congregational observations in the latest available NCS sample (i.e., 2018-2019) rather than limiting by denomination, this study can examine the ways in which clergy gender and egalitarian leadership policies relate to congregational activities regarding “women’s issues” throughout the United States.Footnote 9
Hypotheses
In a political context, women officeholders tend to prioritize “women’s issues”—that is, policies and topics related to equality, children, families, education, and women’s health. This prioritization affects officeholders’ political actions (e.g., Gerrity et al. Reference Gerrity, Osborn and Morehouse Mendez2007; Swers Reference Swers2005; Volden et al. Reference Volden, Wiseman and Wittmer2018) and communication strategies (e.g., Hellwege and Cormack 2022; Osborn and Mendez Reference Osborn and Mendez2010). American clergy and congregations participate in multiple forms of political activities (Djupe and Gilbert Reference Djupe and Gilbert2003; Glazier Reference Glazier2018), and prior studies establish the comparability of women’s leadership in political and religious spheres (Djupe and Olson Reference Djupe and Olson2013). Given prior studies that find American clergy and their congregations act politically and may be comparable to elected officials, I expect that women’s descriptive representation in religious congregations will lead to a congregation’s substantive representation of women’s interests.
For my first analysis, I examine the relationship between descriptive gender representation and the number of “women’s issues” that a congregation reports lobbying or marching to support. I expect that descriptive gender representation in a religious context will be positively associated with “women’s issues” activism.
Hypothesis 1: Count of “women’s issues” activism
H1(a): Congregations led by a clergywoman will participate in more instances of “women’s issues” activism than will congregations led by a clergyman.
H1(b): Congregations with gender inclusive leadership policies will participate in more instances of “women’s issues” activism than will congregations with male-only leadership policies.
After analyzing the number of “women’s issues” congregations support through their political activism, I then examine political activism in two gendered areas: abortion and LGBT rights. Topics of abortion and women’s health fall within the traditional categorization of “women’s issues.” Women legislators (Berkman and O’Connor Reference Berkman and O’Connor1993) and clergywomen (Deckman et al. Reference Deckman, Crawford, Olson and Green2003) are more likely than their men colleagues to prioritize women’s health issues and to hold pro-choice stances. Because of this, I expect that descriptive gender representation in religious congregations will be negatively related to pro-life political activism.Footnote 10
Hypothesis 2: Pro-life political activism
H2(a): Congregations led by a clergywoman will be less likely than congregations led by a clergyman to participate in pro-life political activism.
H2(b): Congregations with gender inclusive leadership policies will be less likely than congregations with male-only leadership policies to participate in pro-life political activism.
Along with issues of women’s health, topics of equality are among those that women officeholders tend to differ from their men colleagues. Among both the public (PRRI staff 2024) and elected officeholders (Krimmel et al. Reference Krimmel, Lax and Phillips2016), women are more likely than men to support LGBT rights. In a religious context, there is evidence that among Mainline Protestants, clergywomen are more likely than clergymen to speak publicly on LGBT rights issues (Deckman et al. Reference Deckman, Crawford and Olson2008), and gender inclusive policies can symbolize a commitment to equality (Chaves Reference Chaves1997). Because of this, I expect that descriptive gender representation in a religious context will be associated with increased participation in pro-LGBT political activism.
Hypothesis 3: Pro-LGBT activism
H3(a): Congregations led by a clergywoman will be more likely than congregations led by a clergyman to participate in pro-LGBT rights activism.
H3(b): Congregations with gender inclusive policies will be less likely than congregations with male-only leadership policies to participate in pro-LGBT rights activism.
Data and methods
To explore the relationship between women’s leadership and a congregation’s pursuit of “women’s issues,” I analyze data from the most recent wave of the NCS, the 2018-2019 dataset. The NCS is affiliated with the General Social Survey and is a nationally representative sample of American religious congregations. The NCS “documents information about worship, programs, staffing, and other characteristics of American congregations” (National Congregations Study). Information about clergy gender, leadership policies, and a congregation’s political activism is the focus of this study.
To explore the relationship between descriptive gender representation and the substantive representation of women’s interests, I use a Quasipoisson regression (for hypothesis 1) and Linear Probability Models (LPMs) and Logistic generalized linear models (GLMs) (for hypotheses 2 and 3). I use a Quasipoisson model for hypothesis 1 because my dependent variable is a count of occurrences (i.e., how often a congregation participates in “women’s issues” activism).Footnote 11 In Tables 2 and 3, I present LPMs due to their ease of interpretability, and all Logistic regression models are available in the data appendix.Footnote 12 LPMs tend to produce similar coefficients as GLMs (Angrist and Pischke Reference Angrist and Pischke2009); in the following analysis, there are few differences between the models in either the statistical significance of the coefficients of interest or the directionality of the observed relationships.Footnote 13
My key independent variables of interest for all models are (1) the gender of the primary religious leader (i.e., the lead pastor or priest) and (2) congregational policy on women’s leadership.Footnote 14 Both variables are dichotomous. The “clergywoman” variable is coded 1 if a congregation’s lead pastor or priest is a woman and 0 if not, and the “gender inclusive leadership policy” variable is coded 1 if a congregation allows women to serve as the head pastor or priest and coded 0 if not. The dependent variable for my first hypothesis is a count of how many “women’s issues” a congregation lobbied or marched to support in the past year. This measure includes issues or policies related to women, children, health, education, and equality.
For my second hypothesis, the dependent variable is a dichotomous variable representing whether or not a group from the congregation lobbied or marched related to a pro-life abortion stance within the past 12 months. For my third hypothesis, the dependent variable is dichotomous, indicating whether or not a group from the congregation lobbied or marched in support of “more rights for gays, lesbians, or transgender people” (Chaves et al. Reference Chaves, Anderson, Eagle, Hawkins, Holleman and Roso2020, 117).
In each analysis, I include a series of control variables accounting for congregational-level, clergy-level, and ideological factors that may affect whether a congregation participates in political activism. Congregational and clergy-level controls contain standard demographic variables (e.g., gender, income, race, education, and region). For ideological controls, I control for a congregation’s political ideology and religious ideology (“theologically speaking, would your congregation be considered more on the conservative side, more on the liberal side, or right in the middle?” (Chaves et al. Reference Chaves, Anderson, Eagle, Hawkins, Holleman and Roso2020, 169).
As previously discussed, “women’s issues,” such as women’s health, equality, and social welfare, are often associated with the Democratic Party (e.g., Fagan Reference Fagan2021) and with political liberalism. Thus, I expect that a congregation’s political ideology will be associated with the likelihood that they take political actions in support of so-called “women’s issues.” Moreover, scholarship demonstrates the connections between theological conservatism (Perry and Whitehead Reference Perry and Whitehead2016; Thomson and Froese Reference Thomson and Froese2018) and a host of political attitudes (e.g., Vegter and Haider-Markel 2020) and behaviors (Leege and Kellstedt Reference Leege and Kellstedt1993).Footnote 15
Model 1 in each table includes only the two key independent variables, model 2 includes congregational-level controls, model 3 includes the congregational- and clergy-level controls, model 4 includes the congregational-level, clergy-level, and ideological controls, and model 5 includes additional controls for religious tradition. In model 5, I include dummy variables representing Roman Catholic, White Evangelical, and Black Protestant (Mainline Protestantism is the reference category). Please refer to Table 1 in the appendix to see descriptions of all control variables. In addition to control variables, each model includes a weighting variable that allows for analysis of the NCS data at the congregational level.
Results
Table 1 presents the relationship between descriptive gender representation and the number of so-called “women’s issues” a congregation pursues.
Table 1. Congregation’s participation in women’s issues activism
Note: 0.1”; 0.05*; 0.01**; 0.001***.
As is evident in Table 1, when controlling for congregational-level (model 2), clergy-level (model 3), ideological (model 4), and denominational factors (model 5), I find support for hypothesis 1b. Congregations with inclusive leadership policies (i.e., those that allow women to serve as head pastor or priest) are more likely than those with male-only leadership policies to participate in “women’s issues” activism. When controlling for congregational, clergy, and ideological controls, I find congregations with gender inclusive policies participate in “women’s issues” activism 4.52 times more often than congregations with male-only leadership policies. I do not find support for hypothesis 1a. Congregations led by clergywomen are no more or less likely than congregations led by clergymen to participate in “women’s issues” activism.
When looking at the control variables, I find the percentage of women within a congregation is a positive and significant predictor of participation in “women’s issues” activism, while the percentage of congregants over the age of 60 is negatively associated with this form of activism. Against expectations, ideological factors are not significant. In this form of gendered political activism, then, a gender inclusive policy rather than the ideological composition of a congregation matters in relation to the political activities a congregation pursues. Among denominational controls, I find that Black Protestant congregations are less likely than my reference category, Mainline Protestant congregations, to participate in political activism on “women’s issues” topics.
Table 2 displays the relationship between a congregation’s descriptive gender representation and the likelihood that a congregation participates in pro-life activism. As a reminder, I expect that the congregations led by clergywomen or those with gender inclusive leadership policies will be less likely than those with clergymen or male-only leadership policies to participate in pro-life activism. As shown in Table 2 model 4, when controlling for congregational-level, clergy-level, and ideological factors, I find initial support for hypotheses 2a and 2b. Congregations led by a clergywoman are 16.6 percentage points less likely than those with a man as their head clergy to participate in pro-life activism (this should be interpreted with caution, as this relationship is not significant in models 1 or 2).
Table 2. Congregation’s participation in pro-life activism
Note: 0.1”; 0.05*; 0.01**; 0.001***.
I also find congregations with gender inclusive leadership policies to be 18.2 percentage points less likely than those with male-only leadership policies to participate in pro-life activism. While this relationship is robust to congregational, clergy, and ideological controls, controlling for religious tradition alters the statistical significance of the key independent variables. As expected, given their history of pro-life activism, the Roman Catholic and White Evangelical variables are both statistically significant and positively related to the likelihood that a congregation lobbies or marches for pro-life causes. When controlling for religious tradition, it appears that religious tradition, rather than descriptive gender representation, may have been driving the observed relationships between clergywomen, leadership policy, and abortion activism observed in models 2 through 4.
When looking at the control variables (and depending on model specification), I find the percentage of women within a congregation, clergy education, and theological liberalism to be negatively associated with pro-life activism. A congregation’s total attendance, income level, and the presence of White clergy are positively associated with pro-life activism.
When examining the likelihood that a congregation participates in pro-LGBT activism, I find support for hypothesis 3b. As seen in Table 3, congregations with gender inclusive leadership policies are more likely than those with male-only leadership policies to participate in pro-LGBT political activism. In model 4, controlling for congregational, clergy, ideological, and religious tradition factors, having an egalitarian leadership policy is associated with a 25.6 percentage point increased likelihood of lobbying or marching in support of LGBT rights (in model 5, this relationship falls just under a standard 0.05 level of statistical significance with a p-value of 0.052, but it is also significant in the logit models presented in the online appendix).
Table 3. Congregation’s participation in Pro-LGBT activism
Note: 0.1”; 0.05*; 0.01**; 0.001***.
Against expectations, I do not find a statistically significant relationship between the presence of a clergywoman and a congregation’s participation in pro-LGBT rights activism. Among the clergy-level controls, I find that as the educational level of a congregation’s head clergyperson increases, so too does the congregation’s likelihood of lobbying or marching in support of LGBT rights. Surprisingly, I find no evidence of a relationship between a congregation’s ideology or religious tradition and the likelihood that they participate in pro-LGBT activism. As with the analysis in Table 1, a congregation’s leadership policy permitting women’s leadership rather than their political, theological, or ideological orientation is associated with an increased prioritization of a gendered issue.
Discussion and conclusion
Analyzing data from the 2018-2019 NCS, I explore the relationship between women’s religious leadership and a congregation’s political activism in gendered policy areas. In so doing, I assess the relationship between women’s descriptive gender representation in a congregation and a congregation’s substantive representation of “women’s interests.” Overall, I find that in a religious context, a congregation’s collective gender representation (as indicated by a gender inclusive leadership policy) matters in predicting a congregation’s pursuit of “women’s issues.” Congregational policy is positively associated with the number of “women’s issues” a congregation pursues and with their political actions in support of LGBT rights.
I do not find evidence that congregations led by clergywomen differ from those led by clergymen in the political actions they take regarding “women’s issues.” However, I want to caution that issues of statistical power may be affecting the ability to accurately analyze this relationship. Due to clergywomen’s vast underrepresentation in American places of worship, when examining political activism taken by congregations, fewer than 50 congregations remain in my sample that have a woman serving as their head pastor or priest, and there may not be enough variation within this variable to conduct a meaningful statistical analysis. Future research with a larger sample size of congregations with clergywomen should continue to explore the relationship between the presence of clergywomen and a congregation’s pursuit of “women’s issues.”
In the area of abortion activism, after controlling for religious tradition (and “Roman Catholic” and “White Evangelical” affiliation, in particular), I no longer find support for my hypotheses. I find initial support for my hypotheses that congregations with gender inclusive leadership policies were significantly and substantively less likely than those with male-only leadership to participate in pro-life political activities. This preliminary outcome supports Deckman et al.’s (Reference Deckman, Crawford, Olson and Green2003) finding that clergywomen and clergymen have the largest ideological divide on the topic of abortion. It also aligns with myriad studies of American officeholders demonstrating the prioritization of abortion legislation among women legislators—although the role of gender in predicting political behavior on topics such as abortion is changing as political polarization increases (Rolfes-Haase and Swers Reference Rolfes-Haase and Swers2022).
The analysis of abortion activism in this study suggests that in a religious context, religious tradition—rather than women’s leadership—matters more in explaining a congregation’s decision to support pro-life political activism. The role of religious communities within the pro-life movement is well documented (e.g., Cook et al. Reference Cook, Jelen and Wilcox1992; Swank Reference Swank2020; Wilcox and Gomez Reference Wilcox and Gomez1990), and the results from this study provide further evidence of the role of Roman Catholic and White Evangelical communities in promoting pro-life causes. Future studies should examine the relationship between descriptive gender representation and pro-choice activism. While the sample size of congregations that participated in pro-choice activism in the 2018-2019 NCS was too small to explore this question quantitatively, this would be a valuable research question for qualitative analysis. It is reasonable to suspect that a different process is at work when deciding not to participate in a pro-life protest versus choosing to participate in a pro-choice protest. I look forward to future analyses exploring the relationship between women’s religious leadership and political support for women’s reproductive rights. Examining this relationship between women’s religious leadership and abortion activism is all the more important in our post-Roe political context.
In the areas of LGBT rights and the number of “women’s issues” pursued, I find that a congregation’s leadership policy, rather than their ideological orientation or religious tradition, significantly predicts a congregation’s political behavior. Congregations that allow women to serve as the head pastor or priest (demonstrating collective descriptive representation) are more likely than those with male-only leadership policies to participate in political activism related to “women’s issues.” Importantly, the observed relationships are robust to numerous controls, including those for congregational characteristics, ideology, and religious tradition. One can be confident, then, that the observed relationship is not the result of congregations with egalitarian leadership policies simply being more politically or theologically liberal.
As a robustness check for my findings, I examine the relationship between women’s leadership and the likelihood that a congregation participates in any form of political activism (that is, that they report lobbying or marching for any political topic in the last year). I find that congregations led by clergywomen are no more or less likely than those led by clergymen to participate in political activism, and congregations with gender inclusive leadership policies are less likely than those with male-only leadership policies to participate in activism. If congregations led by women, or those with gender inclusive policies, were more likely to participate in any form of political activity, then this could mean congregations led by women are more comfortable with political action (regardless of the cause). Instead, evidence supports my theoretical expectation. Congregations with gender inclusive policies are not more supportive of any political action but rather are motivated to action in gendered political spaces.
The evidence supports the idea that gender inclusive congregational policies can matter in both symbolic and substantive ways. While scholarship (Chaves Reference Chaves1997) has long acknowledged that a gender inclusive leadership policy can serve as a symbol in support of women’s equality, my findings in this study suggest that this symbol may prompt congregations to pursue political action in gendered policy areas. I look forward to future research examining the causal mechanism behind this relationship, a task that is beyond the scope of this project.
What this study can do, however, is provide further insight into the ways descriptive gender representation is comparable across political and religious contexts. The findings in this study suggest that collective descriptive representation can facilitate the substantive representation of women’s interests in America’s places of worship. As women remain vastly underrepresented in American religious leadership, this is an encouraging finding for advocates of women congregants’ substantive representation. In gendered policy areas, a congregation’s political action may not be contingent on the gender of their religious leader but rather can be prompted by congregational policies allowing women to lead. Put differently, the opportunity for women to lead rather than the actual presence of women’s leadership within congregations is associated with the substantive representation of women’s interests.
Future studies should take an intersectional approach to exploring the relationship between descriptive and substantive representation in America’s places of worship. In a political context, Demi and Brown (2014) explore the dual roles of gender and racial diversity in representing “women’s issues” in congressional hearings. In a religious context, similarly, scholars should explore how multiple identities interact in relation to the types of policies and programs religious congregations pursue. While in this study, I found little evidence that a congregation’s racial composition predicts their political activism in the observed issue areas, studies should continue to examine the combined and varying effects of gender, sexuality, race, religion, and partisanship in predicting a congregation’s political activism. This would contribute to a growing research agenda examining the intersectional experiences of clergy (e.g., Stringer Reference Stringer2023) and congregants (e.g., Vega Reference Vega2019; Winder Reference Winder2015; Zerai Reference Zerai2016).
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1755048325100163
Financial support
The author declares none.
Competing Interests
The author declares none.
Cammie Jo Bolin is an assistant professor of political science at University at Albany, SUNY. Her research examines questions of women’s leadership, representation, and participation in political and religious contexts in the United States.
Open access
