Examining Risk and Victim Agency in Bystander Interventions in Intimate Partner Violence

Abstract

Across the Western world, support for bystander intervention in intimate partner violence (IPV) is enshrined within government policies, fostering the expectation that community members have a moral obligation to intervene to disrupt and prevent IPV. However, IPV situations are often highly complex, and research demonstrates that there are significant risks that prevent victims from seeking help. Research also shows that victims exercise great care and agency in deciding which risks they are, or are not, willing to take. Critically, there is limited understanding of if and how the risks that prevent victims from help-seeking persist—or, indeed, are exacerbated—when a bystander subverts a victim’s agency to intervene on their behalf. Drawing on in-depth qualitative interviews with IPV victims and service providers, this paper examines victims’ firsthand experiences of bystander intervention, with a particular focus on their perceptions of risk and their sense of agency in managing their own safety. We found that bystanders’ active interventions sometimes led victims to make decisions or take actions that they were not yet ready to take, and often increased the sense of risk victims were feeling. Victims then took responsibility for managing this increased risk, which added to the already heavy burden they were carrying in trying to keep themselves and their children safe. Our findings point to the critical need for more evidence to build a robust understanding of the contexts and complexities that contribute to a bystander’s intervention being experienced as helpful or harmful, and this evidence must be directly informed by victims themselves.

Introduction

Governments across the Western world are increasingly calling for, and promoting the ostensible benefits of, bystander intervention in intimate partner violence (IPV) (Kuskoff & Parsell, 2024a). Bystander intervention in IPV includes any action taken by third-party individuals to interrupt, challenge, or respond to IPV without first being asked for help by the victim (Fenton et al., 2019). Within the existing literature, bystander intervention is presented as a promising solution to IPV (Edwards et al., 2017; Pagliaro et al., 2020; J. Walker, Kelty, et al., 2024), and government policies are increasingly emphasizing the moral and ethical responsibility of bystanders to intervene in IPV (Kuskoff & Parsell, 2024a). Indeed, in the United States, it is legislated that higher education institutions must offer programs that provide ‘safe and positive options for bystander intervention that may be carried out by an individual to prevent harm or intervene when there is a risk of domestic violence . . .’ (The Campus Sexual Assault Elimination Act, 2013). Similar expectations exist in Australian policy, which commits to providing people with the ‘tools and confidence to be ethical bystanders and up-standers in appropriately responding to family and domestic violence (FDV) in their personal and professional lives’ (Australian Government, 2023, p. 7).

The responsibility for citizens to intervene in IPV is echoed by the myriad bystander intervention awareness and training programs that aim to improve community members’ understanding of IPV and increase their feelings of responsibility and self-efficacy to actively intervene when they become aware of the violence (Burnham et al., 2022; Fenton & Mott, 2018; Gainsbury et al., 2020). While such awareness and training programs were initially designed to support bystanders to intervene in sexual assault on school and college campuses, they are increasingly being modified for implementation in broader community contexts to support bystanders to intervene in IPV and other forms of FDV (Banyard et al., 2020; Fenton et al., 2019; Gainsbury et al., 2020; McMahon & Dick, 2011).

One particularly strong motivating factor underpinning the increasing interest in bystander intervention in IPV is the recognition of the many barriers victims face to seeking help. Some of these barriers relate to victims being unable to seek help, for example, due to being isolated or controlled by the perpetrator (Goodman et al., 2023; Lelaurain et al., 2017; A. Walker, Fenton, et al., 2024). Recognizing such barriers, bystander intervention does not require a victim to identify that they require help; rather, active bystanders recognize that a victim needs help and take the initiative to step in and provide that help (Park & Kim, 2023). However, deciding that the victim requires help, even in the absence of their request for help, may be at odds with victims who are unwilling or not yet ready to seek help due to legitimate fears surrounding the increased risk and consequences that they might face when such help is provided (Boethius & Åkerström, 2020; Lelaurain et al., 2017; Sylaska & Edwards, 2014; Thomas et al., 2015). Indeed, research shows that victims of IPV actively exert agency in the ways they respond to and seek help for the violence, and often carefully consider the potential ramifications prior to taking action (Boethius & Åkerström, 2020; Hume & Wilding, 2020; Meyer, 2016; Thomas et al., 2015). Given the centrality placed on agency, and in particular the way victims meaningfully and purposefully identify and manage the risks associated with IPV, it cannot be assumed that victims will experience bystander intervention in the positive way that policy often assumes.

Despite an established body of literature examining facilitators and barriers to bystander intervention in IPV, as well as strategies to overcome these barriers (Alfredsson et al., 2014; Cinquegrana et al., 2018; Pacilli et al., 2017; J. Walker, Kelty, et al., 2024; Weitzman et al. 2020; Wijaya et al., 2022), there is a critical dearth of understanding about how bystander intervention is experienced by IPV victims or what the consequences of bystander intervention are. Scholars such as Banyard (2024), McMahon (2024), and Kuskoff and Parsell (2024b) have drawn attention to the lack of evidence on the outcomes of bystander intervention in IPV and have urgently called for research into how victims experience bystander intervention and its consequences. Indeed, as seminal scholar Banyard (2024, p. 487) argues, ‘to date, our prevention programs seem to start with the assumption that bystander intervention is beneficial and more is better. We still know far too little about whether and under what conditions bystander intervention is helpful’.

This paper presents empirical evidence to advance our knowledge of the helpfulness of bystander intervention. Drawing on semi-structured qualitative interviews with IPV victims and service providers in Australia, this paper examines firsthand experiences of bystander intervention, with a particular focus on victims’ perceptions of risk and sense of agency in managing their safety once a bystander has stepped in to help. Our findings challenge taken-for-granted assumptions about the helpfulness of bystander intervention and point to the need for much more critical engagement with the risks and benefits associated with encouraging community members to actively intervene in volatile and high-risk situations, particularly when bystanders are unaware of victims’ readiness for such an intervention to occur.

Conceptualizing Risk and Agency in IPV

As existing evidence demonstrates, victims face a range of risks of harm when responding to or resisting the violence being perpetrated against them (Boethius & Åkerström, 2020; Lelaurain et al., 2017; Sylaska & Edwards, 2014; Thomas et al., 2015). Hamby and Gray-Little (2007) present a risk-based model of coping strategies that victims use to cope with and respond to IPV in ways that carefully account for and attempt to minimize such risks. The model identifies a continuum of coping strategies with three distinct points. At the lower end of the continuum is what the authors term the ‘conservative strategy’, which tends to involve engaging in avoidance or appeasing behaviors and not overtly challenging the status quo. This strategy is consciously engaged by victims to minimize future risk, including risks of escalated violence, financial insecurity, homelessness, and statutory child protection intervention. While victims who engage in conservative strategies are often seen as passive (Hume & Wilding, 2020; Meyer, 2016), Hamby and Gray-Little’s (2007) model demonstrates that such victims are, in fact, actively and strategically responding to the violence being perpetrated against them. At the higher end of the continuum is what the authors term the ‘venture strategy’. This strategy involves higher levels of risk and victims engaging in this strategy are more likely to leave the perpetrator. This, in turn, exposes them to higher risks of the perpetrator continuing or escalating the violence and making threats against children, pets, and family members. In the middle of the continuum is the ‘balanced strategy’, where victims employ a mix of conservative and venture strategies to accept loss in some areas to increase safety in others. Importantly, victims are generally aware of the risks attached to different strategies, and make carefully considered choices about which strategy to engage based on their needs and priorities.

As Hamby and Gray-Little’s (2007) model demonstrates, victims face a range of risks when it comes to coping with or responding to IPV, yet they also have agency and choice in how they respond to those risks and how they decide which risks they are and are not willing to take on. Existing literature on the help-seeking behaviors of IPV victims provides further understanding into the risk-management processes of victims who choose (or do not choose) to seek help for the violence that is being perpetrated against them. Thomas et al. (2015), for example, examine IPV victims’ perceptions of the ‘trade-offs’ or costs associated with seeking safety. The authors articulate how, when responding to and seeking help for the IPV that is being perpetrated against them, victims must consider and account for a range of volatile factors operating at the individual, interpersonal, and systemic level. Indeed, many victims participating in the study reported having to make difficult decisions about what other needs to compromise on as they worked toward achieving safety.

In more recent studies, Boethius and Åkerström (2020) and Bent-Goodley et al. (2023) similarly find that victims’ decisions to seek help are often carefully considered and highly planned beforehand. By considering who to approach for help, when to approach them, and how much of their situation they disclosed, victims who participated in both of these studies described making decisions that prioritized their safety, as well as the safety of their children, family, and friends. As Boethius and Åkerström (2020) find, some victims recognized that the perpetrator posed a risk to any friends or family members who might try to help, and therefore intentionally distanced themselves from their friends and family as a way of protecting their loved ones. Even in emergency or unplanned disclosures, victims described carefully choosing who they wanted to disclose to and seek help from.

Existing evidence identifying the risks victims face in managing or responding to IPV, including the risks that prevent them from seeking help, speaks to the considerable care and deliberation victims put into their decisions about what actions will keep them safest. As Leone et al. (2014, p. 1870) argue, the ‘help-seeking process can be extremely dangerous and involve a high degree of risk for victims . . . These statistics speak to the legitimate fears that women have about what will happen to them . . . and indicate that women take these fears into account in their decision making’. As research consistently shows, even when they are not taking direct or assertive action, victims often make very conscious and strategic decisions to respond to violence in ways that attempt to protect their own and others’ safety (Bent-Goodley et al., 2023; Boethius & Åkerström, 2020; Hamby & Gray-Little, 2007; Leone et al., 2014).

The evidence presented thus far goes against previously identified—and highly critiqued—tendencies of earlier literature to frame IPV victims as either passive victims or heroic survivors. Many scholars (Broughton et al., 2022; Hamby, 2014; Meyer, 2016) have argued that such framings tend to focus on the autonomy and self-efficacy of victims to overcome IPV, and fail to consider the structural, social, and cultural factors shaping victims’ decision-making. These critiques foreground the importance of conceptualizing victims’ agency in a nuanced way, which accounts both for victims’ capacities to act, along with the constrained conditions within which they are able to act. To this end, and following Showden (2011, p. ix), we conceptualize agency as ‘both deliberating on choices and having choices on which to deliberate. It is thus a product of both autonomy (the individual capacity to act) and freedom (the conditions that facilitate action)’. Critically, Showden’s (2011) conceptualization of agency allows us to recognize victims’ capacity to act, alongside the clear constraints to freedom that IPV represents (Wilcox, 2006). Indeed, given the high levels of control that IPV victims are often subject to, we argue that it is critical to both understand and respect victims’ rights to exert their agency when and how they can. In doing so, however, we are by no means suggesting that victims are responsible for managing or ending the IPV perpetrated against them. Rather, and drawing on the empirical literature identified above, we are simply recognizing that victims do respond to IPV, and do so in carefully considered ways that align with their own priorities and risk assessments. Recognizing how victims exert agency in responding to the IPV they experience is not synonymous with assuming they have the capacity to end the violence or even to keep themselves safe (Showden, 2011).

The above evidence demonstrating that victims actively navigate risk and exercise agency in their responses to IPV raises critical questions regarding the role of bystander intervention as an IPV intervention and prevention strategy. Bystander intervention requires third parties to take the initiative to act when they are aware of IPV, yet the action that a bystander takes may be at odds with the victim’s own risk management or violence resistance strategies. Critically, current knowledge of the experiences and outcomes of bystander intervention in IPV is highly limited, with very few studies drawing on the voices of victims themselves (Kuskoff & Parsell, 2024b; McMahon, 2024). The few studies that do draw on victims’ firsthand experiences suggest that while some victims perceive the bystander intervention they have received to be useful, a significant proportion perceive the intervention to be neutral, negative, or harmful (McMahon, 2024; Taket et al., 2014; Taylor et al. 2019; Trotter & Allen, 2009). Much more research is needed to tease out the contexts that surround these experiences, including how bystander intervention in IPV interacts with risk and victim agency, and how this interaction contributes to positive or negative outcomes for victims.

Methods

This Australian study contributes to addressing the existing evidence gap by advancing our understanding of how bystander intervention is experienced in practice. Drawing on semi-structured qualitative interviews with IPV victims and service providers working closely with IPV victims to provide skilled support, this research aimed to answer the following research questions: (1) How do IPV victims experience bystander intervention? (2) How does bystander intervention interact with the IPV risks victims already experience? (3) How does bystander intervention enable or constrain victim agency? Prior to commencing the study, approval was granted by our institution’s Human Research Ethics Committee.

Data Collection

In-depth qualitative interviews were conducted with 16 victims who have experienced IPV, as well as 7 service practitioners who work closely with and provide support to victims of IPV. Interviews with the victims aimed to elicit their experiences of bystander intervention, including who intervened in what circumstances, what the immediate and long-term outcomes were, and how they felt about the intervention. The interviews also sought to understand what an ideal bystander intervention looked like from the victims’ perspectives. Interviews with the service providers aimed to understand their experiences working with IPV victims who have had bystanders intervene, including their professional assessments of the risks involved in these interventions and the impacts that they had on the victims.

Recruitment began by contacting IPV service organizations that the research team had built strong connections during previous research collaborations. We contacted five organizations, all of which agreed to provide support for victim recruitment, practitioner recruitment, or both. These organizations represent the core specialist and non-specialist IPV service providers within one Australian capital city and surrounding areas, servicing a population of approximately 2.8 million. Recruitment support involved the organizations distributing information about the research to staff and clients, and inviting them to contact the research team to express an interest in participating. Victims were eligible to participate if they were over the age of 18 and had ever experienced IPV and had a bystander witness or intervene. Although victims of any gender were eligible to participate in the research, ultimately, all victim participants were women. Practitioners were eligible to participate if they had experience delivering services to victims of IPV. All eligible participants were included in the research.

All participants were fully informed of the nature of the research and gave their informed consent prior to participating in an interview. Interviews were conducted in-person or via telephone, as per the participants’ preference. With participants’ explicit and informed consent, interviews were audio recorded and transcribed using intelligent verbatim transcription (Eppich et al., 2019). Participants were given the option to review their transcript prior to analysis. Where participants chose not to have their interview recorded, detailed written notes were taken by the interviewer. In recognition of their contribution to the research, participating victims received an AUD50 gift voucher.

Data Analysis

Following Padgett (2017), we conducted a thematic analysis of the interview transcripts and written interview notes. Thematic analysis is a qualitative method of analysis that involves identifying common or reoccurring themes within the data, as well as the relationships between the themes. It provides researchers with a systematic approach to draw out depth and nuance in participants’ experiences (Riger & Sigurvinsdottir, 2016).

We began our thematic analysis by reading the transcripts several times to familiarize ourselves with the data. Then, using NVivo software as a tool to manage and organize the data, we developed an initial descriptive coding frame that was informed by our readings of the data as well as the overarching research questions guiding the study. This coding frame was revised throughout the coding process to reflect changing and emerging themes. To ensure consistency in the application of the coding frame, the first author coded all of the transcripts and written interview notes. After the initial coding, the codes were arranged and grouped under broader analytic themes. We present these themes in our findings section below.

Enhancing the rigor of our research is our engagement with multiple participant voices, including IPV victims and service practitioners. Engaging these different voices allows us to triangulate participants’ experiences, which refers to the process of using multiple perspectives to understand a phenomenon in more depth. As Padgett (2017) explains, engaging in triangulation is not an attempt to find a single objective truth, but rather is a recognition that a rich and deep understanding of a phenomenon is enabled through the bringing together of different experiences and perspectives. Indeed, we take seriously people’s lived experiences, and recognize that the value of bystander intervention can only be understood through people’s experiences of it. Consistent with this position, in conducting and presenting our analyses, we privilege the perspectives and realities of our participants.

It is critically important that our study protects the identities, privacy, and confidentiality of its participants. To provide this protection, when presenting interview data in our findings we have made the conscious decision not to label participant quotes with pseudonyms, descriptors, or participant numbers. Excerpts are thus simply labeled either ‘Victim’ or ‘Practitioner’. Given the relatively small number of participants, as well as the high risk to the participating victims should they be identified, we see this as a necessary step in maintaining participant confidentiality and upholding the standards of ethical research.

Findings

Understanding the Risk

Drawing on our qualitative interviews with victims and practitioners, participants discussed the risks that bystander intervention posed both to victims as well as bystanders. Risks to victims were discussed both in terms of the risk to immediate physical safety, as well as the long-term consequences that were closely linked to the complex dynamics of IPV. Risks to bystanders were discussed in terms of bystander safety, as well as well as how victims took responsibility for and actively sought to manage these risks. We discuss each form of risk in turn.

Perhaps unsurprisingly, victim safety, both in the immediate and the long term, was identified as a key theme within participants’ discussions of the risks that bystander intervention posed to victims of IPV. For many of our victim participants, bystander interventions were experienced as exacerbating the immediate situation, and further provoking the perpetrators’ anger. This, in turn, made the situation worse for the victim. As some participants explained:

He would just always put the biggest scenes on in public, and people would try to be like, ‘Hey, calm down’. And then when you’re in that moment, it makes it much worse, especially for the person who’s in the situation. I feel like it sometimes can make the violence more angry because it's like, why are you involving yourself? (Victim)

[Written interview notes] Reflecting on when her parents got involved, the participant said it often agitated her ex-husband and further enraged him, with that anger translating into further violence against her. (Victim)

The second participant quoted above went on to discuss the complex dynamics of IPV, noting that just because a violent incident might be interrupted, it does not mean the IPV ends. Indeed, in the vast majority of the examples provided by our sample, the victims inevitably ended up back in their home with the perpetrator following the intervention. Feeling provoked by the bystander’s intervention, these perpetrators often engaged in retaliatory violence against the victim. As the victims and practitioners alike explained:

One of [my friends] did tell him at one point that they weren’t impressed with his behavior . . . all it meant was I copped it [received harsh punishment] when I got home. (Victim)

You don’t know where that volatility sits. Is there going to be a rational thought? . . . You don’t know as a bystander, what’s coming next. If you step in, will they go out and buy a gun? . . . Are they going to come back? (Victim)

The risk is that we make these families more unsafe . . . the perpetrator doesn’t always want to hurt us or the bystander, but they will hurt the victim . . . We are talking people who are unsafe, and you never know what anyone is capable of. (Practitioner)

These excerpts speak to the importance of considering the long-term effects of bystander intervention in IPV, as, even if an intervention appears to resolve the violence in the short term, the risk of retaliatory, escalated, or ongoing violence remains very real for victims whose lives continue to be enmeshed with their perpetrators’.

Victims also face risks beyond the threats to their physical safety. Reflecting the complex dynamics of IPV, practitioners in particular discussed the potential unintended consequences of bystander intervention for other aspects of victims’ lives, including the risks of isolation, involvement of statutory child protection agencies, and unwanted legal interventions that have far-reaching impacts on victims’ lives. For example:

The isolation that can happen because a bystander gets involved is really, really worrying . . . I’ve worked with women before who’ve talked about [a neighbor intervening] and then being moved to an isolated community where they’ve got nothing, no connections. (Practitioner)

Of the families I worked with, there was a bystander who called the police and reported violence, but to the family it actually caused disruption . . . police intervened, and the father can’t see [the children]. It became very, very complex . . . it caused a lot of a lot of hurdles and pain . . . And the mum [denied the violence] totally. (Practitioner)

The second practitioner quoted above spoke in depth about the lengths the victim went to in denying the violence, including signing an affidavit. Ultimately, however, the father was required to vacate the home, leaving the mother to single-handedly care for their four high-needs children, whom they had recently regained custody of from statutory child safety services. This practitioner emphasized that bystanders ‘don’t know the whole story’ and, in this particular situation, the victim ‘did not have voice’.

Victims and practitioners also identified that some forms of bystander intervention posed a greater risk to victims than others. Indeed, despite many community education campaigns’ calls for bystanders to stand up to perpetrators as a way of holding them accountable for their actions (Kuskoff & Parsell, 2024a), there was near-unanimous agreement from our sample that standing up to the perpetrator was a highly risky and unwanted form of intervention. As some participants explained:

If any of my friends had stood up to him directly, I’d have panicked. I’d have really panicked, because I’d have known what was coming for me once they were out, well there and then or afterwards. (Victim)

Telling my drunk husband at a barbecue, ‘Mate you’re behaving like a bloody asshole, pull your head in’, he would have got worse when I got home . . . he was always going to do it, like no one could stop him from doing it. (Victim)

[Victims] are the ones that will go, ‘You enter that conversation, you know what’s going to happen to me tonight. Don’t. Don’t do that’ . . . Because you can risk her safety just by having a conversation going, ‘You know what you're doing is [IPV]?’. (Practitioner)

Critically, one victim even discussed how the very possibility that a bystander might take it upon themselves to intervene by confronting the perpetrator was a barrier preventing her from disclosing the violence to, or seeking help from, her friends and family. As she explained:

I would never have disclosed to anyone that the perpetrator knew. There’s not a chance in the world, in case they went, ‘Wow, this is really serious. I’m going to take it to them and talk to them about it’. Like a do-good friend that thinks, ‘Oh my gosh. She’s in over a head. Maybe I can talk to him’ . . . You can’t possibly take the risk . . . because that is straight away going to bring crucifixion to you. (Victim)

Many of our victim participants emphasized that ‘absolutely it can be dangerous to intervene if you intervene in the wrong way’. For the above victim, the risk of a bystander intervening in the wrong way by confronting the perpetrator was not a risk that she was willing to take, and this strongly shaped her decisions around seeking help.

The participants also identified how bystander intervention posed a risk to the intervening bystanders. Recognizing that perpetrators are often highly volatile, victims expressed concerns for bystanders’ immediate safety when intervening in a violent situation. For example:

I don’t think bystanders should ever step into a live situation because they’re putting themselves too much at risk . . . you just don’t know how volatile that the perpetrator is and how it’s going to pan out . . . We’ve seen where other people have stepped in and tried to help, and they’ve been injured or part of the problem. (Victim)

They’re just trying to do the right thing, and then sometimes it will just totally blow out proportion. Or sometimes it can be turned towards them, and that’s also a scary situation to have had that happen before. (Victim)

Bystanders’ longer-term safety is also a concern, even when their intervention is not taken in an actively violent situation. Indeed, some participants described how perpetrators can engage in retaliatory violence after the fact when they discover a bystander has provided help to the victim or called the police. For example:

[Perpetrator] said he was going to drive up there and beat [bystander] up . . . there was still a great risk to them, and they had to call the local police to forewarn them. (Victim)

The retaliation didn’t happen in that moment, but the perpetrator’s retaliation towards them weeks after was obvious. (Victim)

The risk to the bystander as well. Perpetrator’s not happy that you’re trying to interrupt what they’re doing . . . often there’s some kind of, I don’t know that people ever manage to carry it out, but some kind of retribution against the whistleblower. (Practitioner)

Significantly, some victims spoke about how the risk to the bystander then became their responsibility to manage, on top of the risks to themselves and their children that they were already managing. As the victims powerfully explain:

Being a bystander doesn’t mean running into a burning building . . . if you’re running into my burning building, I’ve got to worry about you as well as everything else in my burning building. (Victim)

The circle of risk grew, and because I was at the epicenter of that circle, it was my responsibility to try and predict or mitigate the risks that could occur. . . in addition to managing my son, my feelings, the separation. (Victim)

I was shit scared of retaliation. . . I was renting out two rooms of the house, so it wasn’t just me and [my son], it was these other people that I had their lives in my hands as well . . . I was holding [the risk] because I felt that very, very seriously. (Victim)

These excerpts foreground the complexities of bystander intervention in IPV, as well as the myriad potential unintended consequences. The victims quoted above were acutely aware of the risks posed both to themselves and to the bystanders, and felt a responsibility to mitigate those risks so the bystanders would not be harmed.

While bystander intervention in IPV aims to safely interrupt or prevent violence and send a message to the perpetrator that their behavior is unacceptable (Fenton et al., 2019), these findings suggest that the complex dynamics of IPV make these aims difficult to achieve. These complex dynamics—including victims’ ongoing relationships with the perpetrators, having children with the perpetrators, and often being reliant on perpetrators for financial and housing security—mean that a bystander’s intervention in IPV may have the short-term benefit of interrupting a violent incident, but poses considerable risk to both the victims’ and the bystanders’ long-term well-being. Indeed, almost all of our victim participants described experiences of bystander intervention that, despite the good intentions of the bystander, ultimately made life more challenging or dangerous for the victim.

Undermining Agency

Given the risks involved when bystanders take it upon themselves to intervene in IPV, it is important to consider how these risks interact with victims’ agency to make their own choices about how they want to respond to the violence perpetrated against them. Practitioner participants spoke particularly strongly and explicitly about the importance of respecting victim agency. For these practitioners, it was important to ensure the victims maintained control over if and when they wanted to ask for help, as well as the type of help they asked for. For example:

[Victims] feel they lose their control about when they get to ask for help . . . even if [their partners] are violent, if women are saying that it’s not violence and they’re not fearful, there’s a reason for that, and what they’re feeling is they lose a bit of control of what they want to do about it and sometimes that puts them at more risk of further violence. (Practitioner)

Even if there is violence, it is [the victim’s] decision when they’re ready to move out or do something, but this kind of bystander intervention . . . by taking the power and choice off people or trying to do that, that’s actually very unethical. (Practitioner)

It can be disempowering people . . . You’re fucking my life up here. Thank you for your intervention, I don’t want it. (Practitioner)

Recognizing the importance of victims maintaining control and choice in how they respond to IPV, as well as the risks associated with bystander intervention in IPV, practitioners stressed that to respect victim agency, bystanders should not intervene without first asking the victim what they want to happen.

I think bystander, we’ve taken away the victim being at the center . . . how do you keep someone safe if you’re going to intervene? Has anyone actually asked the woman, victim-survivor, do you want me to intervene? . . . Because they’ll tell you. (Practitioner)

A woman was saying in a group we had that, ‘Why don’t people ask me? And maybe we can start from there, and maybe I’m not ready to do anything about it, but it’s not going to make it any easier having strangers tell me what to do’. (Practitioner)

As discussed in the previous section, bystander intervention at times had unintended consequences for victims, including prompting unwanted police or statutory child protection intervention. These interventions often further removed choice from victims when police implemented unwanted violence protection orders or child protection services required the perpetrator to be removed from the home. The involvement of broader social systems that bystander action led to often had significant impacts on victims’ lives, at times separating them from the perpetrator before they were ready for this to happen. One victim explicitly discussed how a bystander’s intervention started a chain of events that resulted in her ending her relationship with the perpetrator before she felt ready. As the victim explained:

[The intervention] increased the anxiety at the time . . . my greatest fear at the time was care of my son coming into question, thinking, ‘Oh my gosh, does this mean [son] will be taken away?’ . . . The only thing I can liken it to is when you sit on a roller coaster and it starts going, and you think, ‘Oh my gosh, there’s no going back now’ . . . That was the end of my marriage. That was the end of life as I knew it. That was the end of a lot of things for a long period of time for me. (Victim)

Several other victims also spoke to the experience of a bystander’s intervention pushing them to realize that they needed to end their relationship, which was a daunting and anxiety-provoking realization that they did not yet feel ready to grapple with.

The importance of respecting victims’ understanding of their own situations and safety needs was also raised by participants. Victims and practitioners alike stressed that victims often have a deep understanding of their own situations and, compared to a bystander, victims are in a much stronger position to assess and respond to the risks they face.

Going in there, I call it with your cape on, like [a] superhero—taking that direct approach is actually the most unsafe thing that you can do, especially for someone who’s experiencing violence, because they really need to manage their own safety. (Victim)

This victim-survivor has the best gauge of how to keep herself safe. Now that might mean staying in what we would say is an unsafe situation. So I think bystanders don't always fully appreciate the dynamics there. (Practitioner)

While the majority of participants spoke to the importance of bystanders always following the victims’ lead, two victims spoke of their own experiences where a bystander went against their wishes to intervene. For these victims, while the intervention increased their feelings of risk and fear in the moment, in hindsight, they realized that these interventions needed to happen as they were too overwhelmed to be thinking clearly. As these victims explained:

I knew that the relationship had to end, but I was so battle weary that I didn’t have the wherewithal. I’d lost all my confidence, so worried about the implications around joint property ownership, our child, my job, all of those things that I knew someone needed to come in over the top of me and intervene. Now I know that’s at odds with everything I just said about agency, but at this point, I really needed somebody to take the reins on my behalf. (Victim)

It was the people that supported me that helped get me out safely. If I didn’t have that support, I’d probably still be, I’d be either dead or still in that relationship, because when you’re in it is so overwhelming, and often you can’t see the pattern of behavior. (Victim)

While our findings have thus far demonstrated the risks that victims can face when bystanders intervene in IPV, as well as the desire for victims to manage their own safety, the above excerpts suggest that there may be a time and a place for bystanders to make decisions on victims’ behalf. Notably, in both of the above cases, the victims described facing an imminent threat to their lives. As we discuss more fully in the discussion below, given the high risks involved, there is a critical need for more evidence regarding if, when, and how such interventions can and should be engaged in ways that respect the safety and long-term well-being of the victims.

Together, these findings foreground participants’ perspectives on the importance of respecting victims’ agency and control in when and how they seek help. In particular, participants expressed fears that bystanders acting without first asking the victim what they want could lead to interventions that are experienced as disempowering and cause victims to feel pushed into making decisions that they are not yet ready to make. While some victims who experienced this ultimately viewed these decisions as the right ones, the additional anxiety and loss of control they experienced once a bystander stepped in increased their stress in the short term, and compounded fears for their and their children’s safety.

Discussion and Conclusion

The findings presented in this paper speak first and foremost to the complexities involved in encouraging bystanders to intervene in IPV situations without first being asked for help by the victim. Drawing on the experiences of victims themselves, as well as service practitioners working to support IPV victims, we found that not only did victims often experience bystander intervention as disempowering; they also experienced it as something that at times disrupted their own strategies and plans, and led to new, unplanned, or increased risks. These findings represent caution for the burgeoning literature, policy, and social discourse advocating for bystander intervention in IPV.

In much of the existing literature, bystander intervention is presented as a promising solution to IPV (Edwards et al., 2017; Pagliaro et al., 2020; J. Walker, Kelty, et al., 2024). Yet much of the evidence underpinning this assertion is based on its demonstrated effectiveness in instances of sexual assault, predominantly based on college campus settings (Fenton et al., 2019; Shorey et al., 2012; Weitzman et al., 2020). Distinguishing between sexual assault and IPV is important. As Weitzman et al. (2020, p. 1696) explain, ‘Given that definitionally sexual assault and IPV are overlapping but distinct, and perceptions of these forms of violence are quite different, bystanders’ willingness to intervene and their strategies of intervention may differ’. Our findings extend this analysis to suggest that the outcome of bystander intervention may also differ, as IPV involves the victim having an ongoing relationship with the perpetrator. Our findings show that following a bystander intervention, victims inevitably ended up back at home alone with the perpetrator, where the violence continued or even increased. This suggests that while bystander intervention may interrupt an incident of IPV, it does not necessarily mean the perpetrator will end their ongoing pattern of violence. We also identified cases where bystander intervention triggered unwanted formal responses, including police and statutory child protection intervention, which had significant negative implications for the victims. These findings raise the critical question of when it is favorable for a bystander to intervene in IPV given that doing so poses a reasonable risk of future harm.

Reflecting existing literature on victim agency (Hamby, 2014; Showden, 2011; Wilcox, 2006), our analysis suggests that victims have a deep understanding of the risks they face, and are therefore best placed to make strategic decisions on how to respond. For many of the victims in our sample, even though they did not always exert agency in an attempt to end the violence, they did enact agency in how they responded to the violence, and this agency was often subverted when a bystander intervened, which at times led to greater harm. However, it is important here to heed Showden’s (2011) warning against conflating victim agency with safety outcomes—a victim’s exercise of agency does not guarantee a minimization of or end to IPV. By recognizing the value of bystanders respecting victim agency, we are not suggesting that victims have the power to end IPV; rather, we see it as important for bystanders to avoid replicating the lack of control that characterizes victims’ experiences of IPV (Wilcox, 2006), as well as the lack of control and coercion victims can experience when engaging with formal systems and responses (Osborn & Rajah, 2022; Thomas et al., 2015).

From this treatment of agency, and based on the experience of our participants, we suggest the need to further interrogate the common perception that prosocial bystanders should actively take the initiative to intervene in instances of IPV, and instead place greater emphasis on how bystander action can be guided by the needs of the victim. This necessarily involves bystander intervention being directed toward engaging with the victim to gain a deeper sense of what they want. It also requires bystanders to understand that victims are often employing their own strategies and management plans, even if they do not appear to be overtly resisting or responding to the violence. Bystanders engaging with the victim to identify ways to safely support them in these plans is a challenging but critical endeavor.

One factor complicating our championing of respecting victim agency is our finding that two victims valued having bystanders step in to make decisions on their behalf, as they felt too overwhelmed to identify what was necessary to keep safe. Notably, both of these victims described being in life-threatening situations, in hindsight assessing that there was a high risk the perpetrator would have killed them. When considered alongside our finding that intervening to call out a perpetrator’s violent behavior was seen by victims as increasing, rather than decreasing, the risk of further violence, this suggests that there are specific contexts in which bystander intervention in IPV is and is not experienced as helpful. Based on our preliminary findings, it is possible that the outcomes of bystander intervention vary according to what the intervention is trying to achieve. Bystander interventions seeking to change the attitudes and behaviors of the perpetrator were not only seen as ineffective but also dangerous, while interventions seeking to protect victims’ immediate safety in a potentially lethal situation were both effective and valued. In this sense, subverting victim agency and potentially contributing to future risks may be justified if it can reasonably be based on an assessment that not doing so will result in significant immediate harm. Linking back to our point above, a deep engagement with the victim and knowledge of their situation will in many cases, be necessary to enable the bystander to assess when such significant harm is imminent.

These findings, along with our findings regarding the risks bystanders themselves face when intervening in IPV, also speak to ethical questions regarding what level of risk bystanders can reasonably be expected to expose themselves to when seeking to protect others (Banyard, 2024). While these questions are beyond the scope of this paper to address, they are nonetheless critical for policymakers, and society more broadly, to consider when developing understandings of and expectations for bystander intervention in IPV. A deep understanding of the risks involved in bystander intervention is a critical precursor to underpin these discussions.

The findings presented in this paper must be considered in relation to the context and nature of the sample, including its diversity. The research was conducted in a single Australian capital city and the sample was accessed through formal domestic and family violence services. While most victim participants were not accessing services at the time they experienced the bystander intervention, their experiences of IPV and bystander intervention may be different from victims who have never accessed formal services. Moreover, some victims had been separated from the perpetrator for many years and were no longer experiencing violence, while many others had only very recently separated and were experiencing ongoing post-separation violence. It is possible that, with the benefit of hindsight, victims who were no longer experiencing IPV were able to more clearly assess the effectiveness of the bystander intervention. While we did not explicitly ask victims how they identified in terms of race and culture, several participants mentioned identifying as First Nations or as culturally and linguistically diverse. However, due to the small nature of the sample, we were unable to examine differences relating to race or culture. Further, while our study was open to victims of any gender or sexuality, our sample of victims ultimately only included women in heterosexual relationships. It is very possible that victims of other genders and sexualities experience bystander intervention differently.

Finally, we note that our small and purposive Australian-based sample, as well as our qualitative thematic analysis approach, limits the generalizability of our findings. Nonetheless, within the context of an extremely limited body of literature examining the consequences and effectiveness of bystander intervention in IPV, our findings contribute critical evidence that cautions us to carefully interrogate common assumptions regarding the desirability of bystander intervention in IPV. Much more evidence is required in this space to build a robust understanding of the contexts and complexities that contribute to an intervention being experienced as helpful, and this evidence must be directly informed by victims themselves.