Prosecuting strangulation offences: understanding complainant withdrawal using a social entrapment lens

ABSTRACT

In many countries, one of the central challenges to the successful prosecution of domestic violence offences is the complainant’s disengagement from the prosecution process. Withdrawal is particularly likely in cases of serious domestic violence with limited physical evidence, such as strangulation. Recently countries including England and Wales and Australian jurisdictions have introduced discrete offences of strangulation, a behaviour commonly associated with domestic violence and coercive control, that is extremely risky and dangerous. In this paper, we aim to build on our understanding of complainant withdrawal in strangulation cases. Drawing on a sample of strangulation offence prosecution casefiles in Queensland, Australia we examine five casefiles using a social entrapment lens. We argue that using this lens improves understanding of the factual context in which complainants withdraw their support for prosecution and may have implications for prosecution practice.

KEYWORDS:

• Coercive control
• complainant withdrawal
• domestic violence
• non-fatal strangulation
• social entrapment

Introduction

Non-fatal strangulation is recognised as an extremely risky and dangerous form of intimate partner violence (IPV) and coercive control (Campbell et al., 2003; Glass et al., 2008; Pritchard et al., 2018). Strangulation has been identified as a significant risk factor for intimate partner homicide (Campbell et al., 2007; Glass et al., 2008) and is associated with numerous harms including pregnancy miscarriage and brain injury (De Boos, 2019; Glass et al., 2008). In response to this serious form of domestic violence, discrete strangulation offences have been introduced in many countries including throughout Australia (Pritchard et al., 2018; University of Melbourne, 2023). A strangulation offence was introduced in the state of Queensland, Australia in 2016 (Criminal Code 1899 (Qld), s315A). The offence attracts a maximum seven-year penalty of imprisonment and must be heard in the higher (District) court rather than in local or magistrates’ courts.

One of the central challenges to the successful prosecution of IPV-related offences, including strangulation, is the complainant’s withdrawal from the prosecution process (Brady et al., 2022; Messing, 2014). This is a particular challenge in relation to IPV prosecutions because there are often no third-party witnesses to the incident(s) (Ellison, 2002). While evidence suggests that complainants of serious physical IPV offences may be less likely to withdraw (Kingsnorth & Macintosh, 2006; McGuire et al., 2021), the challenges of prosecuting strangulation offences are often exacerbated compared to other serious offences (Brady et al., 2022). In many strangulation cases there are no visible external injuries from the strangulation and the consequences of strangulation, including pregnancy miscarriage and brain injury, may be significantly delayed (Stapczynski, 2010; Strack et al., 2001; Zilkens et al., 2016) and injuries may be even less visible on people with darker skin (Brady et al., 2023). Furthermore, the cause of one of the most common visible effects of strangulation, redness on the neck, has been considered by lawyers as ambivalent evidence of strangulation (Douglas & Fitzgerald, 2021). The lack of physical evidence of strangulation increases the reliance of prosecutors on the complainant-survivor’s testimony in proving strangulation (Brady et al., 2022).

Available data about the prosecution of strangulation offences suggest that complainants withdraw from prosecutions in up to 50% of cases (Brady et al., 2022; Queensland Courts, 2022) which is similar to other IPV-related offences (Robinson & Cook, 2007). While there is a comparatively large literature on complainant withdrawal in domestic violence-related prosecutions, which we briefly outline below, there are no studies that have explored complainant withdrawal through a social entrapment lens. A social entrapment lens would include an analysis of the complainant’s experience of coercive control, their realistic safety options and the social and structural obstacles that restrict the victim-survivor’s agency and autonomy and may make safety more difficult to achieve (Tolmie et al., 2018; Tolmie et al., 2024).

Tolmie et al. have observed elsewhere that ‘recanting an original statement may be a symptom of social entrapment rather than an indication that [the complainant’s] original statement was false, or the violence has stopped’ (2018, p. 183). This article explores this claim. Using a social entrapment lens, we examine five strangulation prosecution casefiles held at the Queensland Office of the Department of Public Prosecutions (ODPP). In doing so we show how using this lens may deepen prosecution authorities’ understanding of the wider context in which a complainant withdraws. We argue that this deepened understanding of complainant withdrawal may assist prosecution services in three ways. First, to develop more nuanced and individualised responses and support for complainants. Second, to improve awareness of the potential unintended consequences of pressure being placed on complainants by ODPP officers and others to stay engaged with the prosecution process. Third, to improve prosecution outcomes through shifting reliance on complainant engagement in cases where this is considered safe and appropriate.

In the next section, we briefly review the literature about why complainants withdraw their support for the prosecution of IPV-related crimes. This is followed by an explanation of the social entrapment lens. After explaining our approach, we analyse five prosecution casefiles using a social entrapment lens. We conclude with a consideration of the potential consequences of understanding complainants’ decisions to withdraw using a social entrapment lens and how this might contribute to improved prosecution understanding, approaches and outcomes.

Reasons why complainants withdraw

The literature addressing complainant withdrawal from prosecution in IPV cases highlights a plethora of reasons for withdrawal. Reasons include a lack of support services (Kingsnorth & Macintosh, 2006; Taylor-Dunn, 2016) and relatedly the time commitment associated with multiple court-related appointments requiring childcare and/or work arrangements to be adjusted (O’Neal, 2016). Court delays also contribute to withdrawal in part as they extend the time in which alleged offenders can influence, intimidate and pressure complainants to withdraw (Bonomi et al., 2011; Sleath & Smith, 2017). While complainants are less likely to withdraw where they suffer visible physical injuries compared to other injuries, where there are third party witnesses, or where forensic evidence that can support the complainant’s claims (especially video and photographic evidence) is available (McGuire et al., 2021; Reckdenwald et al., 2021), as observed earlier, strangulation complainants frequently report no visible injuries and there is usually no third party witness, and these factors also contribute to complainant withdrawal (Zilkens et al., 2016).

Especially in the context of IPV, important contributing factors to withdrawal include the complainant’s fear of the accused (Hare, 2010; Hoyle & Sanders, 2000; Robinson & Cook, 2007) or belief that the criminal justice system cannot keep them safe or solve the problem (Hoyle & Sanders, 2000). Complainants sometimes withdraw their support for prosecution because they seek rehabilitation rather than punishment of the accused (Hare, 2006; McGuire et al., 2021; Robinson & Cook, 2007) or consider arrest to be ‘enough’ (Hoyle & Sanders, 2000; Kingsnorth & Macintosh, 2006). Along this line, wanting closure, or to ‘move on’, is also identified as a common reason for withdrawing (Hare, 2006; Robinson & Cook, 2007). Continued connection with the accused, including cohabitation (Kingsnorth & Macintosh, 2006; McGuire et al., 2021), reconciliation (Robinson & Cook, 2007; Sleath & Smith, 2017), love of and or children with, the accused (Sleath & Smith, 2017) and financial dependency on the accused (Hare, 2010) are also associated with complainant withdrawal. Self-blame or personal responsibility for the assault has been linked to a higher likelihood of withdrawal (McGuire et al., 2021; Robinson & Cook, 2007). Related to self-blame, IPV complainants may also be less likely to support prosecution if the complainant was under the influence of alcohol or drugs at the time of the event or had a history of drug use (Kingsnorth & Macintosh, 2006).

A social entrapment lens

While the overview of reasons in the previous section of this article goes some way to understanding withdrawal, for any given complainant the context is likely to be complex, and reasons for withdrawal could also vary over the period of the legal intervention. As Dawson and Dinovitzer (2001) have highlighted, the prosecution process can be lengthy, and many complainants need to be supported throughout if they are to continue to engage with it. Therefore, understanding the wider context in which withdrawal occurs is important to improve the support provided to complainants and to improve their experience of the prosecution process (Brady et al., 2022; Hopkins, 2023).

As Tarrant et al. (2019) argue, the conceptual lens used to understand facts has implications for how we make meaning of them. Traditionally IPV has been understood in the criminal law response as ‘a bad relationship with incidents of violence’ (Tolmie et al., 2018). This approach sees the relationship as a dysfunctional one where one (or both) of the parties has engaged in acts of physical or non-physical abuse against the other. It presumes that effective safety options could be accessed by the complainant in lulls between the abusive incidents, that the complainant could call the police, leave the relationship or, in the context of this study, support the prosecution of a serious offence and ultimately give evidence about it (Douglas et al., 2021).

However, when IPV is understood as a form of ‘social entrapment’ (Tolmie et al., 2018) or ‘social and systemic entrapment’ (Tolmie et al., 2024)¹ different facts and considerations become visible (Douglas et al., 2020). Using a social entrapment lens involves three interwoven dimensions of analysis. The first examines the ongoing tactics of coercive control used by the predominant aggressor and their impact on the victim-survivor over time. These tactics are designed to exert power and dominance over the victim-survivor using patterns of abusive behaviour to create fear and to deny victim-survivors their freedom (Sharp-Jeffs et al., 2018; Stark, 2007). The tactics of coercive control are tailored to the specific complainant and may include physical violence. How these tactics are asked about and documented is important to ensure that the nature and extent of the abuse, the impact on the victim-survivor, the victim-survivor’s resistance and the offender’s responsibility for their violence are captured (Tolmie et al., 2018). In their study of a sample of cases prosecuted in Texas, Brady et al. (2023) highlighted the value of understanding strangulation, and relatedly complainants’ experiences of the prosecution process, through the lens of coercive control, however we argue this understanding may be enhanced when the additional two dimensions of the social entrapment lens are applied.

The second dimension of social entrapment involves consideration of the realistic safety options available to the victim-survivor. If the safety response is inadequate the victim-survivor may become more deeply entrapped and the abuse may escalate. The third dimension takes account of the impact of structural inequalities resulting from, for example, the enduring effects of colonisation, racism, sexism, poverty and disability, on the victim-survivor’s experience of the abusive person’s coercive control and of safety responses available (Tolmie et al., 2024). A social entrapment analysis facilitates a deep consideration of the location and positionality of the complainant and the accused in their families, communities and in society more broadly (Tolmie et al., 2024). As a result of Australia’s settler colonial legacy and current practices, First Nations people are disproportionately charged and convicted of criminal offences and are more likely to be imprisoned (Buxton-Namisnyk, 2022). Therefore, understanding how the structural inequalities that arise from the effects of colonisation and racism impact First Nations women’s engagements with criminal justice agencies and processes is crucial. In this context, Buxton-Namisnyk (2022) highlights both the under and over policing that First Nations women experience. On the one hand, First Nations women’s calls for help are often ignored, while on the other hand, where police do respond their response may be over-surveillance, child removal and the criminalisation of those who have called for help (Buxton-Namisnyk, 2022). For example, McGlade and Tarrant identify that Jody Gore was wrongfully convicted of the murder of her partner and show how the ‘violence, racism and state sanctioned institutional entrapment’ Jody experienced as a First Nations woman led to this miscarriage of justice (McGlade & Tarrant, 2022, p. 122).

Structural inequalities may serve to exacerbate the violence and weaken the safety options available to the complainant (Douglas et al., 2021; Tolmie et al., 2018). When safety responses such as calling the police or working with prosecution services, risk heightened surveillance, child removal and criminalisation, engaging with those systems is not a realistic safety option (Cripps, 2023; Hobson et al., 2023). This analysis underscores the importance of understanding the interwovenness of the elements of social entrapment. Notably, current scholarship asserts that a social entrapment lens can be used by legal practitioners to shift outmoded thinking about the causes and consequences of IPV, improve understanding and improve legal responses (Douglas et al., 2021).

As noted, Brady et al. (2022) go some way towards using this approach in a recent review of strangulation-survivor statements provided to police. Their study examined strangulation as part of a pattern of conduct perpetrators employed to maintain power, control and compliance instead of viewing it as an isolated incident. Framing their analysis using coercive control, they examined the complainants’ statements to police to explore the context of the strangulation occurrence, the intent of the perpetrator and how victims survived. They showed how this approach could ‘enhance evidence collection techniques and help fact finders better understand suspects’ culpability and the complainants’ “decision” to not cooperate’ (Brady et al., 2022, p. 1100).

However, as observed earlier, we suggest that Brady and colleagues’ (2022) approach may be enhanced by including an analysis of the second and third dimensions of the social entrapment lens outlined earlier. This further analysis extends to understanding the victim-survivor’s realistic safety options and the social or structural barriers they face to allow a deeper and more nuanced understanding of the complainant’s reasons for withdrawal with implications for prosecution approaches and outcomes.

Approach

In the year 2021-2022, most (97%) strangulation matters that were successfully prosecuted in Queensland, Australia resulted in imprisonment (Queensland Courts, 2022). Like other studies of strangulation offending (Pritchard et al., 2018) around 90% of complainant-survivors of strangulation in Queensland are female, and 90% of the accused are male (Fitzgerald et al., 2022) making strangulation a highly gendered crime compared to other matters such as assault where just over 70% of both perpetrators and complainants are male (Australian Institute of Health and Welfare, 2022). Furthermore, a Queensland study identified that most complainant-survivors of prosecuted strangulation offences were in a current (69%), or previous (20%), intimate relationship with the alleged perpetrator (Fitzgerald et al., 2022, p. 9). Notably Fitzgerald et al. (2022, p. 8) also found that 21% of people charged, and 26% of complainants, in strangulation prosecutions were identified as First Nations people.

We used a case study approach (Flyvbjerg, 2011) to undertake an in-depth analysis of five (hard copy/paper) prosecution casefiles involving complainant withdrawal from the prosecution of strangulation charges. The ‘unit of analysis’ (De Vaus, 2001) in this case study research is the withdrawal event and our aim was to undertake a detailed study of this event regarding complainant decision making processes and legal system responses. Notably, each case necessarily sits within the ‘real-life context’ (Yin, 2009, p. 18) of the prosecution process, and in particular the information that prosecutors deem essential to maintain on file through the legal process. From our perspective, this presents an advantage of the prosecution casefile data since it is possible to examine withdrawals from the perspective of different actors (including the complainant through witness statements and other correspondence), but also to analyse these data through the context of the prosecution process. The case study approach makes it possible to assess the withdrawal event, but also to understand whether evidence of social entrapment is apparent, and whether and if so, how it is associated with the withdrawal by the prosecution.

The five cases were drawn from a larger state-wide sample of 210 finalised ODPP casefiles, including one or more strangulation charges, finalised during the four-year period from 2017 to 2020. The full sample is explained and analysed elsewhere.² We began by first assessing all cases in the full file for the presence of the withdrawal event. For this purpose, we defined ‘withdrawal’ as information in the prosecution file that the complainant indicated a lack of willingness to proceed on the strangulation charge(s), changed their statement in relation to the strangulation charge(s), denied that the strangulation incident(s) had occurred in the way that they had first indicated to police, denied that the strangulation incident(s) had occurred at all, or disputed statements by other witnesses about the strangulation incidents. The five files for the case study were selected based on ‘practical consideration’ (Miller, 2018, p. 392) regarding the depth of information available, including from complainants, prosecutors, police and others about the reasons for withdrawal and the context in which complainants’ decisions to withdrawal occurred.

The casefiles are hardcopy (paper) and include a large quantity of qualitative or textual data, for example, witness statements, statutory declarations, correspondence connected to the case, prosecutors’ notes on the progression of the case and in many cases, and prosecution reasons for not proceeding with strangulation. To analyse these textual data, we proceeded with an inductive analysis approach, first reading and rereading all file elements related to withdrawal to develop familiarity, subsequently recording initial notes generated from within each file, and finally developing the key themes impacting on withdrawal that arose in each case. Applying a social entrapment lens to the casefile data involves consideration of the specific social location and positionality of the complainant, the accused and their families and communities. While the complainants’ structural positionality provides insights into their safety strategising concerning withdrawal, what support they can realistically count on from the family violence response system, and how they may be perceived and responded to by the prosecution, there were limitations in the casefiles regarding these issues. Furthermore, given the consequences of colonisation and racism in the Australian context it would have been ideal to include a further case study involving a complainant who was clearly identified as a First Nations woman, however, there were no files that provided sufficient information for analysis.

Complainant withdrawal through a social entrapment lens

In this section we examine the five selected casefiles to consider how a social entrapment lens might be applied to facts presented in the ODPP files. We consider in each case what difference this framing of the facts might make.

Casefile 1

In this matter, the complainant initially reported in her statement to police that her ex-partner had broken into her house, strangled her with an appliance cord and threatened to burn the house down and kill her if she called the police. Less than one month after the initial statement, the complainant recanted her statement. In a signed letter the complainant changed her story and attributed this change to her lack of clear thinking at the time of the incident:

I freaked out and just jumped to “it must be [the defendant]” when in fact I have accused the wrong person. I now know it not to be him at all … If it was him, I would not be writing this letter.

The letter also stated that she had forgotten to tell the police that the burglar was wearing a balaclava and had a tattoo on his arm so could not have been the defendant.

The complainant’s explanation for withdrawal in the letter does not capture additional facts in the case pointing to fear as a central reason for withdrawal. The complainant’s initial report to the police also stated that the accused sent her text messages ‘threatening to kill her’. Moreover, the accused was initially refused bail, but subsequently obtained bail and the complainant’s letter is dated the day after the accused was granted bail. Various other notes on the ODPP file indicate that the accused had been involved in organised crime, had a psychotic illness that he has received treatment for, and had a history of amphetamine and heroin use. The file also detailed the long criminal history of domestic violence perpetrated by the accused against the complainant, including previous breaches of protection orders and bail conditions. The complainant also had six children who were all listed on a current protection order. Despite these additional facts, ODPP staff arranged for the complainant to make a signed statement, in addition to her letter, supporting her rationale for withdrawal—i.e., her initial error in judgment.

Through a social entrapment lens, the case contains clear evidence of coercive control including the use of threats (including death threats) and violence by the accused. There are also aspects of the case that highlight the failings of the safety system, for example the offender’s past breaches of bail and protection orders, and the decision to grant bail on the most recent occasion. Indeed, the accused was on probation for firearms offences at the time these offences took place. These matters highlight a failing of the safety system. Factors that may impede her access to the safety system are also documented—e.g., the offender’s threats should the complainant contact police and his gang membership. There is also evidence that the offender possesses guns. Given the accused’s gang membership the complainant’s interaction with the criminal justice system is likely to be extremely dangerous (Tolmie et al., 2018). Finally, there are other issues identified in the file that would impede the complainant’s access to safety, not least that the complainant has six children in her care. While the strangulation charge was eventually withdrawn, the ODPP noted that they were determined to proceed on other charges. At this stage the complainant discontinued contact with the ODPP. In the absence of the complainant’s support for prosecution, the ODPP withdrew the strangulation charge but proceeded with less serious charges of burglary, common assault, threat and breach of a protection order. The accused pleaded guilty, resulting in an imprisonment sentence of 3.5 years. An unsigned file note observed ‘nil cooperation from complainant’. This comment appears judgmental and dismissive of the complainant and points to assumed standards that prosecutors expect complainants to meet and are judged by. It fails to reflect the risks the complainant faces in proceeding to support criminalisation. There was also no note that the complainant had been referred to the Victim Liaison Officer (VLO).³

Casefile 2

In casefile 2, the police summary on file states that:

The defendant has become angry over the fact there were other men at their address that day. The defendant was drunk and has become verbally abusive towards her [the complainant] … the defendant grabbed a bamboo stick and started smashing things around their room. Fearing for her life the victim left the room and fled to the neighbours’ address … The defendant has followed her inside. The victim was in the lounge room of the address when the defendant walked in and started to push her around whilst accusing the victim of cheating on him. The defendant kept pushing the victim until she fell down. He then tried to get on top of her and she tried to kick him away but was unsuccessful. The defendant then dragged the victim by her hair … then put both of the victim’s wrists on her back and pinned her wrists down with his knee and then placed his right arm around her neck and started squeezing. The victim was able to breathe but felt like she was [going] to pass out. The defendant continued to strangle the victim until the defendant’s son … rushed into the house and pushed the defendant away.

One of the items broken by the accused included the complainant’s mobile phone. In her police statement, the complainant said, ‘he was strangling me and I almost went to sleep’. A third-party witness to the strangulation also provided a statement stating that they saw the accused ‘sitting on top of the complainant strangling her’. However, the complainant later contacted the police to withdraw her statement. The file included a signed statutory declaration made by the complainant stating: ‘he was sitting beside me, holding me down to calm me down. He has done this before, and it calms me. He was not choking me’. The file notes that due to the complainant’s ‘inconsistent’ versions the charge was discontinued.

Considering this case through a social entrapment lens reveals evidence of coercive control through details of the physical violence, expressed jealousy, and destruction of the complainant’s phone—a way to isolate the complainant from her support network (Douglas et al., 2021). The file notes disclose that the accused was on probation for other domestic violence offences at the time of this incident and that he had numerous prior convictions for breaching protection orders and assault. While the complainant had tried to find safety, seeking refuge with a neighbour, this was unsuccessful. These factors highlight persistent failures of the safety system. In terms of structural inequities, it is notable that the complainant and accused lived in a remote Aboriginal community and the accused was identified as a First Nations person. Their remoteness and membership of a small community would have implications for the experience of coercive control, the availability and type of response of services within the family safety sector and the complainants’ experience of the criminal justice response (Blagg et al., 2020). Further, there was no note that the complainant had been referred to the VLO.

Casefile 3

In this casefile, the complainant and accused had been dating for five months. According to the complainant’s initial statement to the police, the accused came into her room and called her a ‘cop rat’ and a ‘two-dollar whore’ and began punching her causing bleeding, he threatened to kill her and then ‘strangled’ the complainant who then ‘could not breathe’. The accused later apologised saying he would kill himself. CCTV footage showed the accused and complainant arguing earlier in the evening at a sports club and this raised an alarm about the complainant’s welfare. In response, the police visited the complainant’s home (after the assaults) to carry out a welfare check. The complainant answered the door and reported she was safe, and that her partner (the accused) was upstairs in bed. The next day the complainant went to the police station to report the assault explaining she had lied to police who had carried out the welfare check because of her fear of being assaulted again. She also explained that she wanted to separate but ‘wanted to find a safe time and place to do so’. The police arranged for a medical assessment which identified ‘bruising over the thyroid cartilage and larynx’ and noted ‘asphyxiation injury’. The accused was charged with three offences, assault occasioning bodily harm (ABH), strangulation and threats.

An undated note written by the VLO stated that the ‘defendant has told the complainant to get rid of the police or he will kill her’. Police took out a protection order and the ODPP opposed bail. However, despite ODPP resistance to bail, the accused was bailed. Subsequently, after the accused was bailed, the complainant made significant efforts to have the strangulation charge discontinued. There were multiple letters, emails and a statutory declaration on file from the complainant. In one email the complainant states:

I feel the DVO [domestic violence order] is sufficient and the terms outlined in it will adequately protect me. To this point I have had no issues and have felt safe … I also have personal issues involving custody of my son that I need to focus on 100% and he is also starting school this year so that will be exhausting. My job is also becoming increasingly stressful, and I don’t want to be worrying about this charge.

Two months after this email the complainant provided a signed statutory declaration to the ODPP that provides five key reasons for wanting the charges to be withdrawn:

1. I do not feel unsafe or threatened in any way … 

2. The stress the matter has already had and will continue to have will cause detriment to my mental and physical health.

3. The stress of this matter is causing stress to my son aged 6.

4. … financial stress caused by time off work.

5. … reputational damage for having time off work for a matter like this.

I can sincerely not stress enough how much I do not wish to attend this matter in court and if made to attend I would not wish to speak. I would like the matter dropped.

Ultimately all charges were withdrawn. The reason identified on the prosecution file was simply: ‘the complainant wants to withdraw’.

Again, reading these facts through a social entrapment lens, it seems clear that the accused is using coercive and controlling behaviour to intimidate and manipulate the complainant. The tactics used by the accused include using violence but also suicide threats.⁴ Failings of the safety system are also identifiable. For example, a police check did not provide safety and the accused was granted bail despite the risks identified. Structural inequality that weakened the safety options available to the complainant included her status as a mother of a young child, and her keen awareness that any engagements with safety systems would have on her child. Furthermore, there is no information on the file regarding the complainant’s work context and whether her employment would be at risk if she took time off for court and further meetings with prosecution. It is well-known that women’s employment and financial independence is often jeopardised as a result of IPV (Summers, 2022). The context of employment is also important in understanding the systems and structures within which complainants make decisions.

Casefile 4

Casefile 4 provides another example of the complex circumstances that lead complainants to withdraw their support for prosecution. The complainant expressed multiple reasons at different times for wanting to withdraw from the prosecution. For example, a VLO note observed that the complainant did not want herself or her child to go to court, and that the complainant blamed her lack of recall on her cancer treatment. Subsequently, in ODPP notes recording an ODPP consultation with the complainant a year after the incident, the complainant said that the complainant had the ‘bad temper’ so ‘got hit a lot’, whereas: ‘ … if [the accused] is not on drugs, he is a good person. I believe he has learnt his lesson. He is not using drugs’.

Ultimately, in the absence of the complainant’s support, the ODPP proceeded with drug charges (supply and possess heroin) which resulted in a seven-year imprisonment penalty, the strangulation charge was withdrawn, and the accused agreed to plead guilty to ABH resulting in a five-month penalty concurrent with drug penalties. The complainant was initially not aware of the serious drugs charges that the accused faced, and it seems the complainant withdrew her support for the violence charges assuming that the accused would be free to visit her. The complainant said there was ‘no problem with him [the accused] coming to [her] home’.

A closer look at the complainant’s initial police statement in this file captures the seriousness of the incident as well as the level of fear the complainant was alleged to have experienced on behalf of her child. The accused, from whom the complainant had recently separated, broke into her house at 2.45am. The accused was in possession of a ‘large knife’ and had entered the bedroom where she slept. The complainant reported to the police that on waking, the defendant accused her of ‘loving someone else’ and that the accused:

… got on top of me and pinned me down on the bed with both his legs so I couldn’t move. He had one hand on my neck and grabbed my neck and squeezed his hand. I couldn’t breathe. I kept thinking to myself: I am dying. I am going to die. I felt like I was going in and out of consciousness. I couldn’t see anything because my vision started to go blurry, and I couldn’t take any breaths.

The accused desisted when the couple’s five-year-old son woke up (he was sleeping beside the complainant). The complainant reported grabbing their son and fleeing to a neighbour’s house. At the neighbour’s house bruises started to develop on the complainant’s neck and the complainant took a picture of the bruising with a phone. The complainant in a statement said:

I got scared because I think he turned off the CCTV because he was planning on doing something to me. I think he put the big knife on the floor to make me scared.

The neighbour ‘insisted she go to the police’.

Viewed through a social entrapment lens, this file underscores the presence of coercive control in the relationship. The complainant refers to being ‘hit a lot’ although there is little detail about these past events recorded in her statement or on the file. The fact that the accused ‘had a key cut’ and ‘broke into her home’ highlights his continuing control over her space for action after separation (Sharp-Jeffs et al., 2018). The accused’s jealousy, expressed during the assault, also highlights his wish to continue to control her. In terms of her realistic safety options, the complainant had taken the step to move away from the accused and had set up CCTV in her new home, but the complainant was ultimately not safe. Despite becoming engaged with the police as a result of the strangulation incident, file notes disclose that the ‘accused sent threatening SMS messages to the complainant’ during this time. File notes also disclose that the complainant was being treated for breast cancer and also had a young child to care for, both factors might exacerbate the experience of coercive control and impact on her safety options. The complainant was clear that her ability to engage with the legal process, as a part of the safety system, was limited, telling the ODPP: ‘I want to focus on my health’ and ‘I don’t want my son to go to court’.

Casefile 5

In this matter, the police attended the address of the accused and complainant as a result of an emergency service call made by the accused. According to police statements, when police arrived, they found the accused ‘straddled over’ the complainant with ‘his hands around her throat’, she was screaming ‘help me’. She had bruises on her face. Police described both parties as intoxicated. The complainant was taken by ambulance to the hospital where she stayed overnight. The accused was charged with strangulation and AOBH (based on the complainant’s allegation that the accused hit her with a wine bottle). The strangulation charge was discontinued as the complainant ‘indicated an unwillingness to cooperate’. An ODPP file note states: ‘ … She said that she wants to cancel everything because she and [accused] are trying to work things out … she wants to get back with the defendant’. The ODPP notes state she wished to have ‘all DV conditions removed’ from the protection order. Another ODPP note states that the ODPP are ‘content’ with resolving the matter as a plea to ABH (the wine bottle assault) as the ‘complainant’s reliability may be validly challenged’. The accused was refused bail throughout and sentenced to a three-month sentence for ABH (suspended).⁵

This file may provide an example of the way a matter could be approached when facts are viewed through a social entrapment lens. The police statements note that both parties are intoxicated, that the accused also has some injuries, and that it was the accused who called the police. The complainant notes in her statement that she was ‘angry’. On their face, these factors might suggest a dysfunctional relationship with incidents of violence and that this was a fight between two intoxicated people. However, details drawn out in the complainant’s lengthy statement point to the accused’s exercise of coercive and controlling tactics over the complainant over time:

My relationship with [accused] has been a violent one ever since we got married. [The accused] and I have had several Domestic Violence Orders, however [the accused] continued to be verbally abusive and physically violent towards me over the years.

The complainant stated that on previous occasions the accused had told her ‘ … before the police arrive you will be dead’. The file also identifies numerous previous protection orders protecting the complainant during which the accused continued to be abusive (and had been charged with breach of the orders). Moreover, at least one breach of a protection order charge on the file that resulted in a fine and notes the accused’s threats regarding the police. Seeing these facts lends support to the possibility that the complainant did not see that contacting the police or following through with supporting charges would realistically deliver safety. This view might have been reinforced by the long history of abuse described by this complainant, and the length of the relationship (47 years, both aged in their sixties) and thus the complainant’s entrenchment in the violent relationship. It is arguable that the detailed documentation of the facts led the police and ODPP to, instead of seeing this matter as a fight between equals, to identify the accused as the predominant aggressor (Tolmie et al., 2018).

Discussion: the importance of the context of complainant withdrawal

Our analysis of five prosecution casefiles demonstrates the complex and shifting engagements that complainants can have with the criminal justice system. It is apparent that the complainants in the casefile analysis were often weighing up their safety options regarding withdrawing from or continuing to support prosecution in the face of constantly shifting terrain. Often the complainants faced impossible trade-offs including prosecution versus their own (casefile 4) or a child’s well-being and employment stability (casefile 3). Such intractable dilemmas underscore the grave limitations of the IPV safety system. We argue that the social entrapment lens can deepen our understanding of complainant withdrawal and might assist prosecution services in three ways. We discuss these in turn.

First, we argue that a social entrapment analysis could help prosecutions to see facts they may miss or disregard as not relevant in the absence of using a social entrapment framework (Douglas et al., 2020). For example, if there was a failure to refer the complainant to VLO support (see for example casefile 2), did this apparent lack of action result from systemic policy apathy by the prosecutor because she was a First Nations complainant? (Buxton-Namisnyk, 2022, p. 1336.) More information is needed in most of the files we explored, but we argue using a social entrapment framework could help prosecution services to develop more nuanced and individualised responses and support for complainants. The complainant’s autonomy and rational decision-making in a dangerous and complex situation should not be discounted (Gruber, 2020). For example, the analysis in Casefile 1 suggests that the judicial decision to grant bail to the accused influenced the complainant’s choice to withdraw because she did not feel safe to continue to support the prosecution once the accused was released. The immediate target audience for the complainant’s reasons for withdrawal is the ODPP. But a second, and perhaps central, audience is the accused—for her own safety the complainant may feel she needs to communicate her efforts to have the accused’s strangulation charge withdrawn. Notably, complainants in our casefile analysis rarely explicitly identified that they were fearful, indeed complainants were more likely to tell the ODPP that they now felt safe (see for example casefile 3). This is understandable given the larger context of complainants trying to influence ODPP decision-making. The ODPP may be less likely to withdraw charges in circumstances where the complainant continues to be fearful. In Queensland the ODPP Director’s Guidelines require that complainants’ views be ‘properly considered’ but also that ‘care must also be taken that a complainant’s change of heart has not come from intimidation or fear’ (Department of Justice and Attorney General, 2016, p. 28). However, the complainant’s fear of the accused is one of the strong indicators of on-going risk (Barlow & Walklate, 2021) and prosecution decisions need to consider risks and respond in a way that does not compromise the complainant’s on-going safety. In some cases, withdrawal or reduction of charges may be a safer option for the complainant and her family (Boxall & Morgan, 2021), in others delayed or continued prosecution may be possible, appropriate and safe.

Second, we argue that a social entrapment analysis lays bare the potential unintended consequences of pressure being placed on complainants by ODPP officers and others to encourage the complainant to stay engaged with the prosecution process. Whether because of pressure from their abusive partner and/or concerns about their own safety (Hoyle & Sanders, 2000), complainants sometimes make a conscious decision to construct an alternative narrative of events, often in the accused’s favour (Bonomi et al., 2011), to effect withdrawal (Rutledge, 2009). In some of the casefiles we analysed (casefile 1 and 2) complainants attempted to withdraw from the prosecution informally but were subsequently required by ODPP lawyers to provide signed statutory declarations giving quite different versions of events compared to versions initially provided to police. While this is understandable, an analysis of the complainant’s circumstances through a social entrapment framework might lead the prosecution to provide more nuanced file notes about the complainant’s circumstances which may avoid the need for signed statutory declarations. A complainants’ disengagement from prosecution and a changed narrative of events, when seen through a social entrapment lens, may be understood as a safety strategy rather than as ‘inconsistent’ (casefile 2) or 'unreliable' (casefile 5). Feeling pressured to come up with ‘better’ reasons for wanting to withdraw and the creation of conflicting statements, especially if recorded in statutory declarations, creates further risks for the complainant. Perjury charges are possible (Rutledge, 2009), although this is unusual in Australia in IPV cases, it has occurred (e.g., R v Brown, 2015). However, there may be other implications resulting from the presence of two conflicting signed statements in prosecution files. For example, the accused (or others) may later subpoena the conflicting statements to use in other proceedings such as civil protection orders, child protection matters or family law applications to challenge the complainant’s credibility and reliability as a witness (Douglas & Fitzgerald, 2021). This may have negative impacts on future safety.

Third, we argue that our analysis underscores the need to improve prosecution outcomes by shifting reliance on complainant engagement and testimony in cases where this is considered safe and appropriate. One response to the issue of high withdrawal rates regarding strangulation offences is to pursue ‘complainantless’ prosecution as this may improve prosecution rates, remove power and control from the accused, reduce the burden for complainants, and open the door to less subjective decision-making and improved police responsiveness (Ellison, 2002). In some cases, this approach may be safe and appropriate, in others the interests of the complainant in avoiding the criminal process or ensuring that the accused is subjected to, at most, a community-based penalty, may be preferable. As outlined earlier, in some of the files we examined it was clear there were plea negotiations, usually a plea of guilty to ABH or common assault was offered by the accused resulting in the withdrawal of the strangulation charge, sometimes this occurred with the consent of the complainant. A plea of guilty to ABH or common assault generally resulted in a community-based penalty.

It is also important to recognise the consequences of mirroring the accused’s coercive behaviour with state coercion through applying pressure to the complainant to continue with a prosecution (Rutledge, 2009). Notably, even when complainants are not required to give evidence, their children or others they care about may be required to do so, there is a stigma associated with being involved in the criminal process and family funds may need to be expended on the accused’s legal representation—all of which may work against the complainant’s interests despite the risk of continued abuse (Gruber, 2020).

The casefiles show the variability of experiences and complainant views and rationales that are often responding to a changing environment. In this way they highlight the need to avoid a ‘one-size-fits-all’ approach to prosecution and we suggest that applying a social entrapment lens can help to ensure that the nuances of individual experiences are considered. The narratives apparent in the casefiles highlight the challenges faced by ODPP officers in determining how best to stay engaged with complainants and when it is appropriate to reduce or discontinue charges and what kinds of supports and responses might be needed by complainants to assist safety and/or support their continued engagement in prosecution.

Conclusion

Using a social entrapment lens when engaging with a complainant may result in VLOs and prosecution officers seeking information from the complainant and witnesses that builds a more accurate and nuanced understanding of the complainant’s circumstances, history and risks, leading to an improved/more accurate understanding of the complainant’s behaviour and decisions and a better informed and individualised response that recognises the limits of the IPV safety system (including the criminal justice system) and the effects of systemic and structural issues on complainants’ entrapment. It emphasises the need for better options for victim-survivors, beyond prosecution and incarceration of their abusers, that can address their long-term safety needs. While the available case files in this study did not permit specific analysis of the experience of First Nations women as complainants, our examination of the full data set (Fitzgerald et al., 2022) and consistent evidence from other research (Douglas & Fitzgerald, 2018) suggests that First Nations women are likely to be most disadvantaged by inadequate recognition of their social and structural entrapment and the reasons for their withdrawal will be least likely to be understood.

On a more immediate and practical level, information recorded on files using this lens (alongside the legal reasons regarding the primary evidentiary issues) might also help to ensure that those who access or subpoena prosecution files in the future gain an accurate appreciation of the complainant’s perspective, rather than seeing her as unreliable or not credible and/or as a person who refused to help herself.

Ethics approval

This research received ethical clearance from the Research Human Ethics Committee at the University of Queensland (no. 2020000558) and at the University of Melbourne.

Acknowledgements

The authors thank Leah Sharman, Eden Pearce and Madison Lloyd for research assistance.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The authors acknowledge that this research was supported by an Australian Research Discovery Project grant (DP200101020). Douglas also acknowledges support of Australian Research Council Centre of Excellence for the Elimination of Violence Against Women (CE230100004).

Notes

1 In a recent publication Tolmie et al. (2024, p. 65) used the terminology of social and systemic entrapment ‘to highlight the unique nature and quality of the social responses to victim-survivors who belong to communities dealing with sustained and intergenerational experiences of state violence and harm.’ In this paper, we use the term social entrapment to capture these unique harms.

2 This research received ethical clearance from the Research Human Ethics Committee at the University of Queensland (no. 2020000558) and at the University of Melbourne. The authors acknowledge that this research was supported by an Australian Research Discovery Project grant (DP200101020). Douglas also acknowledges support of Australian Research Council Centre of Excellence for the Elimination of Violence Against Women (CE230100004). The authors thank Leah Sharman, Eden Pearce and Madison Lloyd for research assistance.

3 VLOs are employed within the ODPP and provide a link between complainants of crime, their families and the prosecution. They can also make relevant referrals. (Queensland Government, 2021, p. 25).

4 Notably suicide threats are recognized as risk factors for future fatality (Campbell et al., 2003).

5 A suspended sentence is a sentence that will not be actually served by the offender unless the offender reoffends or breaches any conditions during the period of the suspension. It has been referred to as a ‘sword of Damocles’ (Bartels, 2007).