Abstract

Background

Coercive control is gaining increasing recognition as a form of intimate partner violence (IPV). Refugee women in countries of settlement are vulnerable to experiencing controlling behaviour. Interventions that support identification of controlling behaviour are an important step in enabling help-seeking for refugee women and prevention of further violence.

Methods

The Safety and Health After Arrival (SAHAR) study tested a culturally tailored IPV screening and response strategy for refugee women accessing Australian settlement services. All women accessing the study sites were asked about controlling behaviours using the ACTS screening tool, which also asks about actions causing fear, threats, and physical abuse. Findings reported here include consultations with a lived experience panel and services, screening results, focus group data and manager interviews.

Results

Of 312 women asked the ACTS questions by caseworkers in four settlement services, 90 women (29%) gave responses indicating IPV with controlling behaviour being the most frequently reported (78/90). Qualitative data indicate that, following consideration of language and diverse understandings of controlling behaviour, settlement service caseworkers were able to identify experiences of harmful forms of control. Conversations about control between caseworkers and women were prompted, and awareness about non-physical coercion increased.

Conclusion

Despite challenges due to differences in language, interpretation and cultural norms, this study found it feasible to enquire about controlling behaviour with refugee women accessing settlement services, along with other forms of IPV.

Background

Coercive control and controlling behaviour

Intimate partner violence (IPV), defined as behaviour by an intimate partner that causes physical, sexual and psychological harm, is estimated to impact one in three women worldwide with devastating consequences for their health, safety and human rights [1, 2]. Coercive control has received growing recognition as a form of abuse where a pattern of behaviours develops within intimate relationships that results in the micro-regulation of the lives of victim鈥搒urvivors [3].

Early conceptualisations of IPV demonstrated a clear relationship between 鈥減atriarchal domination and violence against wives鈥� [4] with violence, both physical and non-physical, directed towards women by men in order to maintain control and personal authority over their female partners [4]. Similarly, Pence and Dasgupta [5] identified power and control as central to all forms of abuse. Over time, the conceptualisation of IPV shifted to specific acts of violence, with physical and sexual violence becoming synonymous with IPV, particularly as these behaviours within intimate partner relationships became criminalised [6]. With this altered focus, awareness of control as underpinning all forms of IPV was diminished [7]. The influential work of Evan Stark [8] repositioned coercive control as a broad context within which IPV occurs and identified controlling behaviour as a form of non-physical abuse that can occur on its own.

Definition and criminalisation

Coercive control has become widely discussed in literature, including recent policy debates, among service providers and in media reporting of IPV [9,10,11]. However, definitional inconsistencies and challenges are apparent [12,13,14]. The Hamberger et al. [12] literature review identified 22 different definitions and forms of measurement; the authors nevertheless concluding that coercive control has three characteristics: (1) intentionality on the part of the abuser (2) a negative perception of the controlling behaviour by the victim, and (3) the ability of the abuser to obtain control through the deployment of a credible threat. This draws on earlier work that conceptualised coercion as involving a demand the controller is able to impose, and a credible threat for non-compliance [15].

Tolmie et al. [16] find that all forms of IPV are controlling, and call for a broader understanding of entrapment that recognises the restrictions placed on women鈥檚 autonomy by their partner鈥檚 behaviour and the systemic patterns of harm that often continue after the relationship has ended.

Coercive control is increasingly being codified in legislation, policy, and programs [17, 18]. In 2015 the offence of 鈥榗oercive control鈥� was introduced in the Serious Crime Act of England and Wales. Following commencement, prosecutions for coercive or controlling behaviour increased in England and Wales [19], with important work recently undertaken to improve the measurement of coercive control within the UK criminal code [20, 21]. In 2018 Scotland and the Republic of Ireland introduced specific offences criminalising non-violent domestic abuse [18, 19]. However, some commentators caution that criminalising coercive control may have unintended consequences, particularly for Indigenous women who face structural inequalities in the justice system and may be at increased risk of being misidentified as perpetrators [14].

Prevalence of controlling behaviour

Limited data exists on the prevalence of either coercive control or controlling behaviour due, in part, to the definitional issues referred to above [14, 22]. In the small number of studies identified in a rapid evidence review, between 7.5% and 28% of participants were identified in general population samples in Europe and North America as experiencing coercive control [10]. A European survey identified 鈥榟igh鈥� and 鈥榤oderate鈥� levels of coercive control with the combined prevalence scores ranging from 11 to 41% across the 28 countries surveyed [23]. In a US survey controlling behaviour was reported by 22% of the 1,039 female participants, comprising 鈥榯hreats of physical harm鈥� (7%) 鈥榯hreats to reveal or use private information or pictures鈥� (3%) and 鈥榩utting down or disrespecting鈥� (20%) [24]. This is similar to the prevalence of 鈥榚motional abuse鈥� found in Australia鈥檚 Personal Safety Survey, defined as 鈥榖ehaviours or actions that are aimed at preventing or controlling behaviour, causing emotional harm or fear鈥� in which 23% of 15,589 female respondents reported having experienced emotional abuse some time since the age of 15, with 6% of women experiencing partner emotional abuse at the hands of a current partner, and 18% by a previous partner [25]. The most common forms of emotional abuse found in that study, were threatening or degrading behaviours, controlling financial behaviours, and controlling social behaviours. These results are similar to another survey of Australian women who reported experiences of controlling behaviours across 13 categories of behaviour, most commonly: jealousy, monitoring of movements, financial abuse, social restriction, emotional abuse, or threatening behaviours [26].

Refugee women and controlling behaviour

Little available data exists on the prevalence and forms of control experienced by refugee and migrant women [14, 27]. In a study of the characteristics of abuse reported by 1,023 Australian women who had experienced IPV by their current or former partner, women who spoke a language other than English at home were significantly over-represented among those who experienced controlling behaviours when compared with the sample of all women in the survey (35% vs. 17%). To meet the definition, participants had to experience three or more emotionally abusive, harassing or controlling behaviours from a list of 13 possible behaviours, demonstrating a pattern [26]. The list included stalking on-line or in person, constant insults to cause shame/ humiliation, and damaging property. The findings by Boxall and Morgan for women of non-English speaking backgrounds are consistent with research that finds women from refugee and immigrant backgrounds in countries of settlement are at increased risk of IPV generally [28, 29]. In a UK study of recently developed scales for measuring coercive control, non-citizen women (refugees and asylum seekers) were found to have experienced slightly more control and surveillance than women who were citizens, although these differences were not statistically significant [30].

The literature contends that the risk of controlling behaviour is higher for refugee women due to language barriers, financial dependency, visa insecurity, limited social networks, and unfamiliarity with laws and services within the country of settlement [27, 31,32,33]. Uncertainty about residency and citizenship for women and children, particularly for women on partner visas, renders them vulnerable to threats. A file review of case notes of 100 women on temporary visas attending a multicultural IPV service found that 55% of women experienced threats from their partner of deportation, and 60% threats to withdraw visa sponsorship [34]. Financial dependency and lack of access to material resources are significant for refugee and migrant women [35]. Some writers propose that in respect of multicultural populations, cultural norms may reinforce patriarchy in the family, manifesting as controlling male behaviour, including control of primary decision-making and finances, guarding against perceived undesirable Western values, and managing family engagement with external services and agencies [36, 37].

Research aim

Within this context the Safety and Health After Arrival (SAHAR) study was undertaken as a three-year study. The overarching study aim was to improve the identification of, and response to, IPV experienced by refugee women settling in Australia. Government funded settlement support programs in Australia include the Humanitarian Settlement Program (HSP), which provides case-management support for refugees during the first 18 months in Australia, and the Settlement Engagement and Transition Support Program (SETS), which offers individual and group support from 18 months to five years after arrival [38, 39]. The study was initially undertaken with four SETS sites and subsequently at one HSP service site.

The aspect of the project reported here aims to identify controlling behaviours experienced by refugee women accessing settlement services and to evaluate the feasibility of screening for controlling behaviour in this service setting. The project trialled a culturally tailored intervention to identify refugee women鈥檚 experience of controlling behaviour, alongside other forms of IPV, with the specific research question: how feasible is it to ask about control when screening for IPV with refugee women accessing settlement services?

Method

The SAHAR study designed and piloted a culturally tailored IPV screening and response intervention for refugee women accessing five refugee settlement services in NSW, Australia, four in the greater Sydney area and one regional NSW site (Author鈥檚 own, 2024). The mixed methods evaluation included: anonymised screening and response data; a three month follow up survey of participants鈥� experience of the intervention; in-depth interviews with a small number of refugee women reporting IPV experience; focus groups with settlement staff (Supplementary file 1) and interviews with service managers (Supplementary file 2). This paper reports on the experience of designing and applying measures of control within the broader study.

The recently validated four item ACTS screening tool [40] was selected and piloted at the study sites over a four-month period. All women visiting the services, who could be seen on their own by a caseworker trained in the intervention, were read a brief preamble and then asked how often in the last 12 months a partner or former partner has made them 鈥楢fraid鈥�, 鈥楥ontrolled鈥�, 鈥楾hreatened鈥� or 鈥楽lapped/ physically hurt鈥� them with the five-point rating ranging from 鈥榥ever鈥� (0) to 鈥榲ery frequently鈥� (4). With a maximum score of 16, a score of 1 or more is indicative of abuse as validated against the Composite Abuse Scale [40]. The ACTS tool was translated into the five most commonly identified community languages most spoken at the sites (Arabic, Farsi, Urdu, Chinese and Vietnamese) and conducted by bicultural caseworkers. Anonymised screening data was collected onsite. The question on control was 鈥淗ow often does your partner/husband control your day-to-day activities? (for example, who you see/where you go? )鈥�. A copy of the ACTS screening tool used is provided (Supplementary file 3).

Bicultural caseworkers who attended a two-day training, at the five study sites conducted screening using the standardised four questions as part of their interaction with a woman over a four-month period with women visiting the service, and/or while undertaking a prescribed six-month case review, when women could be seen on their own in a safe space. According to the study protocol, women whose score on the ACTS tool met the threshold for IPV were offered referral to a dedicated DV worker trained and supported to undertake risk assessment, safety planning and to make external referrals as appropriate. Full results of the screening intervention are reported separately [41].

The data reported here comprises:

• (i) Consultations with women with lived experience as refugees, in the study鈥檚 design phase. A women鈥檚 panel comprising nine former refugee women from seven countries with direct or indirect experience of IPV was established for the study (Table听1). Two consultations were held during the design of the intervention. Additionally, caseworkers at four of the study sites were consulted during the design phase, all of whom were migrants or former refugees. Pre-intervention consultations explored the most appropriate IPV screening tool, as well as understandings and translation of key terms including 鈥榙omestic violence鈥� 鈥榝ear鈥� 鈥榮afety鈥� and 鈥榗ontrol鈥�.

• (ii) Data on refugee women鈥檚 responses to a screening question about control. Data was collated from screening forms completed at each of the five study sites to measure completion, disclosure, and types and intensity of abuse reported.

• (iii) Focus groups with caseworkers and dedicated IPV workers, and interviews with site managers. At the five study sites, focus groups and interviews explored experiences of implementing the intervention. Twenty caseworkers, four dedicated IPV workers and five managers participated (Table听1). Six caseworkers from two sites participated in both the pre-implementation consultations and post-implementation focus groups.

Table 1 Participants 鈥� pre-intervention consultations, focus groups, manager interviews

Findings

Consultation about identification of control through IPV screening

Consultations with the women鈥檚 panel and caseworkers explored the feasibility of identifying and responding to controlling behaviour, and selection of the most appropriate screening tool. At that time, it was planned to use the HITS tool, a widely implemented and validated IPV screening tool containing four items on frequency by a partner of hitting (physically hurting), insults, threats, or screaming/ swearing [42, 43]. Feedback during the initial consultations indicated that threats, insults, and screaming/ swearing were normalised for some women and often not recognised as constituting IPV, with physical acts only being understood as abuse.

Only physical abuse will be seen as DV. (WP3)

Physically hurt is the most relevant. (WP4)

Insult or talk down is not seen as a bad thing. (WP6)

Threats and insults are part of everyday language鈥� [for example] 鈥淚鈥檓 going to chop your arms off鈥�(WP4).

In response to this feedback, the ACTS tool was introduced in the consultations as a potential alternative. The HITS and ACTS tools are similar, with both being four item score-able tools with overlap in relation to hitting and threats. This prompted discussion about the feasibility of including control as a screening item. Case workers and managers suggested that, for women from patriarchal cultures, control by husbands and male relatives was common and not necessarily regarded as problematic. Furthermore, control was sometimes understood as a manifestation of male care and responsibility.

[In] Vietnamese families it is normal for men, husbands, fathers, stepfathers to be controlling. (3-I2)

鈥楥ontrolled鈥� might not be well understood. In some contexts, a man in control is seen as responsible. (3-SM)

Nevertheless, those consulted supported the use of a tool that included a question about controlling behaviour due to this being understood as a significant issue for refugee women. Some argued that to effectively identify controlling behaviour, 鈥渂enign鈥� and 鈥渉armful鈥� forms of control needed to be differentiated. Some caseworkers said that when working with clients they sometimes used language about 鈥渉ealthy and unhealthy relationships鈥� to make the distinction between caring and malign behaviour. As consultations proceeded, participants argued that some forms of control were understood by refugee women as potentially or actually harmful, and this could be illustrated with examples or prompts during screening, such as asking 鈥榙oes he stop you seeing your family鈥� or by enquiring whether the male partner insisted on knowing 鈥榳ho you see鈥� or 鈥榳here you go鈥�. When deciding on selection of the screening tool, some caseworkers said the term 鈥榠nsults鈥� in HITS was not a sufficient marker of abuse, while one manager said in relation to HITS that the tool would not capture psychological abuse. This was supported by others consulted. Participants said the questions in the ACTS tool were simple and easy to understand, such that, on balance, overall agreement was that ACTS was preferable to HITS.

Terminology and translations

Translations were undertaken by a professional translation company then reviewed by bilingual members of the research team or bilingual caseworkers from the study sites before being finalised by the translation company. Consultations yielded strong endorsement for the intervention - including the screening questions, information cards and response booklet - to be translated into community languages and delivered by bicultural workers. Client data from the study sites indicated that nearly half of service users spoke Arabic, with Dari, Farsi, Chinese, Urdu, and Vietnamese the next most spoken languages at those particular services.

We undertook consultation and discussion regarding all five languages with bilingual research assistants and caseworkers to ensure the most appropriate terminology and translation for the English word control.

As Arabic was the language spoken by nearly half of service users, and in response to issues and questions raised by Arabic speaking caseworkers, additional consultation was undertaken about the Arabic translation. Some suggested yuraqib (also mourakhaba) (賲乇丕賯亘丞) meaning 鈥渨atching you鈥� or 鈥渃onstantly watching you鈥�, but this was considered too passive and not involving actual controlling action. Another suggestion was to use the negative version of maswuwl (also masool) (賲爻丐賵賱) meaning 鈥渂eing responsible in a good way鈥�, 鈥渢aking responsibility鈥�, 鈥渓ooking after.鈥�

However, participants generally favoured use of language that conveyed meanings of surveillance, monitoring, being overbearing or restricting, involving close supervision or oversight. This led to consideration in Arabic of yatahakkam (also attahakkoum) (丕賱鬲丨賰購賾賲) or saytara (also saitarah) (爻賷胤乇賴) with a person acting this way being a musaitarah (賲爻賷胤乇赖). Saytara was described as being the more forceful term and the term also used to describe the actions of border control/security forces, so this term had come into common use among refugees and migrants. On the other hand, yatahakkam originally derived from the word for wisdom but was now understood as a negative and the term also used for rulers and dictators who made people follow their rules. As yatahakkam was also recommended by the professional translation company, this was the term selected and included in the Arabic version of the tool.

For all languages, translations were reviewed by caseworkers speaking that language before being finalised.

The consultations also recommended:

• The inclusion of 鈥榦ther physical hurt鈥� with the ACTS question about being hit, to try and capture forms of physical mistreatment often not recognised as abuse such as pushing, shoving, spitting, and choking.

• Endorsed the inclusion of 鈥榓fraid鈥� as a word commonly used and understood in Arabic and said to be clearer than 鈥榰nsafe鈥�.

• Resulted in 鈥榙omestic violence鈥� being translated in Arabic to aleunf almanzilie (also eunf manzali) (毓賳賮 賲賳夭賱賷) meaning violence inside the house or home and inferring abuse between the husband and wife, and potentially children. This was in preference to eunf eayila (also eunf aailie or al eunf al aailie) (毓賳賮 毓丕卅賱賷 丕賱毓丕卅賱賷) meaning family violence that was more likely to be used to refer to abuse in the extended family, for example with relatives-in-law, siblings, or other extended family members.

IPV screening data: disclosure of control

As reported elsewhere [41] IPV screening at the four SETS sites conducted 309 screenings out of 354 attempts (87%). Of the 45 women not screened, 11 were accompanied by their partner or another family member, 18 declined to answer, 15 were not asked because the worker understood the woman was not in a relationship and so elected not to ask the questions, and in one instance no reason was given.

Of the 309 women screened at SETS sites, 90 scored 1 or higher on the ACTS tool giving a disclosure rate of 29% [41]. Scores for this group ranged from 1 to 15 with a mean of 5.43. Control was the most frequently identified form of IPV among the 90 women, indicated by 78 of 90 women (88%). Control was often identified together with at least one other form of IPV with 58 (64%) identifying two forms of IPV, and 34 (38% of those abused and 11% of the 309 women screened) responding positively to all four questions indicating experiences of acts causing fear, control, threats and physical abuse by their partner in the past 12 months. With more than one form of abuse present, this sub-group can be seen as having experienced not only controlling behaviour, but coercive control in its broader sense.

At the HSP site, 45 women were screened with 4 women recording a score of 1 or higher on the ACTS screening tool (a disclosure rate of 9%). ACTS scores for the four women who screened positive ranged from 1 to 3. All had a score for 鈥楥ontrolled鈥�, one also had a score for 鈥楢fraid鈥�. Full reporting of the screening data is published separately [41].

Caseworker experiences of asking about control

In focus groups and manager interviews conducted four months after completing the screening pilot, participants reported that the questions in the ACTS tool were clear and simple to use and, in most instances, were readily understood by women.

The questions were really like鈥� it鈥檚 a very user-friendly for me. So that makes me easy to communicate with the client. (5-C6)

I think the questions were easy enough to ask鈥�They may have different thoughts on the questions but overall, it is very straightforward questions and there was no right and wrong鈥�I think that鈥檚 also an easier way to approach it. So, we do it to everyone, so you are not the only person to be targeted to ask these questions. So, I think this approach helped. (1-M)

Caseworkers also reported that the question about control was the least straightforward of the four screening items, due, in part, to normalisation of some forms of male controlling behaviour, sometimes reinforced by cultural norms that accepted a level of patriarchal control as benign or responsible behaviour. Implementation of the question about control was aided by having brief prompts, for example, 鈥榳here you go鈥� or 鈥榳ho you see鈥�, and it was necessary at times to provide additional explanation and examples to illustrate types of harmful control.

鈥�the second question [about control], I usually spend more time than the other questions. I explain it, 鈥淚s he controlling?鈥� She鈥檚 like, 鈥淲hat do you mean?鈥� I was like, 鈥淥办补测, let me give you an example. You want to go, for example, with your sister, go out for shopping. Does he tell you, 鈥楴辞, don鈥檛 go?鈥欌�� It鈥檚 like, 鈥淣辞, he doesn鈥檛. If he says no, don鈥檛 go, he has a reason. If he tells me, 鈥楴辞, don鈥檛 go, because I want to go with you to visit my parents鈥�, that鈥檚 fine. I鈥檒l call my sister.鈥� I was like, 鈥淥办补测. So, he made you to change what you wanted to do, for example, reschedule your appointment with your sister because you told him that you鈥檙e going out with鈥撯�� 鈥淭hat鈥檚 fine. That鈥檚 okay. He鈥檚 not controlling.鈥� So okay, 鈥淲hat about if someone calls you? Does he ask who鈥檚 calling you?鈥� 鈥淵es, he does.鈥� (3-I1).

The question about control, with examples or prompts sometimes provided by the caseworkers, enabled caseworkers to identify harmful forms of control (examples below).

• (i) Restrictions on women鈥檚 movements and who they could see:

[some men] say, 鈥淥丑, we need to keep her at home. You鈥檙e not allowed to go to TAFE. You鈥檙e not allowed to go to learn English.鈥� Or not much communication with the outside people like neighbours or classmates or whatever. (1-C2)

• (ii) Financial control:

She said, 鈥淚 don鈥檛 get any money鈥�I have a baby. I have another baby. He gets all the money.鈥� And I said, 鈥淪o what about you?鈥� She said, 鈥淚 can鈥檛 say anything鈥�. He says, 鈥淚 brought you here. These are my children. This is my money. If you are very unhappy you can go back.鈥� (2-C1).

• (iii) Threats to take the children:

[women experiencing IPV] think if they say something, they鈥檙e going to have to leave, and they鈥檙e worried about all of those things鈥�and take their children too because in Afghanistan, if somebody divorce, the children goes to Dad. (1-C2)

• (iv) Threats to be sent out of the country:

When we spoke to the clients, they said to us 鈥� the first question is, 鈥淒o you think that he can return me to my country?鈥� (3-I1).

These were contrasted with examples that both caseworkers and clients deemed to be demonstrations of care by the woman鈥檚 partner.

鈥�they feel like when they be controlled, they think it鈥檚 care for them. Or if they manage their money, they feel really happy that he鈥檚 doing everything for me. I鈥檓 happy with it. He鈥檚 managing the money. He鈥檚 managing my bills. And he鈥檚 taking me everywhere. (5-C5)

However, some caseworkers said it was not always possible for them to identify when controlling behaviour could escalate to a dangerous level for their clients.

They say, 鈥淣辞, my husband doesn鈥檛 touch me. My husband doesn鈥檛 hit me.鈥� But financial control is part of it. You just don鈥檛 know鈥�Because it might be a lot happening鈥� They don鈥檛 know what is happening until it gets to the worst situation. But how do you solve this problem? (2-C1)

Caseworkers reported that sometimes the conversations with women led to a realisation by the women that the control they were experiencing constituted IPV.

And they answer the question and then they said, 鈥淥丑, that mean I鈥檓 already living under DV a very long time, but I didn鈥檛 recognise that.鈥� And after that, they recognise they are under DV. With my culture, they just thought if husband hit you physically, that鈥檚 DV. It鈥檚 not control about financial and talk about all the things like, 鈥淗e doesn鈥檛 like me to go out. He doesn鈥檛 like me to study. He doesn鈥檛 like me to do anything that he doesn鈥檛 like.鈥� And I said, 鈥淣辞, that鈥檚 him controlling you.鈥� (3-I2).

Her impression was physical, only physical violence means domestic violence. But some questions like abusive or controlling where you go or how much money, something like that鈥�just short questions, but it sometimes can provide more information, a little bit more for the client as well. (4-C1)

Shared values between caseworkers and clients

Settlement service caseworkers were all former refugees or migrants, many having come from the same countries as their clients. They were well aware that they shared similar backgrounds, beliefs and values with the women they were supporting.

So, our cultural perspective is like it鈥檚 a patriarchal family, right? We are brought up that way. (2-C3)

The understanding of domestic and family violence within specifically the community that we were working with then, the Arabic, had very different meaning or an interpretation鈥�hence why our caseworkers who are also from that community shared the same views of what it is. (3-M)

Caseworkers reflected that implementing the SAHAR intervention had raised their own awareness about different forms of IPV with some acknowledging they previously had limited knowledge about controlling behaviour.

It helped when I was just going through the screening thing鈥�I read it, and I did some research for myself because, obviously, I wasn鈥檛 that much aware about DV. So, I explored that there are not just physical or emotional, but I got to know it鈥檚 financial, and there are so many subclasses in it. So yeah, it actually helped to also understand yourself. (4-C2)

Some of the managers observed that what caseworkers understood as constituting IPV, controlling behaviour, and safety could be different when acting in a professional capacity compared to the views they held as members of a migrant community.

I think the one-on-one work that鈥檚 done as a practitioner can be quite different about when people are looking at the concept within their own community of women鈥檚 safety. (2-M)

It鈥檚 not a matter of difference in understanding of definitions. It鈥檚 a cultural conception of what it [IPV] is and what it is not. (3-M)

Participants acknowledged that the shared perspectives about domestic violence sometimes caused apprehension about broaching such a sensitive issue and had the potential to lead to endorsement of community-normalised forms of control.

From the different cultural background, this is not easy to talk about, okay? Not at all. (2-C1)

I mean, control, like for us, even I think I鈥檓 controlled by my husband鈥� we are so used to this kind of control so maybe that鈥檚 been a hindrance鈥� (5-M).

Coercive control is not something that just happens. It鈥檚 something they鈥檝e been raised up with in the culture. They鈥檝e been taught since they鈥檙e young girls that the male, the alpha, is in control, which is true. (2-C2)

Discussion

The SAHAR study developed and piloted a culturally tailored intervention in settlement services to identify refugee women鈥檚 experience of control, alongside other forms of IPV. Following consultation with women with lived experience as former refugees and experience of IPV, and with caseworkers at study sites, the ACTS tool was selected which includes a question about control. Advice from the women鈥檚 panel and caseworkers enabled translations of the tool into terms for control that reflected community understanding.

Considering the three sources of data together, we find that it is feasible to ask about and identify experiences of control with refugee women. When applying the ACTS tool, control was the most frequently reported form of abuse with the data confirming that, while control was reported as a standalone form of abuse, it was also often accompanied by acts causing fear, threats, and physical abuse, which together meet the definition for coercive control. Bi-cultural caseworkers who implemented the screening indicated that while the concept of control as a form of abuse needed unpacking in some instances, it was possible to distinguish abusive control from what was regarded as benign or caring forms of control. Discussion of this issue often led to service users who were being asked the screening questions, and settlement staff themselves, to re-examine their understandings about IPV and dynamics of control. This reflects the emergent nature of the concept of non-physical coercion which is less well established and less recognised as a form of IPV than physical and sexual violence.

The study findings are consistent with other research that finds controlling behaviour to be the most frequently reported form of IPV for refugee and migrant women [27]. This contrasts with the ACTS validation study undertaken with a different cohort of women accessing antenatal care which found that actions causing fear was the most frequently reported form of IPV [40].

Experiences of control described by caseworkers in focus groups including financial control, restrictions of movement and social restrictions, are found in other studies of women in the wider community [24, 26]. Additionally, women in our study experienced forms of controlling behaviour specifically related to their refugee status such as visa precarity and language barriers. The vulnerability of refugee women experiencing control in the absence of physical forms of abuse is apparent, in the light of other research that finds that women experiencing control only, are less likely to seek support [26].

While refugee women are found to be at heightened risk of experiencing controlling behaviour, this is also experienced by a range of marginalised groups including culturally and linguistically diverse women, First Nations people, people with disability, people living rurally, children and youth, and lesbian, gay, bisexual, transgender, intersex or queer people (LGBTIQ+) [14]. While it may be tempting to ascribe 鈥榗ulture鈥� as a basis for higher rates of IPV, this term is often code for 鈥榥on white cultures鈥� and is profoundly racist and 鈥榦thering鈥� [44]. It also overlooks the fact that misogynistic gender norms and power imbalance exist in all cultures globally and are fundamental to any understanding of IPV as perpetrated against all groups of women [45, 46]. Refugee women鈥檚 heightened vulnerability is also exacerbated by financial dependence, language barriers, lack of understanding of laws and services in countries of settlement, community and cultural expectations, and increased vulnerability to technology-facilitated control [27, 47]. These dimensions are on top of factors known to exacerbate IPV experienced during the humanitarian crises to which most male and female refugees have been exposed such as armed conflict, perceived threats to masculinity, human rights abuses, extreme poverty, loss of livelihoods, disrupted family and community protection structures, and traumatic stress [48,49,50,51,52].

Limitations

Due to resource constraints, it was not possible to conduct extended consultation, regarding translation for all languages. While caseworkers indicated they were able to differentiate between controlling and caring behaviour it is possible that some positive scores in the study did not represent controlling behaviour, due to women providing ambiguous information about their experience, or caseworkers recording a positive score even while understanding the response as non- coercive. It is also likely that some women experiencing control chose not to speak about it when asked the screening questions, given that studies consistently find under-reporting of IPV [53, 54, 55, 56].

Implications for research, policy and practice

This study contributes to a growing body of knowledge about women鈥檚 experiences of controlling behaviour as an element of IPV, but more research is needed to better understand how control is conceptualised in different cultures and contexts, and how it can nevertheless be accurately identified and reported. Additional studies are needed on the prevalence and impacts of controlling behaviour, and its interaction with other forms of IPV, while further theoretical work is required to understand coercion as both a standalone form of abuse and a dynamic that underpins violence against women more generally.

The recent codification of coercive control in criminal law points to the need to evaluate the effectiveness of these laws, particularly with respect to marginalised groups who may be at risk of over-representation and stigmatisation. In light of reduced help-seeking for women experiencing non-physical coercion only [26] our finding of higher rates of controlling behaviour among this cohort, compared to previous testing of this tool, highlight the need for further research on how to support help-seeking for this group.

Our findings point to settlement services having a stronger role in the identification of, and response to controlling behaviour experienced by women during resettlement. This requires sufficient resourcing and support including language-appropriate information, staff training, ongoing opportunities for caseworker debriefing and discussion, and clear protocols for referral and support of women who screen positive.

Conclusion

Despite challenges in enquiring about controlling behaviour due to differences in language, interpretation and culture, this study found it feasible to effectively inquire about this form of abuse in settlement services, along with other forms of IPV. Evidence from this study points to (i) the effectiveness of the ACTS tool in identifying many instances of control that created opportunity for further risk assessment and follow up support, (ii) the value of prompting conversations between caseworkers and their clients about the nature of the control being experienced, and (iii) the importance of awareness raising about non-physical coercion for caseworkers and clients. Identifying controlling behaviour and recognising the diverse understandings of what constitutes control, are essential to enable help-seeking for refugee women and to prevent further violence.