The Impact of Marriage Norms and Gender on Anglican Clergy Actions in Response to Domestic Violence

: Domestic violence (DV) is a gendered issue, with women more likely to be victim/survivors and men more likely to perpetrate abuse. With a strong emphasis on protecting the safety of women and children, the ways in which faith-based communities and leaders engage DV has come under scrutiny. Clergy are potential responders to DV and shape cultural contexts in which DV occurs. Yet, how religious norms relate to actions taken when clergy respond to DV remains under-researched. Using a survey of Australian Anglican clergy, this paper explores how views about “headship” and the sanctity of marriage relate to the uptake of actions by clergy and churches. Increased support for headship predicted a lower take-up of victim/survivor safety-focused actions by clergy and a lower frequency of actions by churches focused on DV organizations. Male clergy were more likely than their female counterparts to engage with perpetrators and to counsel couples. No independent associations were observed between actions and heightened support for the sanctity of marriage. While the strength of relationships between DV actions and both norms and gender was generally weak, these ﬁndings indicate that more work is needed to heighten awareness of the importance of actions focused on victim/survivor safety, connections with DV support services, the problematic practice of couples counselling, and challenges around directly pastoring perpetrators.


Introduction
Domestic violence (DV; also referred to as intimate partner violence or domestic abuse) is a type of family violence between current or former intimate partners, which can include acts of physical violence, sexual violence, emotional or psychological abuse, and patterns of monitoring, isolation, and control known as coercive control (AIHW 2018). DV is a major and widespread social, economic, and public health problem both internationally (WHO 2010(WHO , 2021 and in Australia (AIHW 2018). Religious norms, including norms about intimate relationships, are part of the complex juxtaposition of individual, relationship, community, and societal factors that shape experiences of DV for victim/survivors, perpetrators, and responders Pepper and Powell 2022;Truong et al. 2020;Vaughan et al. 2020). Clergy play an important role both as potential responders for those within faithbased communities and as shapers of cultural contexts within which intimate relationships are situated. Yet how religious norms relate to actions taken when clergy respond to DV in church communities remains under-researched.
DV is a gendered issue, as shown in recent international and Australian research, with women more likely to be victim/survivors and men more likely to perpetrate abuse (ABS 2017(ABS , 2021(ABS -2022Cox 2016;Webster et al. 2018;WHO 2021;Coumarelos et al. 2023). The 2021-2022 Australian Personal Safety Survey (PSS) found that 1 in 4 women experienced church helped them, but others reporting negative experiences (e.g., Barnes and Aune 2021;Knickmeyer et al. 2010;Mulvihill et al. 2022;Popescu et al. 2009; This is Not Big Little Lies 2022).
Norms that operate in religious contexts, as understood and applied by both those who experience abuse and those who perpetrate it, have variously been found to be negatively and positively related to outcomes in DV situations. Recent research studies, particularly those based on victim/survivor testimonies, suggest that certain norms can present risks to the safety of women who are experiencing violence (e.g., Barnes and Aune 2021;Knickmeyer et al. 2010;Mulvihill et al. 2022;Vaughan et al. 2020;Zust et al. 2021). For instance, in Christian settings, the principle of forgiveness of wrongdoing, a mandate for submission of wives to husbands, and the sanctity of marriage can be barriers to women in situations of DV seeking and receiving assistance (Nason-Clark et al. 2017;Truong et al. 2022;Westenberg 2017). Furthermore, male perpetrators can misuse Christian scripture to justify aspects of their abusive behavior or refer to the sacredness of marriage vows when they demand forgiveness (Lock 2018;Knickmeyer et al. 2010;. Conversely, other discourses/norms, such as marriage as a covenant and the equality of partners in a marriage can empower victims/survivors to extricate themselves from abusive relationships Westenberg 2017).
In Christian traditions, clergy play a key role in shaping and reinforcing social norms about intimate relationships. Clergy have regular and varied platforms for expressing Christian values and principles, such as in preaching, biblical interpretation, liturgy, marriage programs, and more . Clergy attitudes regarding the preservation of the sanctity of marriage and the sinfulness of divorce were found to be a barrier for victim/survivors to receive assistance (Zust et al. 2021). These views can also be used by faith leaders to exhort forgiveness and reconciliation (Mulvihill et al. 2022). In contrast, when religious language links forgiveness of wrongdoing to justice and dignity and is framed as a covenantal relationship between partners (and broken by the perpetrator), the discourse can be shifted, enabling victim/survivors to consider other options Westenberg 2017).
Beyond the more general shaping of norms, clergy can offer specific assistance to those who experience abuse by making referrals to DV support services, providing spiritual counsel, and supporting victim/survivors and their families in the long-term process of healing, and some clergy provide pastoral support and accountability for perpetrators (Nason-Clark et al. 2017). Responses by clergy to DV were addressed in a range of recent research studies (e.g., Ames et al. 2011;Behnke et al. 2012;Bent-Goodley et al. 2015;Choi 2015aChoi , 2015bChoi et al. 2018;Davis and Johnson 2021;Dyer 2010;Nason-Clark et al. 2017;Pepper and Powell 2022;Skiff et al. 2008;Truong et al. 2022;Williams and Jenkins 2019;Zust et al. 2021). Some of these studies suggest a connection between clergy upholding gender stereotypes and/or norms protecting the sanctity of marriage and how clergy report they would respond in situations of DV (Behnke et al. 2012;Choi 2015a;Skiff et al. 2008;Truong et al. 2022;Zust et al. 2021). In our review of recent literature, however, we found no quantitative studies that empirically investigated the actions of clergy in actual situations of DV with reference to their attitudes about marriage norms.
Moreover, while DV is well established as a gendered issue, there is limited research in the context of faith-based communities on whether the gender of clergy or other responders makes any difference. For example, while Zust et al. (2021) analyzed attitudes of congregational members by gender and found differences, there were insufficient responses from female clergy in their sample. One reason for this lack of research is that many Christian (and non-Christian) religious traditions adhere to male-only religious leadership.
The National Anglican Family Violence Project Clergy and Lay Leader Survey (Powell and Pepper 2021) provides an opportunity to examine relationships between two norms-headship (the theological position that wives should submit to husbands) and the sanctity of marriage-and the variety of actions undertaken by both male and female clergy and their churches in domains of both response to and prevention of DV. The Anglican dataset also enables us to explore whether there is any interaction between norms and the gender of clergy; that is, whether norms influence actions differently for male and female clergy.

The Anglican Church
The Anglican Church, with its breadth of theological traditions, is a useful context in which to explore the operation of various norms concerning intimate partner relationships and DV.
With a presence in the lands now called Australia since the beginning of European colonization, the Anglican Church remains one of the major Christian denominations in the country today, with some 9.8% of the Australian population identifying as Anglican in 2021 (ABS 2022). The prevalence of DV appears to be similar or greater among Anglicans than in the Australian community at large , and a significant majority of Anglican clergy deal with specific DV situations as a part of their ministry, mainly responding directly to victim/survivors, while a substantial minority engage with perpetrators .
Historically, the founders of Anglican dioceses emphasized different theological influences: evangelical, Anglo-Catholic, and liberal (also called progressive) emphases (Frame 2007). In broad terms, the dioceses continue to reflect the theological emphasis of their first bishops (Hilliard 2009). Sydney, which is the largest diocese in terms of membership, remains largely evangelical/reformed, while other dioceses are more diverse. In 2021/22 across Australia, some 41% of Anglican churchgoers identified with an evangelical or reformed faith tradition, 32% identified with Catholic or Anglo-Catholic traditions, 9% identified with Pentecostal or charismatic traditions and only 5% identified with a liberal or progress approach to faith .
Understanding gender roles within Christian communities is informed by differing Christian theological ideas that tend to be applied across domestic, work, and church spheres. For example, traditional Catholic and Protestant reformed positions hold to a "headship" view for the Christian household ("Wives, be subject to your husbands as to the Lord, for the husband is the head of the wife just as Christ is the head of the church, his body, and is himself its Savior" Ephesians 5:22-23). In recent decades, those from an evangelical tradition extended theological ideas about headship by adopting the terms "complementarian" and "egalitarian" to identify their position on gender relations. This system of biblical interpretation was developed in the USA in the 1970s and 1980s (e.g., Piper and Grudem 1991). A complementarian approach posits that men and women are "equal but different" and that there are specific male and female "roles" (Shorter 2021). Key ideas are that complementarity is part of God's creation, with differences between men and women expressed in marriage, in the family, and sometimes extended to the workplace. Within the church context, as in a family, roles are seen as gender-based and only men are to be elders (leaders). The alternate egalitarian position makes a theological case for women and men to operate in mutual service and interdependence, being unrestricted in the roles they take in the home, workplace, or the church (e.g., Giles 2020; Mowczko 2020) ("Nevertheless, in the Lord woman is not independent of man, nor is man independent of woman. For as woman came from man, so also man is born of woman. But everything comes from God." 1 Corinthians 11:11-12).
These differing theological positions have implications for clergy gender. The ordination of women has been subject to decades of debate in the Anglican Church. While the General Synod approved legislation in 1992 allowing dioceses to ordain women to the priesthood, a minority of dioceses (4 out of 23, including Sydney) have not adopted the legislation. Of the 3831 clergy listed in the Anglican Church of Australia Directory 2020/21, 888 (23%) were women, including seven bishops, 397 active priests, 171 active deacons, 66 "other clergy", and 247 retired clergy (ACA 2022). Broadly speaking, the Anglo-Catholic tradition and the conservative evangelical tradition historically rejected the ordination of women, whereas the charismatic evangelical tradition historically encouraged women's ministry (Fry 2021a). Even in contexts where the Anglican Church has admitted women into leadership at all levels, evidence of gender inequality remains; within the Church of England, it is historic and ongoing, including in the governing structures of the church, selection processes for clergy and pay rates (Fry 2019(Fry , 2021a(Fry , 2021bGreene and Robbins 2015). In an empirical review of gender roles and church tradition in the Church of England, Village (2012) found those with a broad church identity tended to hold views more closely aligned with wider society on matters such as women's ordination, divorce, and remarriage, in contrast to more conservative views held by conservative Anglo-Catholics or evangelicals.
In an Australian study of Anglican victim/survivors, links were observed between experiences of DV and values, beliefs, and practices regarding intimate relationships. Absolutist positions of marriage as a lifelong commitment, the submission of the wife to the husband, unconditional forgiveness, and suffering for Christ-whether propagated by clergy, internalized by victim/survivors, or co-opted by abusers-were harmful for those experiencing abuse .
The data collected from Anglican clergy as part of the National Anglican Family Violence Project Clergy and Lay Leader Survey  provide an opportunity to extend understanding of the relationships between gender, specific norms concerning marriage, and DV responses within the context of a particular, diverse religious community that ordains both male and female clergy. Further, the scope of the survey instrument enables us to focus on a wide array of clergy and church actions with regard to DV.

Research Questions and Hypotheses
This paper poses the following questions: How do norms about marriage, specifically headship and the preservation of marriage, and clergy gender influence uptake of actions by clergy and churches in relation to DV? Is there any evidence that commitment to the norms of headship and the preservation of marriage on the part of clergy reduces actions that focus on the safety of those who experience DV? Conversely, do these norms increase the take-up of actions that compromise safety?
For example, couples counselling can place victim/survivors at risk because it fails to address the unequal power present in an abusive relationship . Recent studies indicated that couples counselling in a DV context is relatively common in Christian churches (e.g., Choi 2015a; Sojourners and IMA World Health 2014;Pepper and Powell 2022). Does commitment to the preservation of marriage relate to the actual uptake of couples counselling by clergy?
When it comes to the uptake of preventative actions, such as preaching or speaking about DV, it is unclear how norms would relate. Increased profile has been given to DV in the churches in recent years, including in the Anglican Church (e.g., ABM 2018, n.d.;ACA 2022;Broughton 2018; Anglican Diocese of Melbourne n.d.; Common Grace 2018; NCCA n.d.; PSU n.d.; Queensland Churches Together n.d.; UCA Assembly 2021). It is therefore possible that clergy who hold to headship and the sanctity of marriage may actually have a heightened awareness of the dangers of these norms and be more likely to speak about DV.
Violence is gendered and norms about gender roles in churches and clergy gender are intertwined. The Anglican sample had enough female clergy respondents to enable us to include clergy gender in the analysis. We expected that male clergy would be more likely than female clergy to engage with perpetrators and to undertake couples counselling .

Recruitment and Instrument
The data were collected as part of the National Anglican Family Violence Project Clergy and Lay Leader Study, commissioned by the Anglican Church of Australia . The study was undertaken by NCLS Research in order to understand how Anglican clergy and local church leaders engage with issues of DV in the context of their local church, and thereby help to guide church policy and practice in relation to equipping clergy and lay leaders to respond to these issues. The research was conducted with the approval of the Human Research Ethics Committee of Charles Sturt University (Ethics Register Number H19306).
The anonymous online survey was intended for Anglican clergy in active ministry and lay people in identified parish leadership positions. From May to August 2020, a series of letters and emails were sent to around 1400 Anglican parishes using contact details for a "location" (congregation or worship center) in each parish, as held in the NCLS Research church database. Due to shortcomings with available contact details, not all parishes or congregations received the invitation to take part.
Recipients were asked to send the invitation to participate in the survey to clergy and lay leaders across the parish. The results presented in this paper are for clergy serving in their church only (i.e., those who indicated that they were the rector, vicar, or senior minister of the parish; a priest in a ministry team in the parish, but not the rector/vicar/senior minister; a curate; or a deacon or deaconess). Two clergy identified as non-binary gender. There were too few such clergy for quantitative analyses of gender; therefore, they are excluded from the analysis presented here.

Instrument
The online survey of approximately 30 min duration included questions on: perceptions of the prevalence, seriousness, and causes and nature of DV; views about the role of clergy and other church leaders in addressing DV; actions taken by clergy and other leaders to respond to DV; awareness of and familiarity with policies, frameworks, training, services, and other resources to help churches to respond; evaluation of how equipped clergy and leaders feel to address DV; beliefs about gender roles and intimate relationships; demographic characteristics and theological tradition of the respondents; details of local church leadership; and number and type of clergy and lay leaders in the parish.

Sample Characteristics
There were 356 clergy respondents from 289 parishes (21% of all parishes in Australia). The mean age was 54.4 years (range 25-86). Some 27.5% of respondents were female and 76.6% were Australian born. Some 35.5% held a degree as their highest level of qualification and 54.4% a postgraduate qualification, including 40.1% with a postgraduate ministry qualification. Some 60.2% identified as evangelical and 29.7% identified as reformed, with almost all who identified as reformed also identifying as evangelical. Some 26.6% identified as Catholic or Anglo-Catholic.
Because this was an opt-in survey, the sample cannot be assumed to be representative of the broader population of Anglican clergy from which it was drawn. It is likely that the results are biased towards those who have an interest in the topic. Respondents overwhelmingly strongly agreed or agreed that DV is common in Australia (93%), that it is reasonable to expect clergy to address it within their ministry (95%), and that the top priority when dealing with specific DV situations is victim safety (99%). They were also highly aware of the nature of DV, with 94% or more of the view that each of a set of specific behaviors across physical, sexual, psychological, emotional, and spiritual domains of abuse were always or usually DV. Nevertheless, the sample was diverse in its theological approach, making it suitable for an investigation of norms regarding marriage and responses to DV.
Clergy actions in response to specific DV situations: Respondents were asked whether they dealt with individuals in specific DV situations as part of their ministry (yes = 1, no = 0). They then indicated which of a set of 16 actions they used in these situations (coded 0 if the action was not selected, coded 1 if the action was selected).
Clergy cultural actions: Respondents indicated the frequency of three actions they had undertaken at their church to create a culture of DV prevention and response (often = 5, sometimes = 4, occasionally = 3, rarely = 2, and never = 1).
Church actions: Respondents were asked about whether four actions happened at their church in the previous 12 months in areas of prayer, support for and work with DV organizations, and the display of information in the church (yes = 1, no = 0, and unsure = 0). Because there were multiple respondents in some parishes, results are presented for one respondent per parish (retaining the most senior clergyperson).
The sanctity/preservation of marriage item was adapted from Jones et al. (2005). The clergy cultural action items and church action items were adapted from Aune and Barnes (2018).
For parsimony in multivariate analysis, scales of DV actions were formed where possible. Exploratory factor analyses (principal components analysis with oblique rotation) were first conducted on each of the three sets of items (clergy actions in response to specific DV situations, clergy cultural actions, and church actions). Reliability analyses were then undertaken to select final items for scales. Three clergy action scales were formed as follows: victim/survivor safety-focused clergy action in response to specific DV situations, perpetrator-focused clergy action in response to specific DV situations, and clergy cultural action. Church actions did not form a sufficiently reliable scale.
The item wordings and descriptive statistics for marriage norms and DV actions and computed scales are shown in Table 1. Table 1. Item wordings and descriptive statistics for study variables.

Headship:
The husband is the head of the wife (as Christ is the head of the church), and the wife should submit to the husband. (n = 354) Strongly disagree (26.6%), disagree (12.1%), neither agree nor disagree (13.3%), agree (22.9%), and strongly agree (25.1%).

Dealt with DV Situations:
Have you, as part of your ministry, ever dealt with individuals in specific domestic violence situations? (n = 343) No (31.5%), Yes (68.5%)

Victim/Survivor Safety-Focused Clergy Action Scale:
Of the following actions, which have you used when dealing with specific domestic violence situations? 7 items (M = 2.69, SD = 1.84, range = 0-7, Cronbach's alpha = 0.703, n = 234), additive scale comprising: Referred the victim to a service agency (e.g., crisis support, counselling, legal support, and financial support) (87.1%). Assisted the victim with a safety assessment or safety plan (49.3%). Assisted the victim with paperwork or statements (e.g., police report, Apprehended Violence Order) (38.3%).
Made an intervention to keep children safe (22.5%). Made a child protection report to government authorities (24.7%). Contacted the police (28.2%). Provided rectory or church property as a refuge (25.1%).
Note: Reported n is the valid number of cases for the question or scale.
Demographic variables: the demographic variables used in the analysis were gender (female 27.5% coded as 1, male 72.5% coded as 0), age (mean-centered, standard deviation 12.2 years), and highest level of ministry qualification obtained (non-postgraduate qualification 59.9% coded as 0, postgraduate qualification 40.1% coded as 1).

Analysis
IBM SPSS statistics package version 29 was used for analysis. We first describe the variation in agreement with headship and the sanctity and preservation of marriage in the sample.
Then we present the bivariate and multivariate relationships between DV actions and both marriage norms and gender, with statistical significance set at p < 0.05.
Bivariate relationships are shown for all individual actions. Somers' d was used for associations with marriage norms. For gender differences, chi-square tests (clergy actions in response to specific DV situations, church actions) and t-tests (frequency of clergy cultural actions) were used.
Regressions were performed with marriage norms and clergy gender, age, and education used to predict DV actions, using action scales were possible. Because previous research found a curvilinear relationship between clergy age and dealing with DV situations , both age (mean-centered) and the square of age were included. We also tested for interactions between marriage norms and clergy gender, to ascertain whether norms function differently for male and female clergy. Linear regressions were conducted for the multi-item action scales: victim/survivor safety-focused clergy action, perpetrator-focused clergy action, and clergy cultural action. Logistic regressions were undertaken for other individual action items that did not contribute to scales, except for pastoral/spiritual care to victim/survivors because uptake of this action was almost universal.

Headship and Sanctity/Preservation of Marriage
The large majority of clergy either agreed/strongly agreed (48%) or disagreed/strongly disagreed (39%) with headship, with a small minority being neutral or unsure (13%). Views about headship were strongly associated with faith tradition and gender (Table 2). Some 9 in 10 reformed clergy agreed or strongly agreed with headship, compared with about 6 in 10 clergy who identified as evangelical but not reformed, and less than 1 in 10 clergy who identified with Anglo-Catholic and other traditions. Very few female clergy agreed with headship, while approximately 6 in 10 male clergy did. The preservation of marriage was normative across the sample. A large majority of clergy (88%) either strongly agreed or agreed that marriage is a sacred covenant that is always important to preserve. Strength of agreement did, however, vary by faith tradition and gender. A majority of clergy from evangelical and reformed traditions strongly agreed, compared with a minority from Anglo-Catholic and other traditions. A majority of men strongly agreed, whereas women were more likely to agree than to strongly agree (Table 2).

Clergy Actions in Response to Specific DV Situations, Marriage Norms, and Gender
Some 68% of clergy dealt with specific DV situations as a part of their ministry, and this was not significantly related to marriage norms or to gender.
With the exception of referring perpetrators to service agencies, actions that were about engaging with the perpetrator (talking, pastoral/spiritual care, and counselling) were associated with headship, stronger endorsement of the sanctity and preservation of marriage, and clergy gender (male) at the bivariate level, as shown in Table 3. 1 Crosstabulation n = 341 clergy. 2 Crosstabulation n = 232 clergy who dealt with specific DV situations. 3 Action differs by headship by at most p < 0.05 (Somers' d with actions dependent). 4 Action differs by sanctity/preservation by at most p < 0.05 (Somers' d with actions dependent). 5 Action differs by gender by at most p < 0.05 (Chi-sq). 6 SD = Strongly disagree, D = disagree, N = neither agree nor disagree, A = agree, SA = strongly agree, F = female, and M = male. 7 Cell sizes for SD/D and N are small. These categories were combined for statistical tests.
Three victim/survivor safety-focused actions (referring the victim to a service agency, contacting the police, and assisting the victim with paperwork or statements) were negatively associated with headship. Among clergy who dealt with DV situations, almost all clergy who strongly disagreed or disagreed with headship referred a victim (98%), compared with 80% of those who agreed or strongly agreed. For contacting police, the results are 38% and 20%, respectively. Clergy who strongly agreed with headship were less likely than other clergy to have assisted victims with paperwork (22%, compared with 43-45% of other clergy). The relationship between several other victim/survivor-focused actions and headship was in the same direction but was not statistically significant. Referring victims to a service agency was associated with clergy gender; 97% of female clergy who responded to DV situations did so, compared with 84% of males.
Counselling victims and perpetrators was associated with a stronger endorsement of the sanctity and preservation of marriage. The relationship for marriage/couples counselling was in the same direction but was not statistically significant; however, undertaking marriage/couples counselling was associated with clergy gender. Some 21% of male clergy who responded to DV situations did so, compared with 8% of female clergy.
Informing or seeking help from the diocese was associated with headship (21% of those who disagreed with headship, 39% of those who agreed) and clergy gender (16% of females, 36% of males).
Associations between clergy action and marriage norms and gender were weak; that is, the effect size was small (Somers' d < 0.3 for norms, Phi < 0.3 for gender), with the exception of the gender differences on talking with perpetrators about the violence (Phi = 0.301).
Prediction of clergy actions in response to specific DV situations from norms, gender, age, and education was weak, with small amounts of variance explained (Tables 4-7).   Increased endorsement of headship negatively predicted victim/survivor safetyfocused clergy action in the regression. Gender, which did not significantly correlate with the action scale at the bivariate level, also held no independent predictive power at the multivariate level (Table 5).
As was the case in the bivariate results, being male predicted undertaking perpetratorfocused action (Table 6). Headship and the sanctity/preservation of marriage, however, held no independent predictive power when gender, age, and education were controlled.
Neither headship nor the sanctity/preservation of marriage independently predicted the other individual actions in response to DV situations ( Table 7). The regression weight for sanctity/preservation of marriage was positive but not statistically significant in the case of couples counselling. Being female significantly reduced the odds of undertaking marriage/couples counselling by 66%.
Interactions between gender and views about marriage were not significant in any of the regressions and are not shown.

Clergy Cultural Actions, Marriage Norms, and Gender
There was a weak positive bivariate association between frequency of preaching about DV and both headship and gender (male). However, talking about DV during worship services or other public gatherings did not relate significantly to gender or views about marriage. There was a weak positive association between the clergyperson publicly saying they were available to support people experiencing DV and sanctity/preservation of marriage (Table 8).
The regression predicting the frequency of clergy cultural action indicated an interaction between headship and gender (Table 9). There was no relationship between frequency of clergy cultural action and headship or sanctity/preservation of marriage among male clergy. In contrast, opposition to headship increased the frequency of cultural action among women. Female clergy who were neutral about the doctrine of headship spoke publicly about DV at a lower frequency than men, whereas female clergy who strongly disagreed with headship spoke at a similar frequency to men. This is illustrated in Figure 1. However, the relationships were weak and the regression model explained little variance in the frequency of clergy action.  The regression predicting the frequency of clergy cultural action indicated an interaction between headship and gender (Table 9). There was no relationship between frequency of clergy cultural action and headship or sanctity/preservation of marriage among male clergy. In contrast, opposition to headship increased the frequency of cultural action among women. Female clergy who were neutral about the doctrine of headship spoke publicly about DV at a lower frequency than men, whereas female clergy who strongly disagreed with headship spoke at a similar frequency to men. This is illustrated in Figure  1. However, the relationships were weak and the regression model explained little variance in the frequency of clergy action.

Church Actions, Marriage Norms and Gender
In terms of church actions, the church donating money/goods to a DV organization and working with a DV organization was negatively associated with increased endorsement of headship at the bivariate level (Table 10). In contrast, those who supported head-

Church Actions, Marriage Norms and Gender
In terms of church actions, the church donating money/goods to a DV organization and working with a DV organization was negatively associated with increased endorsement of headship at the bivariate level (Table 10). In contrast, those who supported headship were more likely than those who did not to report that their church displayed posters or leaflets about DV in the church. There was no relationship between sanctity/preservation of marriage or gender and any of the church actions. Again, statistically significant relationships were weak. In the multivariate predictions, the negative relationships between headship and church actions observed at the bivariate level held (Table 11). Increased endorsement of headship reduced the odds of the church donating money/goods to a DV organization and the church working with a DV organization (a 1-unit increase in headship decreased the odds of donating to and working with a DV organization by 36% and 30%, respectively). However, endorsement of headship did not predict the display of information in the church (the regression coefficient was positive but not statistically significant) and clergy gender did not hold predictive power for any of the church actions. Prediction was poor, with little variance explained. Interactions between gender and views about marriage were not significant and are not shown.
Clergy age and its square were included as controls in all regressions, along with postgraduate education. Age and/or its square were statistically significant predictors in 8 of the 12 regressions. In all but one of these cases, regression coefficients were negative, indicating a decrease in clergy action or church action with clergy age. There was one exception: counselling victims, which increased with age.
In summary, increased support for headship predicted a lower take-up of victim safety-focused clergy actions and less frequent actions by churches focused on DV organizations (donations to and work with such organizations). Among female clergy, but not among males, greater opposition to headship predicted increased frequency of speaking publicly about DV. Male clergy were more likely than their female counterparts to engage with perpetrators and to counsel couples. These relationships, on the whole, were weak. No independent associations were observed between heightened support for the sanctity/preservation of marriage and DV actions by clergy or churches.

Discussion
This study aimed to explore how norms about marriage, specifically headship and the preservation of marriage, and clergy gender influence uptake of actions by clergy and churches in relation to DV. The study looked at a variety of clergy and church actions that address and have the potential to prevent DV and violence against women (DSS 2022).
Much of the evidence for the harmful operation of certain norms concerning intimate partner relationships in Christian contexts is based on the testimonies of those who experienced DV (Knickmeyer et al. 2010;Mulvihill et al. 2022;. Given the role of clergy both in shaping norms and in responding to DV, this paper used a different methodology (a survey study of clergy attitudes and self-reported behaviors) to examine whether there was any evidence that commitment to the norms of headship and the preservation of marriage on the part of clergy reduces their uptake of actions that focus on the safety of those who experience DV or likewise increases actions that compromise safety.
When it came to responding to specific DV situations, increased support for headship did indeed predict a lower take-up of victim/survivor safety-focused clergy actions, such as referring the victim to a service agency and contacting the police. However, no statistically significant independent relationship was observed between increased endorsement of either headship or the preservation of marriage and couples counselling by clergy. Repeating findings from an earlier cross-denominational Australian sample , Anglican male clergy were more likely than their female counterparts to counsel couples.
At the broader level of shaping a culture that recognizes, responds to, and prevents DV, some relationships between norms and actions were observed. Among female clergy, but not among males, greater opposition to headship predicted increased frequency of speaking publicly or preaching about DV in church services or other events. Female clergy who were neutral about headship spoke less frequently than male clergy. However, female clergy who strongly opposed headship spoke at a similar frequency to males. Among male clergy there was no relationship between attitude to headship and speaking publicly. Support for headship negatively predicted church actions that connect with DV organizations. Clergy who were more supportive of headship were less likely to report that their church donated to a DV organization or worked with such an organization, refuge, or safe house. It is widely accepted that building bridges between churches and secular DV services is best practice for supporting those who experience DV (Dyer 2016;Nason-Clark 2009;Vaughan et al. 2020). It appears to be the churches whose clergy are less supportive of headship that are most proactive in this regard. Encouragingly, a positive bivariate relationship was observed between clergy support for headship and the display of posters or leaflets about DV in the church, but it did not persist in the multivariate analysis.
Again, in line with earlier cross-denominational findings , male clergy were more likely than their female counterparts to engage with perpetrators. The positive relationships observed at the bivariate level between perpetrator engagement and both headship and the sanctity/preservation of marriage were due to the association between endorsement of these norms and clergy gender-the norms did not persist in the multivariate prediction of perpetrator-focused action.
The strength of the relationship between DV actions and norms should not be overstated. The results are weak, with little variance in clergy or church action explained. Social psychological theories of human behavior posit a wide array of influences on action, including normative influences of various kinds (both personal and social), values, attitudes, notions of behavioral control, habits, and more (e.g., Ajzen 1991;Cialdini et al. 1991;Schwartz 1977;Stern et al. 1999;Verplanken 2006). It is therefore not surprising that broad norms about marriage appear to weakly influence the take-up of DV actions.
Moreover, while this study found negative relationships between endorsement of headship and actions that focus on victim/survivor safety, we do not claim that clergy who hold to headship are less concerned about DV or about the safety and wellbeing of those who experience it. Almost all clergy in our sample felt that DV is a problem that clergy should address, and that victim/survivor safety is paramount. Public statements and policies from the Anglican Church of Australia in recent years represent investments in a church culture that has heightened awareness of DV and a commitment to improve practices to support those affected by it, such as the ten commitments for prevention and response to domestic and family violence in the Anglican Church of Australia (ACA 2021).
Nevertheless, our findings do indicate that there is work to be conducted by and with clergy who hold complementarian views, and male clergy especially, to address the implications of their actions for the safety and wellbeing of those who experience violence.
Those who hold to headship should be encouraged to have a heightened awareness of the importance of actions focused on victim/survivor safety, especially in connection to specialized DV support services, and equipped to increase their efforts in this regard. As argued elsewhere, education is needed to ensure that clergy, particularly male clergy, are aware that couples counselling is problematic . Additionally, given that directly pastoring perpetrators has the potential to negatively impact the perpetrator's partner, the church, and the clergyperson themself (Nason-Clark et al. 2017), and that little is known about the effectiveness of approaches in holding perpetrators in religious settings to account and stopping them from continuing their abuse (Vaughan et al. 2020), support is needed for clergy, especially male clergy, who do so.
In contrast to headship, the sanctity/preservation of marriage was not determinative of the take-up of any DV actions. It is possible that this might be due to insufficient attitudinal variation towards this norm in the sample; a large majority of respondents either strongly agreed or agreed that marriage is a sacred covenant that is always important to preserve, a norm that we expect would hold in other church contexts as well as our Anglican one. It might also be that such support poses less of a risk than more punitive assertions about the "sinfulness of divorce", as some other studies framed it (Zust et al. 2021). Future work should investigate such possibilities.
Our study is constrained by the quantitative survey methodology and the measures used therein, as well as by the nature of our sample.
While the study shows an association between support for the norm of headship and various DV actions, it cannot tell us how and why these associations occur nor demonstrate causation. Neither can it tell us about discourse, that is, how clergy talk about this norm in the context of DV. For example, how are complementarian clergy talking about headship when they preach or counsel a DV victim/survivor or perpetrator? These are limitations of our methodology. It is up to other studies to further interrogate these topics, and indeed, we invite clergy practitioners themselves to reflect on our findings.
The study relied on self-reports of behaviors. In terms of responses when dealing with specific DV situations, clergy were asked only which responses they used, but not when or how often they had used them-this was a blunt instrument. Clergy were also asked to report on particular actions that their church had undertaken, again, with no timeframe specified. Are clergy reliable informants of their church's actions? Anglican clergy who are newer to their parish may not be so. Moreover, the examination of the relationship between clergy endorsement of marriage norms and church actions presupposes a fit between the clergy's position on such norms and their parish and/or the influence of clergy in their parish as shapers and propagators of culture.
The National Anglican Family Violence Project Clergy and Lay Leader Survey was optin, and the sample cannot be assumed to be representative of Australian Anglican clergy on the whole. As described above, it is likely that participation was biased towards clergy with an interest in the topic. The proportion of participants in the present study who dealt with specific DV situations was similar to Anglican senior parish leaders (senior clergy and lay people serving as principal leader) as reported in a wider cross-denominational survey of clergy in which DV was a very minor topic . However, among those who dealt with DV situations, greater proportions in the present study referred victim/survivors and perpetrators to support services, and a smaller proportion counselled couples. While it is possible that these differences represent a genuine shift over time in clergy actions, with less than four years between the two studies, it is more likely that the present sample was biased. The theological diversity of the sample was a strength, enabling a fruitful examination of covariation in actions and attitudes towards norms and actions, but the study should ideally be repeated with a wider sample. Moreover, studies in other denominations would test the generalizability of our findings beyond the Anglican Church.
The limitations of the present study point toward the importance of further research with wider clergy samples, both in the Anglican Church in Australia as well as other denominations. Sharper operationalizations of both DV actions and norms concerning the ending of marriages are also recommended. Furthermore, studies that take a quantitative approach should look beyond norms that pose dangers for victims/survivors to norms that are empowering (e.g., marriage as covenant; . Cultures can and do shift. For example, a US study showed a marked decrease in support over 10 years among churchgoers for the notion of couples counseling being the best solution for a violent marriage (Zust et al. 2021). Our study provides insight into relationships between norms and actions at a particular time. In line with an increased focus over the last decade or so to reduce violence in Australia (e.g., DSS 2022;Our Watch et al. 2015), attitudes to DV, including the driver that is gender inequality, are gradually improving (Coumarelos et al. 2023). Given the increasing awareness in churches to the realities of DV and violence against women in Christian contexts, and the imperative of reducing violence and responding effectively when and where it occurs, it is important to continue to research experiences of and responses to violence.

Funding:
The National Anglican Family Violence Project (NAFVP), completed in 2021, was funded by the Anglican Church of Australia General Synod, through its Family Violence Working Group https://anglican.org.au/our-work/family-violence/ (accessed on 13 April 2023). Preparation of the present paper did not receive additional funding.

Institutional Review Board Statement:
The study was conducted in accordance with the National Statement on Ethical Conduct in Human Research and approved by the Human Research Ethics Committee of Charles Sturt University (Ethics Register Number H19306, approved on 1 November 2019).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
A description of the anonymised NAFVP Clergy and Lay Leader dataset is hosted on the Charles Sturt University (CSU) library website, together with a downloadable dataset request form. The dataset and associated codebook are stored securely on NCLS Research's servers, as agreed with the CSU library. The Anglican Family Violence Working Group considers and approves dataset requests, including assessment process and conditions of approval. NCLS Research will deliver datasets following the approval of requests.