‘It’s a Gendered Issue, 100 Per Cent’: How Tough Bail Laws Entrench Gender and Racial Inequality and Social Disadvantage

Women’s rates of remand, or pre-trial detention, have grown dramatically in Australia and the rates at which Aboriginal and Torres Strait Islander women are incarcerated without conviction are particularly high. However, there is little research examining bail and remand practices and their relationship to social inequalities. This article presents findings from research on the drivers behind women’s increasing rates of custodial remand in Victoria—a jurisdiction that has significantly restricted access to bail through legislative reforms. Drawing on data derived from interviews with criminal defence and duty lawyers, we outline how bail and remand practices systematically disadvantage women experiencing housing insecurity and domestic and family violence (DFV), increasing their risk of becoming trapped in longer-term cycles of incarceration. Our analysis reinforces the need to move away from ‘tough on crime’ approaches to bail. It also highlights unintended consequences of DFV reforms, including further marginalising and punishing criminalised women who are victim-


Introduction
Australia has experienced unprecedented growth in its remand (pre-trial detainee) populations. Since 2000, the number of people held on remand has more than trebled, a rate of growth that far outpaces those of similar jurisdictions internationally, such as England and Wales and Canada (Walmsley 2020). The explosive growth in remand populations in Australia reflects legislators' increasingly 'tough' approach to bail (Bartels et al. 2018;Auld and Quilter 2020). In the populous south-eastern state of Victoria, two waves of reform to bail laws (in 2013 and 2018) have driven remand rates to crisis levels, particularly for women (McMahon 2019). In June 2021, 54 per cent of those in Victoria's women's prisons were unsentenced, compared to 22% in 2011(Corrections Victoria 2021. For the first time, women on remand outnumber women serving a custodial sentence in the Victorian prison system, and many women leave prison without having spent any time under sentence (Corrections Victoria 2019). First Nations women have been disproportionately affected by changes to bail: between 2009-2010 and 2019-2020, the number of unsentenced Aboriginal and Torres Strait Islander women entering prison rose by 440 per cent (compared to 210 per cent for all unsentenced prisoner receptions; Corrections Victoria 2020). Remandees now represent a significant portion of prison populations, though relatively little is known about how and why people are remanded and the personal and social consequences of the increasing use of pre-trial detention.
This article presents findings from a research project conducted in Victoria, Australia, that investigated drivers of the growth in women's remand numbers. We found that recent changes to the Bail Act 1977 (Vic) that make it harder to obtain bail are having particularly egregious effects on women (Russell et al. 2020b). Our research identified a 'constellation of circumstances' that contribute to women's likelihood of being remanded in custody (Russell et al. 2020b). This constellation ultimately indexes gender and racial inequalities, rather than so-called 'risks' to community safety, and includes experiences of poverty and homelessness, domestic and family violence (DFV), untreated health problems, addiction and childhood trauma. Frequently, these circumstances intersect and exacerbate each other. For example, we found that homelessness is the most significant barrier for women to overcome in an application for bail and that women's lack of safe and secure housing is often the result of DFV (Russell et al. 2020b). This indicates that the crisis of women's remand levels is a product of systemic inequalities in the operation of bail laws rather than individual issues or 'crime' trends.
In this article, we focus on these specific interactions between housing precarity, DFV, policing and courts that occur within Victoria's notoriously strict bail system and the ways in which these create pipelines to prison for women. We draw on data derived from interviews with criminal defence and duty lawyers to explore how police and court responses to criminalised women can exacerbate their systematic exclusion and increase their risk of remand and entrapment in longer-term cycles of imprisonment. We identify six specific ways that this occurs: (1) the denial of bail to women without access to housing; (2) intervention orders precluding women from housing; (3) DFV-related isolation and control disadvantaging women's bail applications; (4) police pursuing other matters when called to respond to DFV incidents; (5) police 'misidentification' of the predominant aggressor in DFV; and (6) a perception of women as less 'innocent' or 'deserving' of protection if they are already criminalised. Each of these examples illustrates how housing insecurity and DFV victimisation can increase women's likelihood of being remanded. Taken together, our findings highlight systemic problems and biases in the bail and remand process and in legal responses to DFV more broadly. Strict bail laws are disproportionately affecting women, especially Aboriginal and Torres Strait Islander women, and entrenching their social disadvantage and exclusion.

Women, Criminalisation and Remand
Women on remand represent a growing subset of criminalised and imprisoned women. Remanded women have been criminalised and imprisoned, though they are not serving a custodial sentence (and sometimes never will). Research on criminalised and imprisoned women has long highlighted that prison fails to deal with-and often exacerbates-patterns of trauma, unmet health needs, structural racism and socioeconomic disadvantage (Carlton and Segrave 2014;Corston 2007;Kendall et al. 2019;Klippmark and Crawley 2018;Moore, Scraton and Wahidin 2017;Pollack 2012;Richie 2012). Women in prison have high IJCJ&SD 3 www.crimejusticejournal.com rates of past victimisation, frequently perpetrated by their intimate partners (Jones, Bucerius and Haggerty 2019;Stathopoulos et al. 2012), and the disciplinary regimes and surveillance carried out in prison can replicate and extend women's prior experiences of abuse, creating a 'carceral continuum' of gender violence (Carlton and Russell 2018;Segrave and Carlton 2010).
Periods of time spent on remand are typically short and unpredictable, making them highly disruptive and often distressing, with limited opportunities for accessing programs or services (ALRC 2017;Forrester et al. 2020;Lachsz and Hurley 2021), including transition or post-release support (Baldry 2010;Carlton and Segrave 2014). Women that are remanded risk being evicted from their housing and losing their employment, and their children may be taken into child protection (Inspector of Custodial Services 2020; Sullivan et al. 2019). At the most extreme end of the harms involved in remand is the risk of dying in custody. With the global surge in Black Lives Matter protests during 2020 and 2021 amplifying First Nations communities' longstanding calls to end Black deaths in custody, bail and remand practices have come under increased scrutiny for their disproportionate effects on Aboriginal and Torres Strait Islander peoples. In 2020 alone, two Aboriginal women died while remanded: Veronica Nelson Walker in Melbourne and Aunty Sherry Fisher-Tilberoo in Brisbane. Veronica Nelson Walker died on 2 January 2020 at Victoria's maximum-security women's prison, the Dame Phyllis Frost Centre, after being refused bail for shoplifting charges. These preventable and premature deaths highlight the racist silences and abject failures of the government to implement the recommendations of the 1991 Royal Commission into Aboriginal Deaths in Custody (Whittaker 2020) and the dangers of unchecked carceral expansion through remand (Russell, Carlton and Tyson 2020a).
As established above, the imprisonment of women reproduces profound gender and racial injustices. Research has shown that women are pathologised and punished in carceral systems in gendered ways (Carlton and Russell 2018;Franich, Sandy and Stone 2020) and that gender inequalities intersect and overlap with race, class, disability and sexuality-based oppression to entrap the most marginalised women and girls (Oparah 2012;Pollack 2012;Richie 2012;Russell and Carlton 2013). However, research detailing the specific ways that bail and remand processes are undermining attempts to reduce women's incarceration rates is only starting to emerge (Kerr and Shackel 2018;McNaull 2017). Our research speaks to this gap, providing a qualitative empirical account of the specific policing and court practices that drive women's criminalisation and incarceration at the 'front end' of the system via bail and remand processes. The analysis developed in this article both builds on and departs from previous Australian research investigating women's imprisonment trends that emphasises the 'triumvirate' of victimisation, substance abuse and mental illness among imprisoned women (Bartels, Easteal and Westgate 2020;Stathopoulos et al. 2012) by directing closer attention towards the systemic inequalities driving gendered carceral growth, rather than individual or psychological issues.
Our findings make a further qualitative contribution to a growing body of Australian research that suggests that legal responses to DFV are having unintended and adverse consequences for multiply marginalised women (Bevis et al. 2020;Douglas and Fitzgerald 2018;Nancarrow 2019). Contrary to their aims of protecting women-who represent the majority of victim-survivors of DFV (AIHW 2018)-police and court processes surrounding DFV are in some instances increasing women's risk of criminalisation and incarceration (Walklate and Fitzgibbon 2021), especially if they have previously experienced criminalisation or are otherwise affected by racism, poverty or poor mental health. This can occur through 'dual arrests' or police 'misidentification' of the 'primary aggressor' (Nancarrow 2019;Reeves 2020Reeves , 2021Ulbrick and Jago 2018), police arresting women on warrants or for unpaid fines (Klippmark and Crawley 2018) or police charging women when they use violence in the context of self-defence (State of Victoria 2016; Wilson et al. 2017). Through analysis of criminal defence and duty lawyers' perspectives on the drivers for women's remand, we identify how narrow preconceptions about 'deserving victims' can further enmesh criminalised women in carceral systems.
In sum, women's prison systems are typically filled with women that have arguably experienced some of the worst excesses of structural, institutional and interpersonal violence. Existing legal responses to DFV, by and large, do not serve to protect these women and can, instead, further punish them. While 'tough on IJCJ&SD 4 www.crimejusticejournal.com crime' reforms to bail laws in Australia have not ostensibly been targeted at women, women have, nonetheless, been profoundly affected by them, as indicated by their steadily climbing remand rates.

Bail in a 'Tough on Crime' Climate
Over the past decade, Australian researchers have observed a pattern of 'an increasingly politicised environment around bail reform', especially the use of the bail regime to send a 'tough on crime' message (Bartels et al. 2018: 91;Brown and Quilter 2014). This has been epitomised in the state of Victoria, which overhauled its bail system in the aftermath of a random attack carried out in January 2017 by the recently bailed Dimitrious Gargasoulas, who drove a car into pedestrians along a busy street in Melbourne, killing six people. As the public reeled from the attack, the Victorian Government announced changes to the bail system and a major review of bail laws. Community safety was given prominence in the reform, alongside the presumption of innocence and fairness and transparency in bail decision-making (see s1B, Bail Act 1977 (Vic)). Notably, bail practices are not intended to involve punishment. In practice, however, it appears that the reformed bail laws are particularly tough on marginalised women, prompting advocates to argue that women are effectively being punished for men's violence (Poole 2019).
Researchers and advocacy organisations are now calling for the repeal of the overly complex and restrictive 'reverse onus' provisions added to Victoria's Bail Act in 2018 (FCLC andLIV 2020;McMahon 2019). These provisions have significantly widened the net of the highest legal thresholds for bail, requiring accused persons to establish either 'exceptional circumstances' or 'compelling reasons' (depending on their charges) to be granted bail. Prior to 2018, the exceptional circumstances test for bail in Victoria applied only to people accused of a small number of very serious offences, including murder, treason, drug trafficking and terrorism offences. Following the 2018 reforms, those accused of committing a Schedule 2 offence while on bail for a Schedule 1 or 2 offence must show exceptional circumstances to avoid remand. In practical terms, this means that people accused of low-level offending, such as shop theft or drug possession, that would not usually attract a custodial sentence, can find themselves having to pass the highest legal test to be granted bail. Notwithstanding the extraordinary effects of the COVID-19 pandemic on the Victorian court and prison systems, this complex legal manoeuvre has fuelled sustained growth in remand numbers and exacerbated the problems of growing numbers of women leaving prison without serving any time under sentence (Corrections Victoria 2019; Walker, Sutherland and Millsteed 2019) or receiving 'time served' prison sentences (Sentencing Advisory Council 2020). Moreover, the broad use of these 'reverse onus' provisions erodes the long-established presumption of innocence in the criminal law (Myers 2016).
In 2020, the widespread effects of COVID-19 were felt in the bail and remand system (Lachsz and Hurley 2021), and magistrates' wariness to send unsentenced people to prison during the pandemic resulted in a 35% decline in the number of unsentenced women in prison in 2020 compared to 2019 (Corrections Victoria 2021). While remand numbers are again surging in 2021, the downturn of the previous year nevertheless indicates that high rates of remand are unsafe, unsustainable and unnecessary and that programs of 'decarceration' offer viable and much-needed alternatives (FCLC and LIV 2020). Deepening understanding of the bail laws' effects 'on the ground' (Russell, Carlton and Tyson 2020a) can support advocacy efforts to reduce political reliance on incarceration as a solution to social problems.

Methods
Our research into women's experiences of the bail and remand system in Victoria involved court observations, interviews with lawyers and analysis of statistical data. Detailed methodological notes and findings from our court observations are available elsewhere (Russell, Carlton and Tyson 2020a;Russell et al. 2020b). For this article, we explore some of the key findings to emerge from the 13 semi-structured interviews we conducted with criminal defence and duty lawyers during June-August 2019. All interviewees participated voluntarily and had at least one year of experience representing women applying for bail in Victoria. We posed seven broad questions to elicit lawyers' perspectives about women navigating the bail and remand system. One question explicitly asked: 'What role, if any, do family violence IJCJ&SD 5 www.crimejusticejournal.com experiences play in women's criminalisation and remand into custody?' The length of interviews ranged from 23 to 57 minutes, with an average length of 34 minutes. They were audio-recorded and transcribed for coding and analysis.
To analyse the interview data, we adopted a 'grounded theory' approach (Birks and Mills 2011). We conducted close readings of transcribed responses to each question and then progressively developed a coding scheme for each question as themes arose in the transcripts. The coding scheme thereby largely emerged from the data itself. We used NVivo software to conduct line-by-line coding of each interview transcript. Each transcript was coded at least twice and by at least two members of the research team. Key themes in the interview data were identified by referring to the frequency with which they appeared across the 13 interview transcripts, as calculated in NVivo.
Drawing on the findings of our analysis of interview data, this article explores lawyers' perceptions of homelessness and DFV as important precursors to and drivers of women's criminalisation and remand. Our aim is to build understandings of how gendered inequalities at social, legal and economic levels contribute to the criminalisation of women. Through a series of examples and accounts provided by interviewees, we demonstrate how these interrelated issues shape women's pathways and pipelines to prison. Based on the results of our study, we argue that there is a nexus between DFV, homelessness and women's criminalisation in the bail and remand system that indicates social and systemic failings, rather than the individual deficits and 'risks' that are frequently attributed to imprisoned women (Pollack 2010;Russell and Carlton 2013).

A Constellation of Circumstances
In interviews, lawyers described a range of hardships that women have generally experienced prior to being remanded, which reflect the findings of previous research on women's criminalisation (e.g., Balfour and Comack 2014). These include poverty, DFV, homelessness, untreated physical and mental health problems, and drug and alcohol addiction. However, the ways in which lawyers conceptualised the relationships between these hardships and the effects that they have on women's lives provide new frameworks for understanding women's criminalisation at the front end of the prison system. For instance, Lawyer 8 described the confluence of disadvantages listed above as a 'constellation of circumstances that operate against [women in a bail application]' (Russell et al. 2020b). Lawyer 6 noted that together these issues create 'a level of chaos in [women's] lives'. The notion of 'chaos' also resonated for Lawyer 1, who suggested that 'often [women's] offences are much lower level, but the chaos that's attached to their lives has meant that they have just really struggled to deal with all of those things'. Rather than an inherent dangerousness or a series of 'bad choices', Lawyer 2 explained that women's offending is: [I]n the great majority of cases, driven by poverty and trauma and disadvantage and things like mental health issues or drug addiction. Those things are often coming from the experience of trauma. … It can either be an underlying issue that has maybe contributed to poverty or unemployment or mental health issues, or it can be something that has been immediate.
The defence and duty lawyers we interviewed reiterated a point that is well established in the literature (e.g., Bartels, Easteal and Westgate 2020;Segrave and Carlton 2010;Wilson et al. 2017): trauma is bound up in many aspects of women's criminalisation. Frequently, women are already traumatised by the time they are arrested and charged, and if they are denied bail, they are then placed at risk of further traumatisation while on remand in the basement cells of the magistrates' court or at the maximumsecurity women's prison (Russell, Carlton and Tyson 2020a). Lawyer 4 stated: A classic example would be a woman living in her car, fleeing domestic violence, selfmedicating with drugs and alcohol. Presents as very unwell, quite traumatised, goes into the prison. The prison decides that she's a bit of a risk to herself so then they put her in an isolation cell, a probation cell, so she's dealing with all of that on her own, which is not great. IJCJ&SD 6 www.crimejusticejournal.com Imprisonment can serve to extend and exacerbate women's experiences of gendered violence and reinforce a sense of powerlessness (Pollack 2012). This process of disempowerment is evident in bail hearings (Russell, Carlton and Tyson 2020a). First Nations women, who are uniquely and acutely marginalised within a colonial and patriarchal legal system, are especially affected (Baldry and Cunneen 2014;George et al. 2020;Klippmark and Crawley 2018). Lawyer 3 elaborated: An Aboriginal woman with a history of being oppressed by the Western world is in a situation where they're having to ask a white decision-maker, often a white male decision-maker, for their liberty, then being told, 'Well, if you don't work with these other white organisations, we're not going to give you bail because you're too at-risk'. You can see how women can get pretty disenfranchised, particularly the Aboriginal women, being told by the people that have oppressed them that … they are being exposed to too much risk or they represent too much of a risk.
This lawyer reflects on the colonial dynamics of the courtroom, which construct Aboriginal women as unruly or 'risky' and obfuscate the structural power imbalances in play (Razack 1998). In criminal justice settings, scholars have shown how discourses of risk become pseudo-scientific tools of racialised and gendered governance, reframing markers of social disadvantage as 'risks' that might precipitate reoffending or non-compliance (Hannah-Moffatt 2005; Struthers Montford and Hannah-Moffatt 2021). This was reflected in our study, which found that gendered inequalities and social disadvantages gave rise to carceral responses from state and judicial officers working under strict bail laws.

Policing DFV and the Penalisation of Homelessness Under Strict Bail Laws
In a context of increasingly stringent bail legislation in Victoria, lawyers identified various ways in which women experiencing housing insecurity and DFV are criminalised or re-criminalised and funnelled into the bail and remand system. Six of these are outlined below: lack of housing; intervention orders; DFVgenerated isolation; policing practices; 'misidentification' of the primary aggressor; and criminalised women's failure to embody ideals of victimhood.

Lack of Housing as a Barrier to Obtaining Bail
Overwhelmingly, lawyers identified homelessness as the most significant barrier that women encounter when applying for bail. While the Bail Act does not stipulate that housing is a specific requirement for bail, having stable housing is 'a pretty big reason' for a magistrate to grant bail (Lawyer 4). Lawyer 6 explained that 'it's definitely harder to get bail without an address, but it's not impossible'. Victoria's 'housing crisis' (Lawyer 8) and 'shortage of emergency housing' (Lawyer 6) can delay women's bail applications or thwart their chance of success altogether. Lawyer 13 gave an example of a magistrate's probable line of reasoning when making a decision about bail: 'Where did she live? Homeless? Couch surfing? Well … how is this person possibly going to comply with their bail conditions if they're just couch surfing?!' The use of remand in response to the perceived insecurity and instability borne of homelessness highlights how stringent bail laws are particularly 'tough' on disadvantaged and marginalised women. In place of remand, the so-called 'risk' of non-compliance with bail conditions could, instead, be mitigated by increasing housing options for women. This would be likely to have widespread benefits since Corrections Victoria (2019) reports that 26 per cent of women entering prison on remand experienced homelessness or housing instability before prison. Our observations of the bail and remand court indicate that the proportion is likely higher: 13 of the 36 women's hearings (more than a third) we observed involved consideration of her homelessness or housing instability (Russell et al. 2020b). As one magistrate exclaimed when denying a woman's application for bail: 'I don't even know where you would live!' Homelessness and housing instability present problems for both men and women seeking to obtain bail. However, insecurely housed women face additional gendered barriers when applying for bail. Lawyer 3 observed that 'magistrates are perhaps more hesitant to bail a woman to the streets because they recognise either subconsciously or consciously the vulnerability of a woman on the street versus the IJCJ&SD 7 www.crimejusticejournal.com vulnerability of a man on the street'. Yet women also experience danger in their homes-DFV is the leading reason that women and children leave their homes in Australia (AIHW 2019). Lawyers reported representing 'a lot of women who have found themselves homeless due to family violence, or they are transient', which then becomes 'a barrier to getting bail' (Lawyer 6). Lawyer 4 described homelessness as a 'nuanced factor when dealing with women … particularly where women are fleeing domestic violence'. DFV represents a gendered injustice at the 'front' and 'back' ends of the prison system; when exiting prison, women often navigate the risks associated with returning to a potentially violent relationship or face DFVrelated debt and homelessness, all of which increase their risk of returning to custody (ALRC 2017; Kilroy et al. 2013; State of Victoria 2016).
On rare occasions, lawyers observed that women might avoid making an application for bail altogether because of the absence of housing and support options or because of fear for her safety in the community. Lawyer 4 recounted that 'the few times that a woman has actively instructed me not to apply for bail is always for fear of her safety in the community, or she's just got nowhere else to go and this is four walls and it's a feed'. Here, the resignation to remand is not an indication of the prison's desirability but women's marked lack of safety and security in the community. In this context, prison can sometimes appear as 'respite' from the chaos borne of the constellating hardships outlined above (Russell and Rae 2020;Schneider 2021). This should not be taken to valorise or legitimise imprisonment for homeless women or victim-survivors (Russell and Gledhill 2012). Rather, the use of carceral systems to fill this 'gap' in safe housing and support systems represents an abject social failure.

Intervention Orders as Gateway to Remand
Intervention orders represent another avenue through which DFV can destabilise women's housing. Lawyer 6 described cross-or dual-applications of intervention orders, where each partner has taken an intervention order out against the other person, as 'quite common'. This can restrict a woman's options for housing, lessen her chances of being granted bail and increase her likelihood of being remanded. Lawyers 7 and 11 similarly remarked that there may be housing available, but an intervention order will prevent a woman from residing there. Some women are forced to decide between breaching an order by returning to their house or sleeping on the street and being 'exposed to the threat of harm as a female homeless person' (Lawyer 3). As Lawyer 2 stated: Intervention orders in general have gone up massively, and that's as a result of all the … changes around family violence. … There's a lot of nuance behind all of that stuff that is not… always positive for women who might have been victims, like they might end up with an order against them as well. … It might be the fact that they have actually done something that would meet the definition of family violence themselves. … But it excludes them from where they live, or they can't contact their kids because of it, and then you'll get repetitive breaches and that will put someone into a remand situation as well.
While an intervention order is typically a civil order, breaching one is a criminal offence. This effectively makes them 'quasi-criminal' or 'hybrid' civil-criminal laws (Nancarrow 2019). The above quote highlights that as their use increases, so too does the risk of criminalisation, including for women (Reeves 2021). Lawyer 11 recounted: We frequently get [told], 'She gives as good as she gets' [and] 'They're both as bad as each other'. … Those lines are very common. … It could be a situation where it's known to police that there have been intervention orders going both ways. They might come to the incident and see that both people are injured, and it might be whoever is more upfront with police and who is more willing to engage with police is the one that's believed and the one who is not remanded, or the one who is allowed to reside at the address where the other one is excluded.
Lawyer 6 reiterated that cross-applications of intervention orders create particular 'difficulties with the woman having somewhere to go. So, it all comes down to housing a lot of the time'. That is, once excluded IJCJ&SD 8 www.crimejusticejournal.com from housing due to the conditions of an intervention order, a woman in the position of applying for bail will be unlikely to succeed, unless alternative housing arrangements can be made.

DFV-Generated Isolation as a Barrier to Bail
Domestic and family violence cuts criminalised women off from support systems that might have enabled them to obtain bail and avoid remand in custody. As Lawyer 3 outlined: One of the first factors that characterises an abusive relationship is isolation from one's family. So, when [women] finally get to this point where they've been remanded, oftentimes [it is] because they've been in that family violence relationship, they're completely isolated from alternative family members that can come and give them support. … It is pretty easy to see the connection between family violence and them not being able to get bail. It's a gendered issue, 100 per cent.
Domestic and family violence can prevent women from staying engaged in the legal process, which can compound their criminalisation and increase their likelihood of eventual remand in custody. As Lawyer 1 pointed out, 'If somebody is in this really violent, unstable relationship, then getting to a court date on the shop theft might seem to have very little importance if every day you're living with the fear of whether you survive'. Lawyer 11 similarly discussed how DFV restricts criminalised women's capacity to navigate the legal system: Frequently, breaching your bail conditions could be in the context of experiences of family violence, whether it be the case that they're being prevented from going to court or their partners are hiding the letters from the lawyer or paperwork, or other kinds of experience of controlling behaviour.
Lawyers noted that in some women's cases, their partner will be the 'co-accused' person in the matter. This limit women's options for housing in a bail application because a magistrate will typically not allow a woman to continue to reside with a co-accused while on bail. Some lawyers suggested there is a gender imbalance in matters that involve 'co-accused' (heterosexual) partners, which leads to adverse outcomes for women. Lawyer 9 described this scenario as 'quite common', stating that 'I always thought this was a bit of a farce when I'd heard this before, but often women are taking charges, or there's a co-accused and she gets done and he doesn't'.
Without financial resources or other forms of support, victim-survivors can find themselves in the position of 'choosing' between living with an abusive partner or becoming homeless. 'The economic situation that women often find themselves in', Lawyer 1 emphasised, 'is so pivotal in terms of the choices that they have, or don't have'. As Lawyer 9 stated, 'poverty doesn't give her opportunities to find an alternative'. This, in turn, reduces a woman's capacity to avoid custodial remand since a lack of safe housing options severely disadvantages women in the bail and remand process. As lawyer 4 points out, 'It's very difficult to get a woman bail to live in a car with her kids'. For these lawyers, a woman's poverty and social marginalisation borne of DFV present significant obstacles in an application for bail, further highlighting the system's capacity to reproduce gender inequality.

Police Pursuing Other Matters on a DFV Call-out
When seeking help from police for DFV, criminalised and otherwise marginalised women are vulnerable to (further) criminalisation and remand since police practices play a crucial role in interpreting and enforcing laws 'on the ground'. This is particularly the case for Aboriginal and Torres Strait Islander women, women who use illicit drugs, migrant women from non-English-speaking backgrounds and women with disability (Nancarrow 2019;Reeves 2020Reeves , 2021Watego et al. 2021). Several lawyers we interviewed recounted instances of criminalised women ending up on remand after seeking help from police for DFV. This includes police pursuing a victim-survivor's other criminal matters, rather than responding to the DFV that she reported: IJCJ&SD 9 www.crimejusticejournal.com Lawyer 1: We've had situations where women have called police to get assistance with family violence and then been arrested on a shop theft warrant. Interviewer: Has the family violence been dealt with by police as well? Or no, just the warrant? Lawyer 1: Well … in their mind, [yes], because they've taken her out of that situation [by placing her] in custody. So, [the police] are going, 'great solution'.
The idea that prison can serve as a 'safer' place or a respite for women experiencing DFV or struggling with addiction or mental health issues was raised on numerous occasions during our fieldwork, both in interviews and in our observations of the court (Russell, Carlton and Tyson 2020a;Russell et al. 2020b). This deeply troubling assumption absolves social responsibility for the lack of housing and mental health and drug and alcohol support available for women in the community and ignores the surveillance, punishment and discipline that is inherent to the prison environment (McCulloch and Scraton 2009).
The re-criminalisation and punishment of victim-survivors seeking protection from police was a significant theme in our interviews. Lawyer 4 recounted another example from their legal practice: I had a client … constantly cycling in and out of custody. … She is so traumatised, just the victim of possibly the worst examples of domestic violence I've ever seen. To a point where she was complaining about a guy to police and 000 constantly. Police then charged her with using a carriage service to harass, calling 000.
In this instance and in others, a woman's trauma and deteriorating mental health is misunderstood and punished, regarded by police as a 'nuisance' or as evidence of 'dishonesty'. Already traumatised and 'cycling in and out of custody', the outcome of her help-seeking was further criminalisation and exclusion.

Police 'Misidentification' of the 'Predominant Aggressor' in DFV
The punishment of victim-survivors through criminal legal responses to DFV is increasingly recognised as a problem, as noted earlier, especially because of police practices of 'misidentifying' the predominant aggressor in an abusive relationship. Most of the lawyers we interviewed confirmed that misidentification is a contributor to the problem of women's criminalisation that places them at risk of remand. Lawyer 7 described misidentification as 'a huge problem' and recounted seeing 'a lot of women brought in before the court' that are victim-survivors. They explained that even women that have taken intervention orders out against their partner and been listed 'as the complainant and the victim' in 'a number of police matters' can be misidentified and criminalised, 'then for whatever reason, the police have been called, whether by the victim or someone else, and the police attend and charge the female' (Lawyer 7). Thus, despite having a history of victimisation on record with police, women may still encounter criminalisation. As Lawyer 3 reflected, 'oftentimes the assessment the police make of a situation is not exactly nuanced. … [Women] could even be confronted with a situation where they are the victims of prolonged family violence but they're the person that gets the notice slapped on them'.
Lawyers recounted how poor mental health can escalate the risk of misidentification for women. Lawyer 11 stated that 'Often, it's the case that you'll find a brief where "miss so-and-so" has mental health issues and so her version of events isn't deemed as credible. So, that's part of the case theory of the prosecution'. Lawyer 13 described a common scenario of misidentification: In this situation, the person that appears less emotional and affected by the violence can speak first to police and define the terms and dominant narrative of the event. This suggests critical problems in legal responses to DFV, which place gendered expectations on women to simultaneously embody constructs of rationality, vulnerability and innocence, to be recognised and protected as 'worthy' victims (Meyer 2016).
The limitations of these underlying expectations were frequently commented on by the lawyers we interviewed. Lawyer 13 provided an example from their work representing a migrant woman, who was unable to effectively communicate her experience to police due to language barriers: [The police] misidentified her as the primary aggressor. She had been whacking [her partner with her handbag, in public], and he was suave, he was a clever man. … She got remanded … and arguably she had committed family violence. I mean, you can't whack people with your bag. But it was in the context of her being victimised really seriously, with control of her passport. … Even despite the Royal Commission [into family violence] … the police got there, and it was late at night and they just went, 'intervention orders for everybody'. [Her partner] spoke very good English.
This example shows how migrant women from non-English-speaking backgrounds experiencing DFV may be particularly disadvantaged and at risk in their dealings with police, especially when an abusive partner is well positioned to leverage the system to their own benefit (Reeves 2020). As many critical researchers note, the capacity for police and other legal professionals to properly account for the overarching structural inequalities and power dynamics of coercion and control in abusive relationships remains highly questionable (Nancarrow 2019;Tolmie et al. 2018;Walklate and Fitzgibbon 2021).

Being Unrecognisable as a 'Victim'
Criminalised women often do not fit neatly into the binary categories of victim and offender that form the basis of the criminal legal paradigm. While the notion of 'misidentification' had currency among the lawyers we interviewed, it was also frequently qualified, for it can fail to capture more nuanced situations. Focusing excessively on misidentification has the potential to overemphasise women's 'innocence' and obscure concerns surrounding the criminalisation of women's self-defensive or trauma responses that result from longstanding violence and multidimensional oppression (Douglas et al. 2020;Tolmie et al. 2018). Lawyer 9 explained that: [Most] of the women that we've seen in that category [of misidentification], I would describe as otherwise having no other precursor/identifier to put them in the courts. They've got a job, they're in a relationship, probably in a house. They had a fight with 'Roger', someone gets hurt, the cops believe Roger, she gets charged, the charges ultimately all go away. … The ones that are more interesting I think are the ones where she has been treated like shit her whole life and one day does something to try and shift things in her favour, or defend herself, and police-there's no language for that that's available.
As this lawyer recounts, women's self-defensive and trauma responses are poorly understood and often met with punishment, rather than support. Several lawyers expressed uneasiness about relying on moralising ideas of 'innocence' to determine deservingness (Wang 2018), especially when considering criminalised women's experiences. Reminiscent of Nils Christie's (1986) oft-cited critique of the 'ideal victim', Lawyer 11 stated that they 'often see cases where women don't behave in what would stereotypically be considered a "victim model"'-such as calmly cooperating with police-which then disadvantages or endangers them in their dealings with police and restricts their access to support services (Meyer 2016;Watego et al. 2021). Poor prior experiences with police can also lead women to avoid contacting police altogether since 'they think that, well, there's no point because I'm never believed' (Lawyer 11).
Lawyer 1 noted a pervasive conception of victim-survivors as 'less deserving' of protection if they have a criminal record or are otherwise engaged in illegal activities, such as illicit drug use. They referred to these women as 'other victims': Lawyer 1: When it's obvious that the woman is a victim of family violence and she's a sympathetic character, that is something that [the police] can invest in very easily. But when IJCJ&SD 11 www.crimejusticejournal.com you bring in other levels, and the woman is offending with that man, or is struggling to get to [appointments] because of the environment that she is in, then those levels of sympathy [wane] …. It's the 'other victims' who are … left in a much more vulnerable position. They might not be as sympathetic, they might be really 'mouthy' … they're battling their way through and sometimes they're exhibiting violent behaviours as well, but it's about stepping back from that and [asking], 'where does that come from?' … I think when you add the offending layer…. They become somehow less deserving. Interviewer: They're not innocent enough. Lawyer 1: That's right, just horrifying.
Lawyer 12 expressed a similar frustration with police dealings with victim-survivors that are criminalised, stating: There's a real mentality of being considerate of the victim, and as soon as that shifts, so as soon as a woman is then seen as being the offender … they're not afforded that respect that they were when they were the victim because they're the offending party.
Lawyers described the 'compound effect' of being viewed as an 'offender' by police, whereby an initial intervention order can quickly escalate to criminalisation and remand. Lawyer 8 stated: Once someone has been identified or potentially misidentified as the person responsible for that violence … when [police are] attending later incidents it's much easier to say, 'there was a family violence safety notice in place, this woman breached it by reattending this address, so we will breach her again'. That is the compound effect. … Rather than interrogating those complex dynamics [of DFV], once that order is in place and it has been breached in a concrete way by attendance … the criminalisation happens so quickly. Then [there is] the flow-on trauma of women in custody, because … they've experienced violence and [are] not getting the assistance that they need.
This lawyer's account encapsulates how police responses to DFV can exacerbate a victim-survivor's trauma, isolation and lack of access to support. The 'compound effect' of criminalisation under strict bail laws, which can swiftly result in a period of remand, not only fails to deal with the dynamics of intimate violence, it also sets women up to become enmeshed in cycles of incarceration.

Conclusion
The findings from our research suggest that for some of the most socially and economically marginalised women, there is a nexus between DFV, homelessness and criminalisation that is exacerbated by strict bail laws. Such bail laws put the onus on women to justify their bail, rather than requiring judicial officers to justify remand prior to sentence. Defence and duty lawyers practising in Victoria report that this is particularly onerous for women experiencing socio-economic disadvantage and gender-based violence. Our analysis shows how bail reforms intended to enhance community safety can instead entrench gender and racial inequalities. Halting and reversing the trend of increasing remand of women requires, at the very least, significantly rethinking and reworking bail laws to make imprisonment a last resort.
Homelessness and DFV are often interrelated for women, especially for criminalised women. Through an investigation of the barriers that women face in the bail and remand system, our research suggests that women are routinely disadvantaged by virtue of gendered and racialised patterns of violence and inequality. We highlighted several specific ways that police and court responses to housing insecurity and DFV can heighten women's prospects of being criminalised, re-criminalised and remanded, which can lead to longer-term cycles of incarceration. These cycles have flow-on effects beyond individual women; they affect families, children, and society more broadly. High rates of remand are costly, and the repetition of short, unpredictable stays in prison-sometimes in isolation-is destabilising and often retraumatising, IJCJ&SD 12 www.crimejusticejournal.com undermining any attempt to enhance community safety through bail reform. As many other researchers and advocates have pointed out before (e.g., Moore, Scraton and Wahidin 2017), expanding holistic and community-based support systems, such as public housing, mental health and drug and alcohol treatment, will be more effective to avoid the long-term criminalisation of women that can be set in train by stringent bail and remand processes. This requires deemphasising and restraining policing, criminalisation and court processing that leads to the exclusionary use of remand as a short-term 'solution' for issues borne of extreme disadvantage and gendered violence.
Our research further indicates that the unprecedented growth in women's remand cannot be considered in isolation from the push towards escalating the policing and punishment of DFV. Any consideration or call for expanding the net of DFV criminalisation needs to reckon with the capacity for these laws to capture and punish women, especially Aboriginal and Torres Strait Islander women, who are victimsurvivors (Watego et al. 2021). Criminalised women face unique risks when encountering police, even when experiencing victimisation. Our data lends support to the findings of prior research that misidentification and 'systems abuse' remain key problems in criminal legal system responses to DFV. However, our research also suggests that it is not simply a problem of misapplication or misuse of the law, or poor police training in DFV. More fundamentally, criminalised women do not fit neatly into the 'innocent victim' category that the criminal legal system projects and expects of women who have been abused. Women already known to police as 'offenders', or who are unable to effectively communicate with police due to deteriorating mental health, language barriers or trauma responses, can quickly become caught up in restrictive civil-criminal legal webs comprised of intervention orders and breaches that lead to remand. These topics require further research, including with those working outside of the legal profession, such as in the family violence, homelessness and mental health sectors. Our study makes clear that stringent bail systems capture and punish women who are victim-survivors, retraumatising them and compounding conditions of isolation and poverty.