‘It’s a work in progress’: men’s accounts of gender and change in their use of coercive control

Over the past ten years the theoretical framework of ‘coercive control’ has been increasingly applied, critiqued and now underpins a criminal offence. While many argue that it more accurately reflects experiences of victimisation, there has been little exploration of coercive control through the accounts of perpetrators. Through two phased interviews with 64 men attending UK Domestic Violence Perpetrator Programmes, we examine how and why men use coercive tactics and how unpicking gender norms enabled some men to recognise and reduce their use of coercive control. We argue that coercive control is more dynamic, contestable and open to change than previous research has suggested. Some men did manage to take steps away from investing in traditional masculine norms and reduce their use of coercive tactics. However, this was an uneven and contradictory process which took time – involving painful realisations of loss and harm alongside a discovery of the benefits associated with letting go of restrictive gender norms. Understanding how and why men invest in or dismantle gender norms that underpin coercive control has important implications for theory and for practice, particularly the content and focus of work with domestic violence perpetrators.

To cite this article: Downes, J., Kelly, L. and Westmarland, N. (2019) 'It's a work in progress': men's accounts of gender and change in their use of coercive control, Journal of Gender-Based Violence, vol x, no x, 1-16, DOI: 10.1332/239868019X15627570242850

Introduction
The perpetration of domestic violence by partners and ex-partners, also known as intimate partner violence, is a complex social problem which, despite its prevalence, is experienced as highly specific, isolating and personalised. Feminist activism and scholarship has succeeded in gaining recognition of domestic violence as a form of systemic global discrimination against women (Westmarland 2015). Much theorisation of intimate partner violence locates it within gender-based inequalities, with intersections of race, class, sexuality, age, religion/faith and disability, which influences both the character of abuse itself and the options available to interrupt and escape it. The United Nations has argued that it is both a cause and a consequence of gender inequality, making its prevention at the global level synonymous with creating gender equality. Policy and practice responses have been developed to support women victimsurvivors 1 and challenge male perpetrators, including Domestic Violence Perpetrator Programmes (DVPPs), also known as Batterer Intervention Programmes (BIPs) and Men's Behaviour Change Programmes (MBCPs).
The study reported on here involved two phased interviews with men (and female partners, discussed elsewhere, for example, Kelly and Westmarland 2015). In these interviews, questions of how and why men use control and coercion was explored, and the data also reveals some of the contradictions and complexities that men face when asked to confront their actions within DVPPs. This paper hence fills a gap in understanding men's use of coercive control through the voices of men themselves. DVPPs can provide space, resources and opportunities for men to disrupt gender at the levels of identity, interaction and social structure (Anderson 2009). We argue that coercive control is more dynamic, contestable and, crucially, open to change than previous research has suggested. Change can happen, but in a less straightforward way than for physical and sexual violence. For coercive control, change is interlinked with the letting go of deeply held, restrictive gender norms, changes can be uneven and contradictory, involving both realisations of loss and of the benefits of letting go of gender norms.
In this article, we start by outlining Stark's (2007) concept of 'coercive control' and how the analysis at its heart has been extended; we then outline previous research on men's accounts of domestic violence. Next, we describe our study and research methods used. This is followed by analysis of how and why men used coercive tactics to restrict the freedom of women ex/partners. Where men took steps to change, this expanded women and children's 'space for action' (Kelly and Westmarland 2015). We conclude with the implications of our findings for DVPPs.
Coercive control: the gendered micro-regulation of everyday life Stark (2007) uses the concept of coercive control to challenge what he sees as an increasing reduction of domestic violence to incidents of physical violence. Feminist domestic violence research and practice has, from the outset, connected men's assaults on women partners and the use of power and control in relationships (see, for example, Dobash and Dobash 1979). More recent research on domestic violence homicides suggests that there may be a group of men who primarily use coercive control (Dobash and Dobash 2015), but the knowledge base here is less developed than on that which includes physical and sexual violence. The concept of coercive control seeks to illuminate how bespoke practices of control serve to entrap women and limit their freedom within heterosexual relationships (Stark 2007). Tactics of control involve the gendered micro-regulation of everyday life including when and what a partner can eat and drink, how they dress and style their hair, how they undertake household tasks, who they can spend time with, how they act around their family and friends, what they watch on television, how they drive, where they can go, what they can talk about, when and where they can sleep, how and when they have sex and how they can spend their free time. Imagined, suspected infidelity is frequently used to justify domestic violence, with jealous surveillance a common backdrop to controlling behaviours (Stark 2007). Victim-survivors make numerous accommodations, including providing access to their private email, mobile phone, social media, home, schedules and body, to demonstrate their loyalty. Stark argues that coercive control undermines women's autonomy and, therefore, should be considered a 'liberty crime'. Crucial to Stark's analysis is that these tactics are effective precisely because they take place within a context of gender inequality in heterosexual relationships, which serves to normalise male control, making abuse difficult to see and name.

Space for action: women and children's freedom
The core challenge Stark made to both researchers and practitioners was to move beyond considerations of safety; to take into account women and children's freedom. This has echoes in the work of European feminists who have interrogated the relationship between violence, gender and power. Eva Lundgren (2004) argued that women's 'life space' was a key site for men to increase their control: both this and women's active adjustments set limits in a woman's life. Lundgren (1998Lundgren ( , 2001Lundgren ( , 2004 analyses how women experience a restricted 'life space' as violence targets a woman's inability to embody the perpetrator's definition of what a woman should be. Drawing on this work Kelly (2003) theorised that while men could increase their 'space for action' through violence, for women, theirs is constrained, not just through the actions of the perpetrator, but also through the failure of the wider community, including agencies to effectively intervene. The concept of the 'abusive household gender regime' developed by Anne Morris (2009) also demonstrated how perpetrators drew on constructions of gender to exert power and control over women and children.
These theorisations document what many victim-survivors and practitioners already knew -that it is the small, seemingly insignificant, actions which, when considered together and reinforced by gendered stereotypes, can be experienced as a 'cage'. The harmful long-term impacts of the imposition of an abuser's 'unreality' upon women and children has been documented (Williamson 2010). This analysis resulted in the creation of a criminal offence of 'coercive and controlling behaviour in an intimate or family relationship' in England and Wales (Home Office 2015) and the introduction of the Domestic Abuse (Scotland) Bill (Scottish Parliament 2018). In one sense, therefore, feminist understandings of domestic violence as a pattern of everyday actions, have achieved a degree of recognition, albeit in a gender-neutral law and without explicit links to gender inequality.
Nonetheless, the concept of coercive control has been criticised, with a number noting the difficulties and challenges that it raises within an array of contexts including activist and advocacy work, public policy, statistical data, practitioner responses, legal practice and public awareness (Arnold 2009;Hanna 2009;Libal and Parekh 2009;Polletta 2009;Burman and Brooks-Hay 2018;Tolmie 2018;Robinson et al 2018;Walby and Towers 2018). To date, most work has focused on women's accounts of coercive control perpetrated by men (Lundgren 2004;Stark 2007;Wuest and Merritt-Gray 2008;Morris 2009;Williamson 2010). Indeed, one criticism of Stark is his failure to examine what it is that motivates some men to use coercive tactics against their partners and ex-partners (Brush 2009). Our data addresses this gap.

Understanding men's accounts of domestic violence
Previous in-depth research with men who have used violence against their partners has explored men's definitions, perceptions and explanations of violence, with specific attention paid to minimisation, shifting of blame and denial (Hearn 1998;Anderson and Umberson 2001;Cavanagh et al 2001;Gadd 2002;Whiting et al 2014). This body of work connects intimate partner violence to wider structures of gender inequality within which men's negotiation of traditional masculinity and use of violence sits (Hearn 1998;Dobash et al, 2000;Wood 2004). Men's use of domestic violence is linked to the affirmation of traditional masculine values that position men as providers, protectors, rational thinkers and authority figures.
These findings are echoed within rare ethnographic and observation-based studies of DVPPs that are informed by a gender analysis, which have cast doubt on the effectiveness of programmes to end men's violence (Fox 1999;Schrock and Padavic 2007). Researchers found that change was inhibited by traditional masculine tropes, such as the 'emotional inexpressive sacrificial breadwinner' (Schrock and Padavic 2007: 643). Men were reported to hijack anti-violence material to present themselves as exploited and downtrodden workers and persuade each other to use rationality to control their feelings and partners. However, both these studies are restricted to a single programme and interpretations of lone researchers, a tendency previously criticised to eclipse the perspectives of those directly involved in violence, particularly an absence of men's voices (Peralta et al, 2010).
British community-based DVPPs draw on a range of approaches in working with men who use violence against women, including cognitive behavioural, pro-feminist and psychoanalytic influences, underpinned by a gendered analysis of domestic violence (Phillips et al, 2013). Nonetheless, the evidence base on the effectiveness of DVPPs has been mixed and issues with evaluation design, including the measurement of 'success', have been highlighted (Akoensi et al, 2013). One key shortcoming of previous evaluations is a tendency to focus on a narrow understanding of 'success' -no more violence -failing to explore the uneven constitution and transformations of gender over time. Most research on DVPPs has focused on whether men change rather than the questions of how and why.
Qualitative studies that have focused on men's accounts of using coercive control have also linked coercive tactics to the affirmation and naturalisation of a traditional gender binary (DeShong 2015; Heward-Belle 2017). It was found that abusive men justified control by constructing women as unable to mother her children (Heward-Belle 2017) or be a 'good partner' (DeShong 2015: 94). This could manifest in repeated criticism of her parenting practices, appearance, mental health and substance misuse problems. Men openly acknowledged the instrumental nature of this tactic, knowing that it would sabotage the mother-child relationship, increase a woman's vulnerability and ensure that he had 'the upper hand' (Hewerd-Belle 2017: 380). Although small sample sizes, and single interviews, these studies offer crucial insights into how wider gender structures can be drawn on to normalise and inflate men's control within intimate relationships.
Quantitative findings from our wider research project (Project Mirabal) suggested that men found reducing their use of coercive and controlling behaviours harder than reducing physical and sexual violence (Kelly and Westmarland 2015). This research also demonstrated that change across different types of violence and abuse was incremental and uneven. Empirical qualitative research that explores change in coercive control is scant; only one study has considered the possibility of shifting the pattern of abusive control within a community sample of women (Wuest and Merritt-Gray 2008). Therefore, the current study advances knowledge of domestic violence based on empirical findings from phased qualitative interviews with a large sample of men in a multi-site study of DVPPs. Underpinned by a gender analysis we examine how and why men use coercive tactics and examine how unpicking gender norms enabled some men to reduce coercive control.

Research methods
As part of a wider study (Project Mirabal) we collected qualitative data via two in-depth interviews from men who had been assessed and accepted onto a community-based DVPP (as opposed to a criminal justice setting). All DVPPs in our sample were accredited members of Respect -a UK umbrella organisation for accredited DVPPs. As an integral part of these programmes, and required for Respect accreditation, women partners and ex-partners were offered support. Participants were a self-selected sample of men from 11 geographically diverse DVPPs that spanned England, Wales and Scotland. While the approaches of these DVPPs can vary, they all use group work exercises to explore content such as understanding the impact of violence, using strategies to stop violence and learning how to build a respectful relationship.
The first interview was completed within six weeks of the man's start on the programme (Time 1), the second within six weeks of the man's end date, regardless of whether he did or did not attend the programme to the end (Time 2). Men who withdrew from the programme were asked to do a separate interview to explore the reasons for their decision to withdraw. The qualitative interview schedule consisted of three sections: on the relationship and the events that led up to involvement with the DVPP; a 'critical incident' section that attempted to get close to specific examples of violence and abuse; and a 'gender' section that explored beliefs, expectations and attitudes about gender. The questions and prompts in the second interview were tailored to explore any change in ideas about gender, the relationship and/or family life, as well as if and how skills learned from the men's programme had been applied in everyday life. Interviews were mainly conducted by four research associates, three female and one male, all of whom were white British.
Sixty-four men took part in the first interview, with ages ranging from 21 to 58, with an average age of 35.8. The majority (54) were white British with five mixed ethnicity (7.8%), three Asian (4.7%), one black (1.6%) and one man chose not to specify his ethnicity. At Time 1, the abuse for 41 men involved an ex-partner, for 24 it was within a current relationship and one man had an unclear relationship status. Some men were addressing abuse perpetrated against their current partner and ex-partner/s. Most men (57) were fathers with a total of 138 children. Sample attrition (either the man declined or we were unable to re-contact) meant that 36 (56%) men completed the second interview. At Time 2 there were changes in the relationships for 14 men, with 22 remaining the same and 28 not known (as no Time 2 interview was completed). One man started a new relationship, four men moved back in with a partner, two chose to live separately and seven men separated. For those whose relationship status did not change, nine couples chose to stay together and 13 chose to remain separated.
Qualitative interviews were recorded and transcribed in full. All identifying information, such as names and locations, were removed. Pseudonyms were allocated to all participants. Transcripts were organised using NVivo 10 and codes were developed from the data using thematic analysis. Inductive coding of qualitative data was completed by one research associate. Initial themes for the 'space for action' indicator included key areas of everyday life in which men used controlling tactics such as affect, body, relations with others, time, money, home, children and movement. Each theme included numerous subthemes that captured the particular form and character that this control took in participants' lives, such as 'restricting her to the home' in movement, 'monitoring her phone' in relations with others, and 'his emotional needs come first' in affect. Initial codes were reviewed in the research team and a coding frame was finalised. Thematic data analysis, taking into account theme frequency and change over time, was then finalised by one research associate. Ethical clearance was granted by the Durham University Research Ethics Committee.

Results
In terms of the wider question of whether men attending the DVPPs changed their behaviour, quantitative data was collected from female ex/partners covering five points in time. These results are reported on in-depth elsewhere, but in summary there was positive change across all six indicators, with the most dramatic being the almost complete cessation of physical and sexual violence for female ex/partners (Kelly and Westmarland 2015). Our focus here, however, is on coercive control. Here, the quantitative data showed change to be far more limited and uneven. For instance, only marginal shifts were seen for items such as 'I feel afraid of how he would react if I got a new partner' (91% of women said this happened always, often or sometimes before the programme compared to 65% 12 months after the programme started) and 'I feel like I have to be very careful around him if he is in a bad mood' (96% before compared to 75% after). Larger reductions were seen on other items, such as 'tries to restrict where I go' (64% before compared to 21% after), 'tells me to change the way I dress or my appearance' (57% before compared to 13% after), 'tries to look at my messages and contacts' (62% before compared to 22% after), and 'prescribes or criticises the way housework is done' (55% before compared to 16% after).
While this quantitative data added to our understanding of what was happening, the qualitative data is crucial in understanding how and why men use coercive control as well as how and why some men reduced their use of coercive control. To get as close as possible to specific examples of men's use of coercive control in intimate relationships, and the ways in which men reduced women and children's 'space for action', men were asked to talk in detail about an example of a time when he tried to control the way his ex/partner did something or when he tried to stop her from doing something he did not like. Men were encouraged to describe in detail actual examples that were of particular significance to them -reducing the possibility of purely repeating things they had been told within the DVPP and giving the 'correct answer'. Prompts were used to examine men's thoughts, feelings and actions before, during and after these moments. Men were asked to provide examples before the programme started (Time 1) and after the programme started (Time 2).

Understanding men's coercive control: normative gendered expectations
At Time 1 qualitative interviews men were able to offer a multitude of examples which they understood as attempts to control their partners, most of which could be legitimated through gendered norms about heterosexual and familial relationships. Women often accommodated to these demands, but on occasion would also resist: the latter response was often the context for more explicit forms of abuse. However, the gendered underpinnings of controlling tactics could be hidden by men's talk about gender equality. Most participants talked about living in a liberal era of gender equality in which the life choices and opportunities of women were considered equal to those of men. For example, Aaron asserted that, 'If you want to go and do something do it…There [are] no gender guidelines I don't think'. In the first interview 41 men made similar remarks, with only seven making any reference to gender inequality. When asked to reflect on how gender shaped their lives many men said that they had never been asked about, or explicitly thought about, this before.
When reflecting on the wider expectations upon them as men, however, gender operated through taken-for-granted ways of being associated with traditional masculinity: investments in being a protector, a provider and a father who was the legitimate head of the household or family. For instance, Roger commented: 'I believe what's expected of a man is that he's supposed to be masculine, take care of his family, protect his family.' Similarly, Todd explained: '[what is] expected of men in a family sense I suppose would be to look after your family, provide for your family…From a society point of view…probably about being strong, not showing weakness'.
When reflecting on their expectations of women, men spoke about wanting their partners and ex-partners to conform to aspects of traditional femininity, especially being a nurturer and homemaker. Fred described how 'I just wanted her to be basically -a mother to my son and then like stay at home, do the house tidying, do the shopping, and then basically -basically look after me'. The successful accomplishment of a 'normal' and 'nice' family was considered to be a widelyaccepted social norm for many men, as Will noted 'I wanted what everyone wants, to settle down, nice family…nice job if I possibly get back to work…just everything what a normal family would want, what I had when I was getting brought up…love…a nice happy family.' This investment in traditional ideas of gender, family and intimate relationships became entwined with coercive tactics that men used to affirm their position and have needs met. Brendan reflected on how his need to accomplish a family underpinned his control: I was a bit 'come here let's protect this little family'. At the time, it was really…I genuinely wanted to do that because I craved a family. I never really had that proper family. My dad left when I was very young and I didn't want that to happen so I kind of smothered everybody and wanted to protect them.
However, as Brendan suggests, men's accomplishment of the 'ideal' or 'normal' family or relationship, could be a complex and contradictory process.

Regulation as protection: enforcing standards, routines and being right
For several men, the sense that he should, and did, 'know best' as the head of the household informed their use of coercive control. This could manifest itself in the regulation of household tasks, for instance as Ivan spoke about his enforcement of a 'better' washing up regime on his partner.
The way she washed up…the order that she washed the pots and pans and glass and things like that…she doesn't have an order…she'll literally wash things as and when she grabs them basically. With myself personally I'll have an order. I wash up the glasses first or the bread board first and the glasses. Everything that's non-greasy first and then work my way to the plates and the knives and forks and then the pots that we've actually used to cook in. That's the way I would do that and basically try ask her to do it that way and explain why…that it's a better way of doing it and yeah she wouldn't have it and we would have an argument because I wouldn't see her way and she wouldn't see my way.
When his 'better way of doing it' was not accepted and validated by other members of the family a sense of frustration led to arguments, building longer term resentment. Such expectations and standards made criticism legitimate in men's eyes, they would, however, always target her 'failures' as a woman. Grant drew attention to his partner's ineffective use of time and inability to look after her home and children properly: Well if I came in and there was a pile of dishes and loads of washing all over the floor, I would say 'you've got a big 8 kg drum washer there, well why don't you put something in it?' I: And what would she say? You're always having a go at me!' I said 'well look, you can stand there and smoke 50 tabs [cigarettes] a day but you can't do the washing?' I said 'the washer does the washing, all you do is put the powder in and switch it on'. Criticism…I was always classed as critical. If I went in and there was no food in her cupboards '[Partner] you need to go shopping' it was 'don't you tell me!' and all that. You couldn't even in the end suggest anything because it was just classed as criticism and 'what do you know?'…Well you'd look and she'd never been anywhere, she'd never done anything. And I thought well if you're doing nothing and you're going nowhere, why don't you do your housework? Tidy up this and tidy up that because with children in the house it does get messy, you know what I mean? And anything I said would just be same again 'shut up this is my house!' Here Grant positions resistance to his 'suggestions' as irrational ('you couldn't even in the end suggest anything') and uses children to demonstrate how reasonable his expectations are ('with children in the house it does get messy, you know what I mean?'). Gendered expectations of her as a mother, and thereby responsible for cleaning and cooking, were mobilised to inflate the power of his point of view even while he simultaneously undervalues her work ('the washer does all the washing'), her time ('you're doing nothing') and focuses on her negative habits ('you can stand there and smoke 50 tabs a day').
In these examples it is women's resistance to the standards being imposed that becomes a legitimation for violence and abuse. In their initial interviews many men made clear that if women had not 'wound them up', had accepted the rightness of their critique, life would have been less fraught. For some men, the fact that they had not used violence and abuse in previous relationships was evidence that there was something about 'this relationship' that was the problem -women they had been in previous relationships with had not been resistant so the associated conflict, violence and abuse had not been present.
Some men did choose to take responsibility, albeit for 'incidents' of violence, and acknowledge a contradiction between their self-identity and use of violence. Giles reflected: 'I always frowned at a guy hitting a girl like I used to think what a dick, for me to have an incident where I was that guy was horrible, it was just something that was so against the way I'd been brought up.' Despite this recognition, men invested in a 'protector' masculinity making it reasonable to expect their partner to fulfil her duties and, when needed they would take responsibility to help her achieve this for the 'good' of the family.

Deficiencies inflate responsibility: reforming women
Henry explained his view on his partner's social life: 'Yeah, well going out is one thing but going out and you know meeting different men…and just getting drunk and very little by way of responsibility around the house around my little girl, missing work because out being drunk so just very irresponsible.' Disrespectful constructions of women as deficient enabled some men to bolster their power and need for his intervention, help and guidance. This could be done in a number of ways: being deemed irresponsible easily led astray by various 'bad influences'; struggling with substance misuse and health problems; not coping as a mother; being poorly educated; references to infidelity; or simply being untrustworthy and manipulative. Some men infantilised women by assuming that she had no mind of her own and was therefore highly likely to be corrupted by dangerous external influences. Henry explained: 'she gets easily influenced if she is especially in the wrong company, wrong company will take her the wrong way on the wrong path and she'll just…carry on and she'll just blow with the wind…she can even become a prostitute as far as I'm concerned'. Constructing his partner as easily influenced to her detriment is seen by Henry as conferring an authority to determine who she can spend her time with: I'll put a judgement I don't like this person, I don't want them to ever come to our house…so therefore she knows…that I know they are a bad influence…other times she corrects herself and sees it for what, it takes some time but she says it afterwards 'You know what, you were right, because that person is…this way and they are trying to use me.' This tactic becomes effective as she 'corrects herself ', meaning that she now conforms to his judgements on whom she should and should not spend time with.

Pulling towards as pushing away: managing jealousy and insecurity
Jealousy over a partner's contact with other men was often normalised as a feature of heterosexual relationships. At the heart of men's accounts was needing to control their partner's interactions with other men. As Fred explained: 'I was probably jealous -I was worried in case she met someone when she went out, or started flirting…she went out with her friends in case she got influenced, flirting with a bloke or went -went off with a bloke'. This might also extend to wanting to minimise women not only having contact with men, but also being potentially desirable, often requiring them to dress differently: I wouldn't say to her like, 'You're not wearing that,' I -I'd never -I'd never say that, but I would give that disapproving look, but basically coz I -I didn't like -I'm a man and I know what men like, and I -I don't want other men looking, so I -I think to a certain extent it's understandable, but I never said to her, you know, not -not to wear anything.
Other men used more direct ways to prevent their partners from going out including refusing to do childcare, threatening to leave with the children, causing an argument or commandeering car keys. Falsely claiming to have no money or withholding money was particularly effective: 'on purpose…so if she wanted even a tennerI would say "I haven't got it"' (Guy); 'I said, "I haven't got no money."' And I -and I full well did have money, I had money in my wallet' (Fred). If she had access to money, Fred would use a different tactic, threatening to take the children to his accommodation and not bring them back for a number of days. These accounts illustrate the individualised nature of coercive control, adapted to specific relationship contexts.
The more women were limited in this way, the more some sought out autonomy, be it spatial or sexual, within the home. Roger described how his partner increasingly chose to sleep separately from him. sometimes I would go down and ask her to come to bed, plead with her, beg her to come to bed, and she would refuse, then I would feel, completely unwanted, it would end up sometimes being an extremely bad argument then, sometimes I would -I would -I'd lash out. I got so angry I'd think why -why are you doing this to me?
Unwilling to recognise his role in creating a hostile situation in which his partner sought out refuge, or his need to manage his own emotional response to this situation, Roger instead focused on how 'unfair' she was being to him. For many of the men a woman choosing to spend time with friends and relatives was experienced as rejection, leading to pressure to spend time together as a family or couple.
If I knew she was going out drinking that week then I would say 'Do you not want to make plans instead of going out drinking with ya mates and spending money, we'll…us will do something like me and you and the bairn or me and you.' (Casper) Here men intentionally used gender norms related to heterosexual relationships and the family to obscure the power inherent in restricting the freedom of their partners.

Challenging coercive tactics: letting go of gender expectations
A DVPP can provide men with opportunities to reassess their everyday lives. The men we interviewed participated in community-based DVPPs underpinned by a gendered analysis of domestic violence (Phillips et al, 2013). The group work men participated in involved a critical and collaborative approach that encouraged them to unpick gendered power relations. Some were able reflect on this in their Time 2 interviews, which are the focus of this section. Sebastian reflected on how letting go of his investment in traditional masculinity was central to his process of change 'it [was] always that, "Right, OK, I'm head of the household, I -you know, I'm going to do what I want to do, whether you like it or not"…you just stopped doing it'. Change for this sample of men was an uneven and incremental process, in some ways particular to each person and their relationship context (Kelly and Westmarland 2015). Damian described how 'it's a work in progress, it's not like a light switch that you can turn on and off…it's a thought pattern or a thought process which you need changing completely and it takes time to change it'. In this section, we explore some of the key lessons that some men spoke about in relation to the positive impact that letting go of traditional ideas about gender, family and relationships, and through that the control they had exerted to maintain these, had upon themselves, their ex/partners and children. Understanding how and why men can dismantle investments in traditional masculine norms offers insights into understanding coercive control as dynamic, contestable and possible to change.

'You don't always have to be right': finding collaborative ways of being
Letting go of gender expectations, particularly in being the traditional head of the household, meant that men could more readily admit mistakes, accept flaws and imperfections in themselves and their ex/partners and release themselves from being superior and always needing to win. This could coincide with a difficult recognition of how 'the right way' had had a negative impact on family life, as Jasper comments: 'it does annoy me, coz that's my family's gone, just for male pride or stubbornness -I just thought I was doing the right thing, you know, at the time. And I wasn't'. Some men were able to let go of the need to be right about the intricacies of everyday life, which meant that being wrong was no longer experienced as a demeaning exposure: You don't always have to be right. What I've learnt on the programme and other people have picked up on it as well is turning round and saying 'I'm wrong'. Sometimes in the past you've argued about something so insignificant you just think how embarrassing is that, so what? Just admit when you're wrong and go 'Yeah, I was actually wrong on that and you were right it's on Channel 4 not ITV, sorry.' (Brendan) This ability to be wrong and step back from solely determining what is 'right', opened up new ways of relating to ex/partners. Instead of viewing ex/partners as subordinates or opponents that required control men explored more collaborative ways of being in relationships and family that took responsibility off their shoulders. Matthew explained: 'I am focused differently on how I want a relationship to be. It's not just me, I'm the alpha male, I'll look after everything and if there are any problems then it's all down to me.' Instead when confronted with a problem some men were able to better recognise that ex/partners were on the same side and solve the issue together: 'I think now that I prefer more to talk and sort the issue out as a team rather than try and get my own over on [partner] or over on anyone' (Simon). This allowed some men to give more space for others to take responsibility and make their own decisions: 'I always felt responsible for everyone where like now it's…I'd point out stuff calmly but it's down to them what they do at the end of the day' (Brendan). This enabled women and children to expand their space for action.

'None of us are perfect': questioning the ideal
This process of questioning and dismantling gender norms also allowed men to uncover newfound respect for imperfections and flaws. This challenged perfectionistic ideals that ex/partners were previously required to live up to. For instance, in loosening his grip on his idea of the 'perfect' woman Todd found a new sense of empathy, respect and care for his ex-partner.
I had no empathy for her, I just didn't care where she was coming from. But now it's like I do care, I know she's not perfect, none of us are perfect, it's a matter of concentrating on the positive instead of taking all her negative traits into account…The thing is -if I came out with controlling arguments or actions that was because I had no respect for [ex-partner] and I just didn't care whether I was controlling or upsetting her. But now I do care, now I realise that it matters. Some men gained an understanding of how disrespect of their ex/partners legitimated abuse. Justin reflected on the infantilisation of his partner: I've justified that to myself, because this woman is behaving like a child, and so I'm -can slap her round the face, I can slap her round the legs, because she needs to be shocked and she needs to be frogmarched, she needs -for her own safety because she's lost her mind. D'you know what I mean? And of course that is awful. Some men were able to admit their own imperfections, including a reluctance to view themselves as 'cured', 'perfect' or 'fixed' and instead understand change as a long-term commitment and process.

'My heart is more open': men's emotional work
Men were invited by the programme to confront the anxiety and jealousy that underpinned their coercive control. Fred described the benefits of taking a more patient approach: If I look back on my previous relationships, if I didn't hear from my partner or something, then I'd start -I'll start panicking and then I'll start accusing them of stuff, but now I step aside and just let them have their own space, and, then I just wait till they contact me when they've got the time to do it.
This shift enabled those women still in relationships to restore their autonomy and expand their space for action.
Engagement in the group work component of DVPPs encouraged some men to explore a wider spectrum of emotions, to work towards being a more positive influence on their children. Simon described the significance of being an emotionally expressive role model for his son.
As a man now, I'm proud that if my son sees me cry it shows that he knows 'My daddy is not a robot, my daddy is actually human and he does have feelings and he is sad and he is happy and my dad has different emotions, just like me'…I used to lock my sadness away. I was only happy when I wanted to be happy, whereas now my emotions come out more and he gets to see properly what a man is meant to be.
Dexter spoke about how he is better able to consider the needs of his partner and her children, and the difference this made: 'It has opened me up as a person. I feel more comfortable around men and women and I'm more open, my heart is more open, I feel good within myself.' Dexter went on to explain: 'I think the difference is amazing. I'm more thoughtful, I'm more emotional, I'm more emotional about things people say. I'm not as hard like "Get over it ya bitch" or whatever. I'm more into their thoughts of how they feel. I'm just totally different'.

Conclusions
In this paper we have presented a gender analysis of new empirical findings that advance current understandings of coercive control based on a large dataset of phased interviews with men participating in multiple DVPPs across Britain. Although our sample, which was a self-selecting sample of predominantly white heterosexual men means that our results cannot be generalised, we have demonstrated how and why men's use of coercive tactics are effective in a wider context of gender inequality within which heterosexual relationships operate. This entrenchment of power and control in gender norms offers powerful insight into how and why reductions in coercive control are more limited than reductions in physical and sexual violence evidenced in quantitative findings.
Our findings also indicate how some men have managed to take steps away from investing in traditional masculine norms and reduce their use of coercive tactics. This advances knowledge of coercive control as more dynamic, contestable and open to change than previous research has suggested. This uneven and contradictory process can involve painful realisations of loss and harm towards the self, ex/partner and children alongside a discovery of the benefits of letting go of traditional masculine norms. In stepping back from a need to be right, responsible, rational and strong, and in accepting mistakes, emotional vulnerability and imperfections in the self and others enabled some men to find a non-abusive way of being in which they could be more open, confident, comfortable and considerate to themselves and others. This restored autonomy and expanded women and children's space for action.
Questions remain about whether men who attend a DVPP and undertake changes in their relationships are more able to unpick and deviate from restrictive gender norms; however, these findings offer potential in contributing to the ongoing development of DVPPs to better challenge men's use of coercive control. There are important implications here for organisations that provide DVPPs, most obviously the need to keep gender norms and expectations as a central role in work with violent men. There are also implications for programme content in terms of the balance of focus between reducing physical and sexual violence and the time, knowledge and other resources needed to create shifts in coercive control. For example, we are aware that some DVPPs have used the Project Mirabal findings as a springboard to discuss how they might include more material on aspects of coercive control such as financial abuse. Our findings highlight the centrality of DVPPs, underpinned by a gender analysis, in providing long-term supported opportunities for men to reflect on, unpick and experiment with letting go of gender norms, which in turn can lead men to construct non-abusive ways of being a man that increase women and children's space for action and enhance wellbeing for women and children and also for the men themselves.

Conflict of interest
The authors declare that there is no conflict of interest.