Work with men to end violence against women: a critical stocktake

Abstract

This paper provides a critical assessment of efforts to involve men in the prevention of men's violence against women. Although there is a substantial evidence base attesting to the effectiveness of at least some strategies and interventions, this field is also limited in important ways. Violence prevention efforts often have focused on changing men's attitudes, rather than also seeking to transform structural and institutional inequalities. While feminist and queer scholarship has explored diversities and pluralities in the organisation of sexuality, much violence prevention work often assumes a homogenously heterosexual male constituency. Too often this work is conceptually simplistic with regard to gender. Against this background, this paper contests and complicates several assumptions that are part of an emerging consensus in men's violence prevention: first, that it is in men's interests to support progress towards non-violence and gender equality; second, that the best people to engage and work with men are other men; and finally, the strengths and limitations of inviting and drawing on ‘real men’. A critical assessment of the field's working assumptions is vital if it is to contribute to the future prevention of men's violence against women.

Introduction

It is time for a critical stocktake of efforts to involve men in the prevention of violence against women. Efforts to engage men in preventing and reducing men's violence against women, and more generally in building gender equality, are increasingly well established, signalled by new regional and international networks, conferences and campaigns, an expansion of domains of intervention, an orientation towards ‘scaling up’ and an increasing engagement with public policy (Flood 2015). In particular, it is time to assess a series of assumptions about the meaning and practice of this work that are becoming influential across the field and yet are unsupported by evidence or politically problematic.

Those who advocate that men should become active participants in ending violence against women often argue that, as part of this, men should engage in critical reflection on their own social locations and practices. In this spirit, I should acknowledge that I write in part as an ‘insider’ to men's violence prevention. Indeed, I have been something of a ‘cheerleader’ here, whether through participation as an activist in grassroots pro-feminist men's groups and national campaigns or through my research and education work (Flood 2010, 2011a). But advocacy must be accompanied by reflexive and critical assessment. This paper's assessment is guided by one overriding question: what will be most effective in reducing and preventing men's violence against women and building gender equality?

A critical stocktake is particularly timely now, as ‘engaging men’ is an increasingly common element in violence prevention work. In the USA, for example, a growing number of domestic and sexual violence organisations now have men-focused positions or resources (Macomber 2012), and a significant number of US colleges and universities include programmes focused on men (DeGue 2014). Internationally, campaigns and interventions engaging men are proliferating. The recent 2nd Menengage Global Symposium: Men and Boys for Gender Justice (New Delhi, November 2014) provided an impressive expression of this trend, with 1100 people from 94 countries participating.

This paper offers an assessment of three dimensions of the men's violence prevention field: (1) its practical relations with feminism, (2) its understandings of men and gender and (3) its approaches to engaging men. But first, what has this field achieved thus far?

Achievements

The field of efforts to involve men in violence prevention has some notable achievements. Perhaps most importantly, there is a growing body of scholarly evidence attesting to the effectiveness of particular strategies in engaging men to prevent and reduce men's violence against women.

Contemporary scholarship does document that particular interventions successfully have shifted the attitudes, behaviours and/or inequalities associated with violence against women. A 2007 international review documented 15 interventions involving men and/or boys in preventing and reducing violence (World Health Organization 2007). Of these, four were judged as effective, seven as promising, and four as unclear (where ‘effective’ entails a rigorous design and high or medium impact or moderate design and high impact, ‘promising’ entails moderate design and medium or low impact or rigorous design and low impact, and ‘unclear’ entails a limited design regardless of impact, or limited impact).

A more recent systematic review examined interventions for preventing boys' and men's violence, focusing on studies with a randomised controlled or quasi-experimental design, although it also included non-randomised studies with treatment and control groups (Ricardo, Eads, and Barker 2011). The review examines 65 relevant studies. It found that such interventions can change boys' and young men's attitudes towards rape and other forms of violence against women, and the gender-related attitudes associated with these, but evidence of their effectiveness in changing behaviours is far more equivocal. Only one of the well-designed studies demonstrated a significant impact on sexually violent behaviour, while only seven studies with strong or moderate research design demonstrated an impact on the perpetration of non-sexual violence.

Another systematic assessment of interventions aimed at heterosexually active men and aiming to produce more gender-equitable relationships found eight interventions addressing the perpetration of violence against women. Of these, three were not also in the review already discussed. All three reported declines in the perpetration of violence, but only one could be classified as methodologically ‘strong’ (Dworkin, Treves-Kagan, and Lippman 2013). It would be wrong therefore to proclaim simply that ‘engaging men to end violence against women works’, as most interventions have not been evaluated or evaluated only in methodologically or conceptually weak ways (Murray and Graybeal 2007), but it is encouraging to note that well-designed interventions can make change.

Other achievements, although without data regarding their impact, at least signal some kind of influence and momentum. The first is the fact that the notion that men should be engaged in prevention is increasingly accepted in violence prevention programming and policy. This is a feminist achievement, locating the problem of violence against women firmly with men, although also politically delicate as this paper notes later. Large numbers of men have participated in anti-violence groups, networks and campaigns, albeit with varying levels of involvement and commitment. In some countries, large masculine organisations and workplaces (including national sporting bodies, trucking companies and other entities) have supported men-focused violence prevention initiatives. Inspiring alliances have formed between women's and men's networks and organisations. Efforts to engage men in the prevention of men's violence against women show increasing global reach and mobilisation, rising community endorsement and increasing sophistication (Flood 2015).

Relations with feminism

While most violence prevention work with men shares a broadly feminist approach to gender, from its beginnings there have been feminist expressions of disquiet regarding its practice and politics, including the concerns that the development of efforts to engage men in preventing violence against women may reduce funding for women's programmes and services, dilute the feminist orientation of prevention agencies, marginalise women's voices or involve only rhetorical rather than substantive support from men. Research from the USA suggests that women's concerns have intensified as men's involvement has increased (Macomber 2012).

Anti-violence work with men around the globe often is undertaken by organisations with a broader focus rather than by dedicated men-focused organisations. A global survey of organisations that seek to engage men in violence prevention found that at least in terms of numbers of organisations, most of this work is being done by organisations with a wider agenda involving sexual violence prevention, batterer intervention, domestic violence service provision and so on (Kimball et al. 2012). Only 20% of respondents reported that engaging men in violence prevention was the sole focus of their organisation's work, and most of those solely focused on engaging men were from North America. Seventy-six percent of respondents, reported collaborations with women's organisations, although this may not mean effective partnerships or the absence of resource competition or the dilution of feminist efforts. Similarly, over one-third of registrants at the 2 ^nd Menengage Global Symposium self-identified as women's rights representatives, suggesting that much of this work is done in, by and with women's rights organisations.

The ideal or principle of accountability is widespread in gender-conscious work with men. Its practice may be more uneven, with research in men's anti-violence groups in the USA, for example, finding two problems: definitions of accountability often were absent or diverse or unclear, and the burden of policing men's sexist behaviour typically fell to women (Macomber 2012). On the other hand, two international initiatives show promise. The Engaging Men through Accountable Practice intervention provides a curriculum for engaging men in change in relation to personal and relational accountability (International Rescue Committee 2014). MenEngage, a global alliance comprising over 700 non-government organisations, country networks and UN partners, recently developed accountability standards and guidelines for its members (MenEngage 2014).

Around the world, there are few if any instances where violence prevention work with men has directly taken funding away from work with women. One could argue that directing any resources to work with men by definition takes resources away from work with women, given a limited funding pie. However, assessing the implications of this then is a matter in part of assessing their relative value and effectiveness in ending violence against women. Funding support for work with men and boys, as a proportion of all work addressing gender equality, appears to be very small. For example, direct support provided to organisations or programmes targeting men and boys by the Swedish International Development Cooperation Agency in 2012 comprised only 0.8% of its total funding for gender equality (Dover 2014).

There have been tensions between efforts to engage men and boys in preventing violence against women and girls and other feminist efforts focused on women and girls themselves. For example an international study among representatives of organisations that engage men and boys (in Africa, Asia, Europe, Oceania and North and South America) found that many spoke of experiencing suspicion from or conflict with victims' organisations and feminist and (other) women's groups. Interviewees expressed concern regarding the allocation of resources, ideological compatibility and leadership sharing (Casey et al. 2013).

Another concern is that men may ‘take over’ violence prevention campaigns. While there are international cases of men taking over programmes on gender, there are few if any documented cases of men taking over women's or feminist violence prevention campaigns. Men in communities often argue for their right to involvement in women-focused events, such as Take Back the Night marches in the USA (Kretschmer and Barber 2014). However, this demand rarely if ever comes from male anti-violence advocates themselves. While men-focused organisations are increasingly visible, especially in North America, most work engaging men in violence prevention around the globe is done by broader women's and violence prevention organisations (Kimball et al. 2012). Perhaps the greater problem here is not that men will take over, but that they will not turn up, in that few men join efforts to prevent violence against women and much of the work is done by women. In Australia, for example, while the White Ribbon Campaign is described as a ‘male-led’ effort to end violence against women, only 36% of the community events in 2013 were organised by men (L. Davies, personal communication, May 22, 2014).

A more obvious problem is that the small numbers of men who do participate in violence prevention advocacy sometimes do act in patriarchal ways. It is an article of faith in men's anti-violence work that men should strive for non-violent and gender-equitable practice in their own lives. The small number of studies among male activists and educators – nearly all from the USA, and none which are longitudinal – does find that these men do develop more anti-sexist forms of practice (Flood 2014). At the same time, this research also shows that some male activists and educators espouse stereotypical notions of their roles as protectors and defenders of women, emphasise their homosocial investments in evaluations by male rather than female peers, or respond in defensively homophobic ways to others' perceptions of their transgressions of masculinity (Flood 2014). Men may not take over entire campaigns, but Macomber's (2012) careful research among US ‘engaging men’ groups finds that some men in the movement do dominate interactions and interactions, claim unearned expertise or act in other patriarchal ways. This is not surprising given the patterns of masculine socialisation to which most men are subjected.

Male advocates in the violence prevention field may be given greater status, power and recognition than women doing similar work and rise more quickly to leadership positions (Macomber 2012). This echoes the ‘glass escalator’ effect documented among men in other feminised professions such as nursing and primary school teaching (Williams 1995). At the same time, other axes of privilege and disadvantage in any particular context are likely to intersect with such processes. Research from the USA on Black men's experience as male nurses (Wingfield 2009) suggests that the glass escalator effect in men's violence prevention may more available to white, heterosexual, economically privileged men than to other men.

While the growing emphasis on the need to involve men in stopping violence against women can be seen as a feminist achievement, it also may have had negative consequences for feminist work, in three ways. First, it indirectly may diminish the legitimacy of women-only and women-focused programmes and services, in prompting a mistaken belief that all interventions should include men. Corroborating this, some women's organisations report that they now are subject to pressure to include men (Meer 2011), although to their credit, international networks such as MenEngage have affirmed the vital importance of women's autonomous organisations and leadership. Second, an emphasis on and practice of including men in this work can fuel the invalidation and marginalisation of the expertise of women and the women's sector. Third, it can add to women's work and divert energy and focus, with women working to thank and reassure men and to manage or censor their own critical responses (Castelino 2013).

Understandings of men and gender

To what extent does men's anti-violence work embody feminist theoretical or conceptual frameworks? Most efforts to engage men in preventing violence against women share a social constructionist and feminist approach to gender. International research among representatives of organisations that engage men and boys in preventing violence against women and girls finds that they typically take as given that gender relations are relations of inequality and injustice and that the social construction of masculinity as dominant and aggressive is central to men's violence against women (Casey et al. 2013).

There are few if any examples where organisations or campaigns focused on men's roles in ending violence against women have adopted bluntly non-feminist or anti-feminist positions, arguing, for example, that domestic violence is gender-symmetrical or that women ‘provoke’ violence against women. However, while there is no systematic data on this, there are likely to be variations in the character and centrality of the feminist analysis adopted in this field and, of course, feminist understandings themselves are diverse and contested. In addition, as one might expect, representatives of organisations that engage men and boys in preventing violence against women and girls report tensions between engaging diverse men and maintaining a clear pro-feminist ideological stance (Casey et al. 2013).

There are weaknesses in the typical conceptual approaches of men's anti-violence work. In some instances, these weaknesses reflect influential understandings in fact in the violence prevention field more generally.

Not only attitudes

Violence prevention efforts often have focused on changing men's attitudes. This is evident in the articulated goals of interventions and programmes, in the kinds of strategies used in them and particularly in the outcome measures used. In a systematic review of interventions for preventing boys' and men's sexual violence, across 65 studies, outcomes in 47 studies included attitudes towards violence and in 25 included attitudes towards women, while outcomes in only 16 and 9 studies included the use of non-sexual violence and sexual violence, respectively (Ricardo, Eads, and Barker 2011, 24). Yet violence-supportive and sexist attitudes are only one determinant, albeit a significant one, of men's perpetration of violence against women. Changing attitudes does not necessarily change behaviour, and the relationship between attitudes and behaviour is both complex and bi-directional. More fundamentally, a focus on attitudes neglects the structural and institutional inequalities that are fundamental in shaping men's violence against women (Pease and Flood 2008). Yet interventions among men addressing structural-level factors are rare, with a systematic assessment of interventions aimed at heterosexually active men and with an experimental or quasi-experimental design finding only a single study (Dworkin, Treves-Kagan, and Lippman 2013).

Violence prevention efforts focused on men and boys are influenced to some degree by the public health approach to interpersonal violence that has become influential in the violence prevention field in general. Public health approaches have obvious advantages for the prevention of violence against women, given their focus on prevention, orientation towards social and collective determinants of health and wellbeing, reliance on evidence-based approaches to programme and policy development, emphasis on comprehensive and multi-level interventions, emphasis on collaborative work across sectors and integration of evaluation into prevention (Chamberlain 2008, 7; McMahon 2000, 30; Noonan and Gibbs 2009, 6s). At the same time, there are some differences in emphasis between public health approaches and (other) feminist approaches. For example, public health approaches are more likely to frame violence against women as a contributor to poor health than as a social injustice (Parks 2009). In other words, public health approaches show some orientation to violence against women in terms of its impact on morbidity and mortality, while feminist approaches show a greater orientation towards violence against women as a symptom of gender inequalities and oppressions. Public health approaches have been criticised for narrow standards of evidence, privileging the collective wellbeing of society over the interests of individuals and groups, and in practice neglecting the collective and institutional factors shaping health. Perhaps because of its feminist roots and activist orientations, however, international men's anti-violence work also frames violence prevention in terms of social justice. This is evident in both national organisations such as Sonke Gender Justice (South Africa) and influential international networks such as MenEngage.

Which men?

A second weakness in the typical approaches of men's anti-violence work, particularly in the global North, is its treatment of both men and violence as homogenous. Contemporary feminist scholarship takes as given an intersectional approach, centring on the insight that gender intersects with other forms of social difference and inequality such as those of race and ethnicity, class and sexuality. Women and men occupy multiple social locations, each positioned in relation to intersecting social divisions and inequalities (Mann 2012). Critical masculinities scholarship has taken up this same insight, with an increasing investigation of the complex interweavings of privilege and disadvantage in men's lives. With regard to violence against women, feminist attention to intersectionality is visible in both theoretical work on how conceptualise men's violence against women and in empirical examinations of the intersections of violence with particular social, cultural and political contexts, processes and populations. For example when men use violence, their social locations influence how they are treated. Male perpetrators are more likely to be held accountable and criminalised, and their crimes are more likely to be seen as linked to their ethnicity if they are poor, Black or men of colour (Russo 2001).

There is widespread recognition of these complexities in the men's anti-violence field, but a substantively intersectional approach does not seem visible in practice in countries such as the USA or Australia. Men's anti-violence scholarship so far shows very little direct or comparative assessment of the value of approaches tailored to specific populations. In one of the first and few such assessments, Heppner et al. (1999) found that in violence prevention work with racially diverse groups of men, culturally relevant interventions are more effective than ‘colourblind’ ones. There is also a small body of evidence that particular interventions can be transferred from one culture or context to another, with Ricardo, Eads and Barker (2011, 38) documenting examples of programmes developed in one country then also proving effective in other countries. One of the most substantial challenges to Anglo-centric and even racist discourses regarding men and violence has come from South Africa, with the development of a rich scholarship on these issues (Morrell, Jewkes, and Lindegger 2012).

Although there is little systematic knowledge of how to address intersecting forms of gender, class and race in working with boys and men, there is at least recognition among the leaders of men's anti-violence organisations of the challenge of intersectionality. In interviews with 29 representatives of organisations that engage men and boys in preventing violence against women and girls, based in countries in Africa, Asia, Europe, Oceania and North and South America, they emphasised the ways in which poverty, racism, migration, food insecurity and other issues complicate the conceptualisation, implementation and prioritisation of engaging men in violence prevention (Casey et al. 2013). Racism, poverty and other factors push issues of men's violence against women to the margins, make it harder for men to become and remain involved, limit the sustainability of programmes and unsettle strategies focused on asking disadvantaged men to critically evaluate their power and privilege.

Where the treatment of men as homogenous in violence prevention may be most obvious is in relation to sexuality. While feminist and queer scholarship has explored diversities and pluralities in the organisation of sexuality, much violence prevention work often assumes a homogenously heterosexual male constituency. Although there is widespread recognition of the centrality of homophobia to the construction and policing of masculinity, it is harder to tell if there is widespread acknowledgement of actual diversities in men's sexual practices and relations. The terms ‘men’ and ‘masculinities’ themselves are taken for granted in the field, despite increasing scholarly and activist emphases on the fact that sex and gender may not be synonymous. As a consequence, as Murphy (2010) noted, for example, for a US national conference for campus-based men's gender equality and anti-violence groups, essentialised notions of gender remain dominant and there is very little attention given to transgender people.

The language of ‘men as partners’ emerged in the 1990s in the fields of sexual and reproductive health, violence prevention and elsewhere, with men described as allies with women in working to end men's violence or in supporting women's reproductive health and promoting gender equality. It rapidly became a pervasive way of articulating men's roles. While I have earlier argued that the language of ‘partnership’ is valuable politically in emphasising men's alliances with and accountability to women (Flood 2003), it also has several dangers. It may erase women's and feminist work and leadership in the field of concern, like the broader emphasis on engaging men per se (Meer 2011). It may give the false impression that equal numbers of men and women are doing the work and taking action, whereas typically the number of active men is small. Returning to the issue of sexuality, the framing of ‘men as partners’ also risks heterosexism, if one often assumes that all men have female partners and that they can be motivated to a concern about violence against women through their intimate relations with women and girls. Yes, this is one way to motivate men's concern for violence against women and girls, but ideally it is complemented by or moves to a wider concern with the rights of women and girls in general (Manjoo 2014).

Good men, bad men

A wider issue concerns the ways in which men's engagement in violence prevention sometimes is framed with regard to men and masculinities. It should be said at the outset that the evidence is that interventions aimed at men are more likely to be effective if they explicitly address masculinity – the practices, norms and relations associated with manhood (Dworkin, Treves-Kagan, and Lippman 2013; Fulu, Kerr-Wilson, and Lang 2014). Such interventions are one example of ‘gender-transformative’ efforts that seek to transform gender norms and relations.

The interrogation of masculinity is a valuable part of the men's anti-violence field. At the same time, research documents some problematic framings of men and masculinities among US and Australian participants that may also have wider currency. This section focuses on male anti-violence advocates' articulations of their own gendered positions and practices, while a later section of the paper addresses the field's use of ‘real men’ in seeking to engage other men.

First, Macomber's (2012) US research found a pervasive distinction between ‘masculinity’ and ‘men’, allowing a critique of sexism and violence as a problem of ‘traditional masculinity’, while minimising men's potential discomfort. While this rightly identifies that violence is grounded in the socially produced ideologies and structures of gender, it can also lessen attention to men's violent practices. Second, participants made simplistic distinctions between two types of men: between ‘well-meaning men’ or ‘men of conscience’ on the one hand, and those ‘other’ men who assault women on the other. Similarly, writing on Victoria, Australia, Castelino (2013) reports that some men involved in violence prevention position themselves as ‘good men’, offering a simplistic dichotomy between perpetrators and good men. Such accounts may risk placing male advocates' relations with women beyond critical assessment and neglect the privileges and entitlements that men in general receive in patriarchal contexts (Castelino 2013). In addition, male participants' emphasis on the point that ‘not all men use violence’ can involve a focus only on obvious physical and sexual forms of violence and not also on other forms of coercion or violence-supportive attitudes and relations. Third, male activists may position themselves as virtuous men – as strong or courageous or bold – with a pervasive emphasis on them acting ‘as men’. Such framings may be useful in fostering positive, anti-patriarchal identities and in mobilising support and commitment and may represent the strategic use of common constructions of masculine strength. At the same time, they may place male activists beyond scrutiny, intensify problematic investments in gender identity, discourage critical self-examination and exclude women.

Engaging men: contesting an emerging consensus

As the field of violence prevention and work with men and boys has become established around the world, a number of problematic assumptions have become visible. While it is premature to declare that these assumptions represent a consensus in the field, they are evident in public discourse and documents of various men-focused interventions and organisations, particularly in North America but also elsewhere. These assumptions are problematic because they are based on shaky evidence, have potentially dangerous effects or should be articulated more carefully. The first is that it is in men's interests to support progress towards non-violence and gender equality.

Men will lose

The idea that men will benefit from the reduction or prevention of men's violence against women, and more broadly from progress towards gender equality, is a common element in appeals to men in violence prevention. The notion of benefit to men is visible, for example, in various overviews or background documents on men and violence prevention, typically in terms of whether men are constrained by dominant constructions of masculinity or the ‘costs of patriarchy’ (Ferguson et al. 2005; Kaufman 2003; Lang 2003; Minerson et al. 2011). Such a perspective is given routine emphasis in the wider field of engaging men in building gender equality, as shown, for example, at the recent UN Commission on the Status of Women in March 2015 (Anderson 2015).

There are few if any violence prevention interventions or organisations aimed at men that show a single-minded focus on the benefits to men of non-violence and gender equality. Still, an emphasis on benefits to men should avoid downplaying the patriarchal organisation of gender and violence and thus the actual obstacles to change. In the first instance, men who use violence against their partners or other women benefit directly from this. When men engage in the coercive micro-regulation of their partners' lives, which Stark terms ‘coercive control’, this ‘affords benefits that can be measured in increments of resources garnered, personal service, sexual exclusivity and subjective reinforcement of gender identity’ (Stark 2010, 207). Men's use of coercive control exploits persistent gender inequalities, and also works to maintain them. Similarly, interviews with convicted rapists suggest that men who rape often see rape as a low-risk, high-reward act, a means of revenge and punishment, an added bonus while committing other crimes, a way to gain sexual access, a source of impersonal sex and power, a form of recreation and adventure, and a source of male bonding (Scully 1990). If it is true that men's violence against women has the social consequence of reproducing men's authority over women (Eisenstein 1984), then there is a sense in which men in general benefit from this violence.

In some ways then, men will ‘lose’ from progress towards non-violence and gender equality. Some men will lose the benefits they accrue from the perpetration of violence, while others will lose the unjust and unearned privileges associated with gender inequalities. Efforts to involve men in violence prevention must acknowledge the costs to men of undermining the patriarchal privileges that underpin men's violence against women. They should also acknowledge the potential costs of involvement in violence prevention itself, given that the men and boys who participate may be ridiculed or harassed for lack of conformity to hegemonic masculine norms (Crooks et al. 2007).

At the same time, it would be a mistake to avoid all reference to how men may benefit from a non-violent future, to portray progress necessarily as a zero-sum game in which men will lose and women will gain and to appeal to men purely on altruistic grounds. To end men's violence against women, we will need to secure the support of at least some men, and, to do that, we will need to appeal, in part, to the ways in which they will gain. Yes, the overarching reason for men to support an end to violence against women should be ethical, moral or political – this violence simply is unjust. But we should also appeal to men's reconstructed or anti-patriarchal interests – and the investment that some men already feel in freer, safer, more egalitarian lives for women and girls.

Patterns of male privilege, alongside other intersecting forms of privilege and injustice, are part of the landscape in which men's and boys' engagement in violence prevention takes place. Those who work with men encounter tensions in asking members of a privileged social group to critically interrogate their privilege, for example by examining their deeply held beliefs about being a man (Casey et al. 2013). A persistent challenge in this work is how simultaneously to invite and involve men on the one hand, and avoid colluding or reinforcing male privilege on the other. Some common ways of inviting men into violence prevention are based on complicity with notions of ‘real’ manhood or stereotypically masculine attributes, as I note later. Institutionalised male power – in governments, criminal justice systems, religious institutions, communities and so on – poses a wider challenge for this work. Representatives of prevention efforts report that their challenge to entrenched gender inequalities, and men's defences of these inequalities, generates attacks and ridicule and makes it harder to gain resources, legitimacy, support and membership (Casey et al. 2013).

Men should listen to women

A second common assumption in the violence prevention field is that the best people to engage and work with men are other men. There are two sides to this: an emphasis on all-male groups, and an emphasis on male educators and trainers. As mentioned earlier, this paper is guided very much by a pragmatic question, of how best to engage men to prevent violence against women and build gender equality. The non-academic and ‘grey’ literature on violence prevention education among men tends to emphasise the need for male-only groups and, certainly, there are pedagogical and political advantages to single-sex groups (Flood, Fergus, and Heenan 2009). However, evidence regarding the merits of single-sex versus mixed-sex groups is more mixed. Some reviews find that separate-sex programmes are more effective than mixed-sex programmes (Vladutiu, Martin, and Macy 2011), while others report more varied patterns of change (Ricardo, Eads, and Barker 2011; Clinton-Sherrod et al. 2009). Anderson and Whiston's (2005) review of 69 education programmes for university students on sexual assault found little evidence that men are more likely to benefit from programming administered in all-male groups compared to men in mixed-sex groups.

The most effective sex composition of groups may depend on such factors as the age of the group, the focus and goals of the teaching sessions and the nature of the teaching methods used. Mixed-sex groups may be more effective if the programme or session is intended, to encourage male empathy for females or for victims of violence, to create gender dialogue or to create opportunities for males to listen to females. On the other hand, if the programme or session is intended, for example, to encourage men's ‘ownership’ of the issue or to facilitate their move from bystander to ally, then single-sex groups may be more effective. There is some argument for using different sex compositions at different points in one's education programme (Flood, Fergus, and Heenan 2009). Finally, there has been some recent discussion of ‘gender-synchronised’ approaches, which coordinate work across the logics, participants or strategies in men's and women's programmes (Flood 2012; Greene and Levack 2010).

There is in the field a complementary emphasis on the need for work with boys and men to be conducted by male educators in particular. For example, in a study of 25 men active in all-male anti-rape prevention groups on 11 US campuses, participants said that one element that made these groups effective was that they are delivered by men (Piccigallo, Lilley, and Miller 2012). Again, the use of male educators has particular pedagogical and political advantages (Flood, Fergus, and Heenan 2009). However, female facilitators can work very effectively with boys and men, and there are benefits to women and men working together. Such partnerships demonstrate to participants a model of egalitarian working relationships across gender; they model women's and men's shared interest in non-violence and gender justice; they give men opportunities to hear of women's experiences and concerns and to further mobilise their care for the women and girls in their own lives; and they enhance accountability to women and women's services. In addition, simplistic assumptions about ‘matching’ educators and participants, for example by sex, may not address the complex interactions and negotiations that take place regarding a range of forms of social difference, from age and ethnicity, to class and sexuality. Indeed, sharing a biological sex is no guarantee of individuals' compatibility, given women's and men's diverse gender identities and relations.

Again, what works best? There is little robust research evidence in the violence prevention field regarding the effectiveness of matching educators and participants by sex. In relation to violence prevention, there is anecdotal evidence that men will listen more readily to other men than to women, with men in all-male anti-rape prevention groups on US campuses sharing the belief that men are more receptive to hearing anti-rape messages from other men than from women (Piccigallo, Lilley, and Miller 2012, 513). Research in higher education documents that male teachers addressing gender issues are evaluated by students as less biased and more competent than female teachers (Flood 2011b). These findings suggest that using male educators with men may be advisable, although it would be valuable to know whether using male educators makes a difference to generating men's initial participation and engagement, lessening their violence-related attitudes or reducing actual perpetration. On the other hand, various studies find that many men's initial sensitisation to the issue of violence against women was fostered in particular by listening to women and women's experience (Casey and Smith 2010; Piccigallo, Lilley, and Miller 2012).

This is a good example of a wider tension in violence prevention work with men, between meeting men ‘where they are’ on the one hand, and seeking to transform the gendered identities and relations among men which sustain men's violence against women on the other. If men will listen more readily to men, then violence prevention efforts among men may, understandably and pragmatically, rely on male educators and leaders. Yet men's greater willingness to listen to other men also reflects men's homosocial investment in evaluation by male peers and the social marginalisation of women's voices and experiences, and both can feed indirectly into violence against women (Schwartz and DeKeseredy 1997). Thus, prevention efforts also should find opportunities to foster men's cross-gender investments and attention to women's voices. This brings us to a third problematic emphasis in violence prevention work with men, on the use of ‘real’ men.

Use diverse men

Education and marketing interventions aimed at men and boys sometimes make use of ‘real’ men, men who embody at least some of the qualities associated with hegemonic masculinity. Face-to-face educational programmes on university campuses may use male leaders in sports or fraternities (all-male residences on campuses) as peer educators, while social marketing campaigns often rely on male sporting celebrities. Such men are seen as ‘bell cows’, able to lead other men into this work because of their conformity to gender norms (Murphy 2010). Men in anti-rape groups on US campuses reported that peer recruitment to these groups was facilitated by the presence and profile of men who were ‘really cool … the kinds of guys you really looked up to … guys’ guys' (Piccigallo, Lilley, and Miller 2012, 517).

The use of admired peer leaders or opinion leaders is well established in various health promotion fields. While it has an obvious logic in violence prevention as well, the evidence for its effectiveness is weak. In fact, a 2005 meta-analysis of 69 studies of sexual assault education on US college campuses found that professional presenters were more successful than either graduate students or peer presenters in promoting positive change (Anderson and Whiston 2005), while another found that professional-facilitated and peer-facilitated programmes differ in their domains of effective impact (Vladutiu, Martin, and Macy 2011).

In violence prevention with men, the use of ‘real’ men (whether as peer educators, professional facilitators or marketing spokespeople) embodies the same tension noted above. An exclusive reliance on men who conform to (many) gender codes may lessen the effectiveness of this work, by compromising long-term impact for short-term appeal. Men's violence against women is sustained in part by rigid gender codes, the policing of manhood and rigid constructions of a gender binary between masculinity and femininity, men and women, and male and female. As Murphy (2010, 106) asks:

what contradictions are involved when males who already do masculinity ‘normally’ ask other males to do masculinity ‘differently’ (i.e., embrace non-violent masculinity)? What message is subtly conveyed about gender variance when only gender-normative males are deemed competent to persuade other males to transgress gender norms? What incentive is there to do gender differently if we continue to reward, celebrate, and affirm only the gender conforming? What real-life lessons, practical tools, and survival skills do gender-normative males have to offer on the topic of gender variance?

Violence prevention efforts among boys and men at times should affirm and promote men who do not fit dominant codes of masculinity: ‘girly’ men, gay men, ‘sissy’ men and transgender men. They should break down powerful gender binaries. They should ‘turn up the volume’ on the actual gender and sexual diversity in men's lives, including the fact that many heterosexual men experience and demonstrate ways of being outside hetero-normative masculinity (Heasley 2005). At the same time, interventions targeted at general populations of men should not be so determined to challenge existing constructions of gender that they fail to engage the majority of their audience.

Remove, don't only reconstruct

A fourth assumption is that the goal is to encourage new, positive masculinities among men. Overlapping with the use of actual ‘real’ men as peer leaders and educators is an explicit appeal to ‘real’ men in some marketing campaigns, where audiences are told for example that ‘real men don’t use violence'. More subtly, some initiatives appeal to stereotypically masculine qualities, while simultaneously seeking to redefine them. The most prominent example of this approach is Men Can Stop Rape's social marketing campaign, ‘My strength is not for hurting’, which encourages norms of sexual consent among young men, while its more recent bystander intervention campaign encourages men to be ‘the kind of guy who takes a stand’.

There is an obvious logic here, an effort to undermine the socially produced association between violence and masculinity by re-defining masculinity as non-violent. More pragmatically, appealing to men's investments in manhood may be done in the hope that it will produce greater interest and engagement. I have argued earlier that violence prevention campaigns aimed at boys and men should speak to questions of identity, offering males new ways to negotiate and perform gendered identities (Flood 2002–2003). At the same time, campaigns also should actively encourage men's disinvestment in gendered identities and boundaries. One prominent advocate here is John Stoltenberg, who recently argued that ‘talking about “healthy masculinity” is like talking about “healthy cancer”’ (Stoltenberg 2013). He notes that ‘Real men don't …’ campaigns obscure the actual links between manhood and men's violence against women: that men rape or hit women in order to experience or demonstrate themselves as real men; that maintain men's anxieties about their successful performance of gender; and that keep ethical choice-making locked into gender identity. Violence prevention work among boys and men should seek not only to challenge the dominant cultural meanings given to manhood, but the gender binaries and hierarchical policing of gender that complement them.

There is, in contrast to appeals to ‘real men’, a prominent form of intervention among men which invites men to adopt a stereotypically feminine behaviour in the service of ending violence against women. At Walk a Mile in Her Shoes marches, men wear stereotypically female shoes, typically with high heels, and literally walk a mile, to show their concern about violence against women. (See http://www.walkamileinhershoes.org/.) Even this feminine behaviour, however, is vulnerable to co-option by wider patriarchal dynamics. As Bridges (2010) emphasises, drag (the explicit performance of gender for an audience) does not automatically produce politically progressive understandings of gender, and what counts as a progressive gender and sexual transgression is shaped by both its context and its performance. In the marches, the men's actual use of drag involved the playful disruption of gender boundaries in ways which marked these performances as temporary and inauthentic, reinforcing rather than destabilising gender boundaries. Their performances also showed a homophobic avoidance and renunciation of challenges to heterosexuality (Bridges 2010).

Negotiating the complex terrains of gender and sexuality in involving men in violence prevention is a complicated business. On the one hand, if men's engagement with feminist activities and issues is framed as a gender transgression, a kind of drag, then this may suggest that men's adoption of feminist politics will be only in certain situations or in inauthentic ways and that a concern about violence against women is feminine. On the other hand, if it is ‘real men’ who refuse violence against women and if men are told to strive for a non-violent masculinity, then this may collude too much in gender itself.

Change men by empowering women

The final assumption contested in this paper is that changing men necessarily involves working with men. While there is a powerful feminist rationale for working with men, changing men may be best achieved in some circumstances by engaging and empowering women and by focusing on transforming inequitable gender relations. Among programmes and organisations focused on engaging men in ending violence against women, while there is unanimous support also for work with women, it seems to be widely assumed that the most effective way to change men's attitudes and behaviours is to work directly with them. Instead, changing men also can be achieved by working with women, and by shifting the wider conditions within which men make choices about violence and non-violence.

Educating women can change men: by shifting women's expectations of partners and intimate relations, interventions may increase the pressures on and incentives for heterosexual men to adopt non-violent practices and identities. Interventions can harness men's motivations to be accepted and liked by women by encouraging women's unwillingness to associate with sexist and aggressive men. As Adams-Curtis and Forbes (2004, 114–5) argue, ‘When women as individuals and as a group loudly and publicly refuse to associate with misogynistic individuals and groups … traditional sexual scripts will change.’ Yes, it is unfair to ask women to play a role in changing the behavior of men, but it is no more unfair or damaging than the consequences of current gender relations (Adams-Curtis and Forbes 2004).

There are other strategies that also ‘force’ men to change. Violence prevention efforts should include efforts to change the structural and institutional conditions within which men make choices about how to behave. They should change the actual structures of costs and benefits, and not just men's calculations of them. One obvious example of such an effort is to increase the criminal justice system's policing and punishment of men's violence against women. Other strategies include empowering women, decreasing their economic dependence on men, shifting workplace and sporting cultures, and changing laws and policies. Although interventions among men addressing structural-level factors are rare (Dworkin, Treves-Kagan, and Lippman 2013), there are some encouraging examples. The Mobilising Men programme, working in countries such as India, Kenya and Uganda, seeks to change systemic and structural gender inequalities (Greig and Edström 2012), while Sonke Gender Justice, among other strategies, works to hold the South African government accountable for the adoption and implementation of policies regarding gender-based violence (Sonke Gender Justice Network 2009). Moreover, the 2 ^nd Menengage Global Symposium showed an increasingly prominent attention to the structural and transnational forces that shape gender inequalities and both personal and collective forms of violence.

Engaging men has become almost a routine element in efforts around the globe to reduce and prevent men's violence against women. This is a significant feminist achievement, feminist because it embodies the fundamental recognition that violence against women is a problem overwhelmingly for which men are responsible and that men must join with women to address. The field of men's violence prevention shows increasing sophistication in its intervention strategies, institutionalisation in organisations and regional networks, and a growing evidence base (Flood 2015). As it develops, however, it risks the uncritical adoption of some taken-for-granted truths that are inaccurate, dangerous or simplistic. A critical assessment of the field's working assumptions is vital if it is to make progress in reducing and preventing men's violence against women.

Disclosure statement

No potential conflict of interest was reported by the author. The author did not receive any financial interest or benefit arising from the application of the research with which this paper is associated.