The Domestic Violence Protection Order System as Entry to the Criminal Justice System for Aboriginal and Torres Strait Islander People

The domestic violence protection order (DVO) system is a hybrid system of criminalisation in which the DVO itself is a civil order, but any contravention of that order may result in a criminal charge. Limited attention has been paid to the potential consequences of criminalisation through the hybrid DVO system in the Australian context. We use Queensland, Australia, as a case study and examine administrative data gathered through Queensland courts. We show that a disproportionate number of Aboriginal and Torres Strait Islander (ATSI) people are named on DVOs, charged with contraventions of DVOs and significantly more likely than non‐Indigenous people to receive a sentence of imprisonment for a contravention of a DVO, compared to non‐Indigenous people. We find that ATSI women are particularly overrepresented in this system. We review explanations for these startling figures and emphasise the need for a change in approach.


Introduction
Domestic and family violence (DFV) is an issue of national concern in Australia with approximately one woman killed every week by a current or former intimate partner (Cox 2012). The most common legal response to DFV in Australia is to obtain a civil domestic violence protection order (DVO) and over 30,000 such applications were made in Queensland in the [2015][2016] year (Magistrates Courts of Queensland 2016: 21). The DVO system is a hybrid system of criminalisation in which the DVO itself is a civil order, but any contravention of that order may result in a criminal charge. The prima facie aim of a system combining civil DVOs with the threat of a criminal charge for a contravention is to promote the safety and protection of the victim through deterrence for non-compliance with the DVO. It is well recognised that the outcomes of criminal justice policies are often variable across groups and may at times produce results disconnected from the initial aims (Crutchfield 2016). There has been significant scholarly attention paid to the unintended consequences of various criminal justice responses to DFV. For example, while the United States (US) evidence is mixed and complex, some studies show greater criminalisation of domestic violence victims, primarily women, through pro-arrest policies (Chesney-Lind 2002;Miller 2001) and mandated victim participation in DFV prosecutions (Hanna 1996;Sack 2004). Some studies have looked at the factors that are present in dual arrest cases. Studies have identified that DFV incidents that involve alcohol or drugs are more likely to result in dual arrest (Muftic, Bouffard and Bouffard 2007;Roark 2016), with one study finding that women's odds of dual arrest increased when her male partner alone was under the influence of drugs or alcohol (Roark 2016). Limited attention has been paid to the potential consequences of criminalisation through the hybrid DVO system in the Australian context. However, Cunneen (2010) and Nancarrow (2016) have observed that the civil DVO system is particularly likely to enmesh Aboriginal and Torres Strait Islander (ATSI) people within the criminal justice system. There is a persistent and significant overrepresentation of ATSI women as victims of DFV (Judicial Council on Cultural Diversity 2016: 6), and a significant overrepresentation of ATSI people as offenders in Australian criminal justice systems more broadly (Cunneen 2013). Of particular concern is that the imprisonment rate of women has doubled in the past decade and this is almost entirely due to the imprisonment of ATSI women (Australian Bureau of Statistics (ABS) 2016). In this paper, we further investigate the possible role of the DVO system in this overrepresentation.
We use Queensland as a case study through an examination of administrative data gathered through Queensland courts. We focus our attention on the ATSI and gender distributions among DVO respondents (those named as perpetrators on DVO applications), and those charged with, and/or found guilty of, a contravention of a DVO. Our article makes a number of important findings in the Queensland context. It identifies that a disproportionate number of ATSI people are named on DVOs (as both aggrieved and respondent) and subsequently charged with contravention of a DVO, compared to non-Indigenous people. Moreover, our analysis shows not only that ATSI women are overrepresented in these charges compared to non-Indigenous women but also, in particular, that 69 per cent of women who were sentenced to serve a period of imprisonment for a contravention of a DVO in the 2013-2014 year were ATSI women. Given that, in the 2011 census, ATSI women accounted for 3.3 per cent of all Queensland women (Queensland Treasury and Trade 2011) this demonstrates a startling overrepresentation.
We begin with a brief overview of the DVO system. We then explain our methodology and its limitations and present our analysis of the Queensland data. In the final sections, we consider a range of overlapping and intersecting theories that might help to understand the picture the data present.

Gradual shifts in the DVO system
Since the 1980s, all Australian states and Territories have introduced legislation that provides a civil legal remedy for a person experiencing DFV. DVOs are now the most common legal response to DFV in Australia (Douglas and Fitzgerald 2013: 57-58 DFV can apply to the lower courts for a DVO to prohibit a perpetrator of DFV (in Queensland referred to as a 'respondent') from committing further acts of DFV against the person (in Queensland referred to as an 'aggrieved'). The police can also apply on behalf of the aggrieved for the DVO.
The definition of DFV is broad in Queensland and encompasses emotional and financial abuse, sexual and physical violence and behaviours that are threatening, coercive or controlling. 2 DVOs issued in Queensland include a standard condition that the respondent should be of good behaviour and not commit DFV (s 53 Domestic and Family Violence Protection Act 2012 (Qld)) and many DVOs include conditions tailored to the circumstances of the aggrieved. Conditions may include requirements that the respondent does not contact or come within a certain distance of the aggrieved. Depending on the conditions of the DVO, a contravention offence may be charged in a wide range of circumstances, from a telephone call or visiting the aggrieved to physical violence towards the aggrieved.
While a DVO is a civil order designed to protect the applicant from future harm, where the respondent has contravened conditions of a DVO, this may result in the respondent being charged with a criminal offence of contravention of the DVO. Previous research has identified that contravention charges often proceed in place of substantive offences such as assault (Douglas 2008). Similar to sentencing for other offences, relevant considerations include prior convictions for contraventions, the impact on the victim and the gravity of the behaviour (Australasian Institute of Judicial Administration (AIJA) 2017: [9.3.1]). Where multiple charges arising out of the same conduct are pursued-for example, both assault and contravention-double punishment will be a consideration for the sentencing magistrate or judge to take into account (AIJA 2017: [9.3]).
The police are often the first responders to a call about DFV and play a pivotal role in this context. The recent Victorian Royal Commission into Family Violence noted that '[f]or many women and their children, police not only provide protection at a time of crisis but are the entry point to the broader family violence system. The quality of the police response is therefore crucial' (2016: 34). Police exercise their discretion and apply for DVOs on behalf of aggrieved people in over 70 per cent of cases and, where the police are the applicant, the DVO is more likely to be granted (Douglas and Fitzgerald 2013). Police also exercise their discretion regarding whether to investigate and charge a contravention of a DVO.
The penalties available for contravention of a DVO vary across Australia but, in all jurisdictions, it is ultimately possible for the offender to be imprisoned in response to a conviction for a contravention of a DVO. Previous research has identified that penalties for contravention are most commonly fines (Douglas 2008;Sentencing Advisory Council 2009: [3.37]). However, a significant proportion of contravention charges do result in a penalty of imprisonment. In a review of penalties imposed for contraventions of DVOs in Victoria, 16 per cent received a custodial sentence (Sentencing Advisory Council 2015: 31); in New South Wales (NSW), the figure was 12.4 per cent (Trimboli 2015). In their research in New Zealand, Towns and Scott found that around 10 per cent to 14 per cent of those convicted for contraventions of DVOs received a custodial sentence (Towns and Scott 2006: 158).
The purposes of sentencing include retribution, deterrence, rehabilitation, denunciation and community protection. 3 However, courts in all jurisdictions have emphasised the seriousness of DFV and that the need to deter future DFV is an important aim of sentencing for offences committed in this context (AIJA 2017 [9.5] essential that those orders are not ignored. When they are repeatedly breached, the need for general and individual deterrence will ordinarily outweigh subjective and other mitigating considerations. Statutory maximum penalties for contravention of a DVO in Queensland have gradually increased over time. From 2008 until 2012, the maximum penalty for a contravention was one-year imprisonment for a first offence (s80(1) of the Domestic and Family Violence Protection Act 1989 (Qld)) and, in circumstances where there were two or more prior convictions within the past three years, the maximum penalty was two years' imprisonment. In September 2012, the maximum penalty for a contravention was increased to two years' imprisonment for a first offence and three years' imprisonment if the offender had been previously convicted within the past five years. 4 The Queensland data In this section, we present ATSI and gender distributions for individuals named as DVO respondents, as well as those charged and or found guilty of DVO contraventions in Queensland. These figures are drawn from the Queensland Wide Interlinked Courts (QWIC) system, an administrative data source containing the elementary information required to process offenders through court, including information about court appearances, and outcomes (for example, sentences) for criminal matters. The data were provided by the Statistical Analysis Unit of the Department of Justice (DJAG) and include information on DVO applications and contraventions in the years 2008-09 to 2013-14. For the purposes of this brief analysis, we focus primarily on the latest year of available data, 2013-14. Since it is not possible to link specific DVOs with contraventions, we present results for DVOs and contraventions separately.
The QWIC database captures information on individuals (using a single person identifier). To avoid the possibility of over-counting individual characteristics-for example, for those who may be accused of multiple DVOs or contraventions-we present results based on a count of unique-DVO respondents and unique-contravention defendants.
At the time our data were collected, DVO contraventions were the only domestic violence-related criminal matters recorded through the QWIC system. Details of the contravention behaviour or any other offence connected to the DVO contravention (for example, assault) were not available.
The QWIC database includes an ATSI status flag for individuals identified as DVO respondents, and/or defendants for DVO contravention charges. The classification of individuals into ATSI and non-ATSI categories in criminal justice data typically requires 'self-identification' by an individual who comes in contact with the police, and data quality can be affected by individuals' willingness to respond as well as police consistency in asking the question (ABS 2015: note 60). It is important to note that evidence suggests that ATSI status is generally under-reported in criminal justice statistics (Hunter and Ayar 2009;Kennedy, Howell and Breckell 2009). In the QWIC data, the proportion of 'unknown' ATSI status individuals was higher for data related to DVOs than contraventions. For example, ATSI status was unknown for about 22 per cent of DVO respondents (n = 5,245) compared to less than 1 per cent (n = 58) of defendants facing contravention orders. One possible explanation for this discrepancy is that, at the DVO application stage, a respondent's ATSI status is an optional category on the form completed by the aggrieved or by the police on his or her behalf. However, individuals who have progressed to the stage of facing contravention charges in court will have had greater opportunity and perhaps incentive to self-identify as a person of ATSI status. Given the greater weight of the unknown ATSI category in relation to DVO applications, in the results presented here, we maintained ATSI status unknown in the total for the analyses-related DVO applications and orders, but we excluded the smaller ATSI status unknown figure from the total in the final section on contraventions of DVOs.

DVO applications and orders
In 2013-14, DVO applications lodged in Queensland were associated with a total of 23,492 unique respondents with at least one order in the year. 5 ATSI people accounted for about one in five (21%, n = 4,891) of these respondents. This represents a significant overrepresentation of ATSI people, given their approximate 3.6 per cent share of the Queensland population (Queensland Treasury and Trade 2011). Men accounted for a larger proportion (80%) of all respondents than women (20%); however, among ATSI people, women's share of respondents was slightly higher (23%).
ATSI respondents were also overrepresented relative to respondents of the same gender. In particular, ATSI women accounted for just over 24 per cent of all women respondents, whereas ATSI men accounted for 20 per cent of all male respondents.
Examining the order outcomes associated with each respondent showed that, in most cases (88.7%), orders against respondents were granted rather than dismissed or withdrawn (11.3%).
Orders against all ATSI respondents were granted (88.9%) and dismissed (11.1%) in roughly similar proportions; however, there were gender differences in these figures. Orders were granted against ATSI women in 92 per cent of applications, in contrast to 88 per cent of applications for men.
Finally, we examined patterns of police involvement in the initial DVO application process. Here, we limited the analysis to the 20,829 defendants whose order outcome was granted, and consider ATSI-status and gender differences in cases that originated as either police-lodged or aggrievedlodged DVO applications. Overall, the police lodged the application for a majority (79%) of respondents, but this figure was higher among ATSI respondents (90%). The proportion of police-lodged applications was roughly similar for ATSI women (90.3%) and ATSI men (90.2%).
The within-gender overrepresentation of ATSI women among respondents whose initial application commenced with police involvement is also evident. ATSI women respondents comprised about 28 per cent of all women for whom the initial application was lodged by the police, while ATSI men accounted for 22 per cent.

Contraventions of DVOs
In 2013-2014, there were 6,888 unique defendants in total facing DVO contravention charges in the state (Table 1). Overall, men accounted for the majority (86.6%, n = 5,967) of these defendants, while women accounted for about 13.4 per cent (n = 921). During the year, ATSI people comprised over one-third (34%, n = 2,322) of all defendants. Among all contravention defendants, most (87.7%) were found guilty, a roughly similar proportion as for all ATSI defendants (88.8%). ATSI people were also overrepresented among defendants found guilty of a DVO contravention. This was particularly the case for ATSI women who accounted for a slightly greater share (39.4%) of all female defendants found guilty, than was the case for ATSI men's share (33.3%) of all male defendants.
Among all defendants found guilty of a contravention, most (39%) received monetary orders, followed by community orders (28%), which were primarily made up of probation orders but also included orders such as community service, parole and Intensive Correction Orders (ICOs).
Just over one-quarter (27%) of all defendants received a custodial order; however, this figure rose to 43 per cent for all ATSI defendants. In contrast, the proportions of ATSI defendants receiving monetary (23%) or community (26%) orders were below the total sample proportions for these order types (39% and 28%, respectively).
The custodial order share is driven particularly by ATSI men who made up about 49 per cent of all defendants receiving custody, with about 53 per cent of all men receiving custody. In fact, nearly one-half (47%) of all orders that ATSI men received for a contravention were custodial.
Among ATSI women, custodial orders comprised 22 per cent, a large difference relative to the proportion of custodial orders for non-Indigenous women (6.5%). Among all women who received a custodial order, ATSI women accounted for a large majority (69%). To further investigate the within gender share of custodial orders for ATSI defendants, we looked at annual unique defendant figures over the six-year period from 2008-09 to 2013-14. Figure 1 shows that, over the period, ATSI defendants have comprised a consistently large proportion of custodial orders among both men and women. However, the share among female defendants for ATSI women, in particular, rose from 52 per cent in 2008-09 to a high of 74 per cent in 2011-12 before declining slightly to 69 per cent in 2013-14.

Understanding the contravention picture-in the shadow of colonisation
The QWIC data tell a familiar story about ATSI people's overrepresentation in the justice system (Bartels 2010;Weatherburn 2015) and add new knowledge about the persistent and high overrepresentation of ATSI people in the DVO system. Specifically, it shows that ATSI people are over-represented at each stage of the system, including as named respondents on DVO applications, as respondents for whom DV orders are granted, and as respondents for whom there was initial police involvement in the application. ATSI people are also overrepresented among those sentenced for a charge of contravention of a DVO, as well those who receive a custodial order for contravention offences. The QWIC data results also underscore gender differences in the overrepresentation of ATSI people that fit with a worrying Australian trend of ATSI women's rising criminal justice involvement both as victims and perpetrators (ABS 2016). Notes: Represents ATSI proportion of annual total number of unique defendants for each gender. Where defendants had more than one contravention in the year, the last contravention outcome was counted. Excludes <1% of cases for which ATSI status is unknown.

Source: Queensland Wide Interlinked Courts (QWIC) system
We know that factors including the oppressions associated with race, gender, class and geographical location combine to provide an intersectional, layered identity and social location (Crenshaw 1990;Stubbs and Tolmie 2008). Recently, Crenshaw has also emphasised that 'intersectional dynamics are not static, but neither are they untethered from history, context, or social identity ' (2012: 1426). While there are several overlapping and intersecting accounts offered for why ATSI people, especially ATSI women, are overrepresented in the prosecution of criminal offences and in imprisonment statistics generally, the myriad of interrelated factors associated with Australia's history of colonisation frame all of these. Colonisation strategies included removal from land and forcible takings of land, intergenerational child removal and lack of recognition of ATSI law (Cunneen 2013: 377;2010: 26-32). The fragmentation of kinship systems and breakdown of culture, social and economic marginalisation, homelessness, and drug and alcohol abuse have followed for many ATSI people (Gordon, Hallahan and Henry 2002: 54). Connell (2014) points to the relationship between colonisation and gendered violence. She observes that 'the colonial state was built as a power structure operated by men, based on continuing force' and she posits a direct relationship between this history and the continuation of gendered violence in the present, with gender and race being inseparable in that experience (Connell 2014: 556; see also Behrendt 1993 andNancarrow 2016: 51-53).
Researchers have identified the increasing frequency and severity of DFV experienced by ATSI women (Cheers et al. 2006: 52;Olsen and Lovett 2015: 2;Snowball and Weatherburn 2008).
Notably hospitalisations for non-fatal family violence-related assaults perpetrated against ATSI women in 2012-13 were 34 times the rate for non-Indigenous women (Judicial Council on Cultural Diversity 2016: 6), and ATSI women living in rural and remote Australian communities experience DFV, especially physical and sexual violence, at particularly high rates (Australian Institute of Health and Welfare 2015: 44-46). While ATSI women are at the receiving end of increased frequency and severity of DFV, we demonstrate here that they are also significantly overrepresented in the QWIC data, compared with non-Indigenous women, as respondents to DVOs, in contravention prosecutions and in sentences of imprisonment for contravention.
Researchers have considered the different cultural mores that influence ATSI women not only to use violence but also to sometimes fight back, in circumstances where non-Indigenous women might remain more passive. Researchers have also identified ATSI women's reduced access to services and support as explanations for these statistics. However, one of the key explanations for ATSI overrepresentation across genders may be that the criminal justice system treats ATSI people differently. In the context of contravention charges, this is reflected at both the level of policing and prosecution but also at the sentencing level. In this context, we argue that the political commitment to getting tough on DFV is now being reflected in legislative sentencing responses and associated judicial decision-making and has a disproportionate impact on ATSI people. These explanations are considered further below.

Women as both victimised and criminalised
Some research has suggested that ATSI women are more likely to use physical violence and fight back in response to violence than non-Indigenous women (Langton 1988). ATSI women's use of physical violence may increase the likelihood that DVOs will be made against them by the police, which, in turn, increases their chances of being charged with contraventions. The use of violence in the commission of a contravention may also contribute to higher sentences. In 2013-14, illicit drug-related offences comprised the most common index offence category for imprisoned non-Indigenous women (26%), while 'acts intended to cause injury' or assault was the most common category (31%) for ATSI women (ABS 2016).
Earlier research by Yeo (1996) posited that ATSI women may be reluctant to call the police and may choose instead to respond to an attack with violence. This reluctance may result from fears, underscored by personal or community experience, of negative police engagement or children being removed (Willis 2011). More recently, the Aboriginal and Torres Strait Islander Social Justice Commissioner has supported Yeo's comments, citing anecdotal evidence that ATSI women's higher arrest rates for violent offences result from the greater likelihood of their using violence to protect themselves in the absence of what they perceive to be adequate police protection (2002: 29; Bartels 2010). Interviews with incarcerated Aboriginal mothers in Western Australia found that most of the women who used violence had also been victims of violence (Wilson et al. 2017). While using violence in self-defence may be more common among ATSI women, it is also more likely to place them at greater risk of an escalated violent response by an abuser and this may help to explain the large overrepresentation of ATSI women as victims of assault and homicide (Cussen and Bryant 2015;Kerley and Cunneen 1995).
The connection between victimisation and violence among ATSI women has also been made in recent Queensland research using administrative and police data for 185 complete domestic violence histories for people dealt with in the Mount Isa and Cairns Magistrates courts (Nancarrow 2016). Mount Isa is a regional town of 21,000 people located over 1,800 km from Brisbane, the capital city of Queensland. Cairns is a large coastal town of 140,000 people located approximately 1,600 km from Brisbane. These data point to the greater vulnerability of ATSI women with higher chances of having complex case histories involving victimisation by multiple perpetrators, and situations in which their perpetration of violence is most often preceded by their own victimisation (Nancarrow 2016: 116). To add further context, Nancarrow's interviews with legal and justice system professionals, some of whom were ATSI people, suggest four underlying reasons for ATSI women's greater propensity for violence (2016: 116-128). First, interviewees identified the regular occurrence of 'fights' including fights provoked by women, defensive fights to resist violence and generally using violence to resolve problems. Second, most also identified the day-to-day environment of 'chaos' associated with the legacy of colonisation, including disadvantage, lack of rules, lack of priority given to court orders and more generalised fear, powerlessness and lack of trust of police. Third, the police response was often implicated in the legal outcomes for ATSI women. For example, police often employed a 'formulaic application of domestic violence legislation' (Nancarrow 2016: 127) regardless of whether they were responding to a pattern of coercive control, responding to fights arising from anger, or simply resolving disputes in circumstances where the criminal response may be more appropriate. In many cases ATSI women with complex cases, including both victimisation and violence, were not well served by these methods. Finally, Nancarrow identifies that 'race relations' was identified by some study interviewees as an explanation for DFV between ATSI people. Grouped under the general heading of 'race relations', Nancarrow identifies concerns held by ATSI people about under-policing/over-policing, ATSI people's mistrust of police, a lack of police awareness of ATSI culture and police bias against ATSI women (Nancarrow 2016: 125-127).
Researchers have also pointed to practical issues confronting ATSI women living in rural and remote areas including potentially being less likely to have access to services (Blagg, Bluett-Boyd and Williams 2015;Owen and Carrington 2015). A lack of services clearly leads not only to reduced access to help, including safe housing and financial and legal aid, but also to police and court intervention. Even where services are available, they may be culturally inappropriate and interpreters may not be available (Cunneen 2010: 25). A lack of appropriate services may also contribute to ATSI women taking self-help measures and using defensive force as a protection strategy.
When ATSI women fight back, they may be perceived to deviate from the stock expectations of the 'perfect victim' (McMahon and Pence 2003), or how a woman 'normally' behaves in response to DFV, and this may influence arrest and sentencing decisions. For example, some police and sentencing justices may expect women to be passive, vulnerable and unable to help themselves (Stubbs and Tolmie 2008) and women who fight back unsettle these expectations. Notably, Hester's (2013: 628) research has identified that, when women are violent, they are three times more likely than men to be arrested. She describes this as 'gender injustice' (Hester 2013: 634). Similarly, Dell (2001) found that Canadian Aboriginal women who were labelled 'violent' received particularly harsh treatment from the criminal justice system, and Stubbs and Tolmie (2008) have made this point in their analysis of cases where ATSI women have been charged with homicide.

Differential treatment at the policing-level
Given the consistent evidence of ATSI people being overrepresented at every stage of the criminal justice system as well as in social welfare and child protection systems (Cunneen 2013: 388-389;Weatherburn 2008), the overrepresentation of ATSI people in the DVO system is not surprising. Of course, a contravention of a DVO can only be charged when a DVO is in place. The Queensland data presented here indicate that ATSI people, and particularly women, are overrepresented among respondents for whom the initial DVO application is lodged by the police. This finding has also been previously observed in Queensland samples. For example, Cunneen (2010: 59-65) has reported that, in Queensland, ATSI people are nearly six times more likely than non-ATSI people to be an aggrieved in an application for a DVO, and that police were more likely to be the applicant in cases involving an ATSI aggrieved. Similarly, Nancarrow found that police had made the applications for DVOs on behalf of ATSI women in 97 per cent of the cases she examined compared with 79 per cent of cases involving non-ATSI women (2016: 111). She also found that applications were frequently made by police on behalf of both parties leading, in many cases, to cross-orders (Nancarrow 2016: 97). Where cross-orders are in place, both parties may be charged with a breach of the order, potentially further enmeshing both parties in the criminal justice system (Douglas and Fitzgerald 2013 The overrepresentation of ATSI people as aggrieved and respondent in DVO applications leads to overrepresentation in orders being made, especially where applications are made by police (Douglas and Fitzgerald 2013). Cunneen (2010: 85) found that DVO cases involving ATSI aggrieved were more likely to be uncontested and courts were more likely than not to grant the order-and more likely to do so in the absence of the respondent than in cases involving non-ATSI people. A DVO necessarily puts a person at greater risk of overrepresentation in contravention of DVO charges (Victoria Legal Aid 2015).
The high rates of overrepresentation of ATSI people charged with contravention offences may reflect more serious levels of DFV (Cunneen 2010: 69-70;Nancarrow 2016: 99). However, research suggests that attempting to understand ATSI overrepresentation by asking whether the root cause lies in the offenders' behaviour versus the processing of those offenders are unhelpful ways to frame the discussion (Piquero 2008: 60). Analogously, in empirical research into ATSI status and the risk of bail refusal, Weatherburn and Snowball (2012) found that ATSI people were nearly two times more likely than others to be refused bail. The authors also noted that ATSI people have much longer criminal records and are more than twice as likely to have 10 prior convictions and more than three times more likely to have 12 or more prior convictions (Weatherburn and Snowball 2012: 57).
Weatherburn and Snowball speculate that these greater numbers of bail refusals might be due to a perception that ATSI people are more likely to breach bail conditions. However, these numbers may instead reflect more intense police scrutiny of ATSI people (Weatherburn and Snowball 2012: 57). Research also points to differential treatment of ATSI people in relation to the use of DVO contravention charges. In the DVO context specifically, a study in NSW found that male, young and Indigenous offenders were charged with breach of a DVO sooner than other defendants after a DVO was ordered (Poynton et al. 2016: 5). The overrepresentation of police applications for DVOs involving ATSI people may suggest increased scrutiny and lead, inexorably, to a greater likelihood of behaviour being further scrutinised and contravention charges laid. As Piquero (2008: 59, 69) observes, police are afforded a high level of discretion and may be exercising this discretion differently in their responses to violence involving ATSI people compared to non-Indigenous people, contributing to the enmeshment of ATSI people in the criminal justice system via the DVO system.

Differential treatment at the sentencing-level
The over representation of ATSI people sentenced to imprisonment generally has been identified in previous research (Cunneen 2013). Consistent with this previous research, the QWIC data point to an overrepresentation of ATSI people in imprisonment outcomes for contravention charges. As noted earlier, it is possible that ATSI people are engaging in higher levels of DFV and have longer criminal histories. These factors may, in part, contribute to higher sentences.
However, from their analysis of a group of NSW cases flagged as relating to domestic violence offences and processed between 2009 to 2012, Jeffries and Bond concluded that Aboriginal 'violent offenders were more likely to be sentenced to prison than similarly situated non-Indigenous defendants' (Jeffries and Bond 2015: 466;Piquero 2008: 65). Notably, contravention charges are usually heard in the Magistrates Courts. Other research undertaken by Jeffries and Bond suggests that the location of legal processing of charges may influence ATSI overrepresentation. In their analysis of sentencing cases in the lower courts in NSW, Western Australia and Queensland, Jeffries (2011, 2012: 181) examined the cases from the focal concerns perspective. Bond and Jeffries (2011: 258) state 'this framework argues that, in making sentencing decisions, judicial officers are motivated by three focal concerns: blameworthiness, community protection or risk, and practical constraints and consequences. They suggest that sentencing decisions at the lower court level are likely to be affected by practical constraints; for example, the limited availability of time and of contextual information (Bond and  273). They argue that the constraints on magistrates in the lower court may make magistrates more likely to use 'perceptual shorthand', pointing out that there are certain perceptual cues that are often linked to racialised interpretations of criminal behaviour, rehabilitative capacity and social relationships (Bond and Jeffries 2011). 6 Bond and Jeffries note, for example, that in the US 'young black / Latino' has been identified as a perceptual shorthand for being crime prone or dangerous Jeffries 2011, citing Steffensmeier, Ulmer andKramer 1998). Similarly, Cunneen points to current justice processes built around risk assessment and therefore singling out ATSI people as a 'crime-prone' population (2010: 387, 391). These implicit biases may be operating on the minds of sentencing justices (Greenwald and Krieger 2006) in response to ATSI offenders. In Queensland, contravention charges are usually disposed of within a few minutes. These factors may provide further explanation for the higher penalties being applied to ATSI men and women charged with contravention offences.
The maximum penalty for contravention of a DVO has increased steadily in the past ten years and with the increase in penalty has come increasingly strong rhetoric from judicial officers to deter and denounce DFV and to protect women and children from it (AIJA 2017: [9.3.1]). Previous reports have identified that ATSI women prioritise stopping violence as the most important priority for the justice response (Hennessey and Willie 2006: 7-9;Nancarrow 2006;Robertson 2000: [4.7.8]), pointing to the least punitive aim of rehabilitation, in the sense of stopping the violence, as the preferred sentencing aim. However, the courts have found that sentences that focus on deterring future violence and community protection reflect the priority aims of sentencing in the context of DFV. The focus on these aims may encourage a more punitive approach.
The courts have largely resisted taking into account the effects of colonisation generally and its ensuing particular disadvantage in sentencing, finding it difficult to reconcile individualised justice with 'Indigenous justice' (Anthony, Bartels and Hopkins 2015 While Rothman does not suggest an alternative sentence in the Munda case, he points to the need for a proper consideration of the history and effect of colonisation in sentencing. His analysis identifies that all ATSI people have experienced colonisation and discrimination. Thus, recognising these factors allows for substantively equal treatment rather than negative racial stereotyping, an interpretation that is not supported by the Munda case.

Discussion and conclusion
Our analysis of Queensland courts administrative data shows that ATSI people are significantly overrepresented within the DVO system, as both aggrieved and respondent, at the application stage, in contravention charges and in resulting imprisonment outcomes. The overrepresentation of ATSI women as compared to non-Indigenous women in this system, and especially with respect to imprisonment statistics, is particularly stark. Coker and Maquoid have made the link between DFV and hyper-incarceration in the US, noting the biggest increase is among African-American women (2015: 588). In the US, one factor that contributes to the vulnerability of women, especially African-American women, to incarceration is abuse by men (Lipsitz 2012). This may well be the case in Australia.
Our analysis suggests that the hybrid DVO system contributes to the further enmeshment of ATSI people in criminal justice processes. Of particular concern is the overrepresentation of ATSI women in the DVO system. The Queensland findings reflect 2016 Australian prison statistics that show that the number of ATSI women in prison has risen by 20 per cent since 2011, compared with a rise of three per cent for non-Indigenous women (ABS 2016). We have outlined several intersecting and overlapping explanations for these statistics, including issues with policing and in decisions around sentencing. However, in the US context, Crenshaw has warned that neither 'race-based' nor 'gender sensitive' discourses have effectively served women of colour, pointing out that surveillance and punishment are intersectionally scripted in many ways, 'including the ways in which race, gender, or class hierarchies structure the backdrop against which punitive policies interact ' (2012: 1424, 1427). Crenshaw has identified that, while there is increasing knowledge and public awareness about the mass-incarceration of black men and boys in the US, women and girls of colour are largely marginalised from this discourse (2012: 1430, 1434). She argues that, while it is important to recognise and respond to the sheer numbers of black men and boys in custody in the US, their numbers mask the fact that racial disparities in incarceration are similar for black women as compared to white women in the US (Crenshaw 2012(Crenshaw : 1437. Because black women and, in Australia, ATSI women are more likely to be poor and the primary carer of children, they may be more likely to be the subject of surveillance by authorities including police and child protection services (Crenshaw 2012(Crenshaw : 1441. Such heightened surveillance puts black women at particular risk of criminalisation especially where there is a high level of discretion involved in the exercise of power (Crenshaw 2012(Crenshaw : 1441. This analysis resonates with the Australian context, especially in the context of the DVO system. ATSI women are disproportionately represented in DVOs, often because the police have taken out an order on their behalf as an aggrieved or against them as a respondent (or both) and wide discretion can be exercised with respect to charges under this hybrid system.
While space precludes a deeper discussion of possible ways to address the concerns raised in this article, some have suggested alternative approaches that might help to address the overrepresentation of ATSI people in criminal processes. These alternative approaches include the decolonisation of justice (Blagg 2016) and, relatedly, justice reinvestment (Brown et al. 2016 courts, to reduce magistrates' reliance on 'perceptual shorthands' and resultant negative sentencing discrimination and to provide a more meaningful sentencing process for ATSI people (Jeffries and Bond 2013: 111;Marchetti and Daly 2016). These strategies may offer some pathways towards change. However, we must attend to the specific justice needs of ATSI women, the most marginalised group in Australia, who are both under-protected and over-policed and are being incarcerated in ever increasing numbers under a DVO system originally introduced to protect women from DFV. September 2012, the definition was much narrower including wilful injury or damage, intimidation or harassment, indecent behaviour without consent or a threat to commit any of these behaviours: Domestic and Family Violence Protection Act 1989 (Qld) s 11. 3 See Penalties and Sentences Act 1992 (Qld) s 9(1). 4 Domestic and Family Violence Protection Act 2012 (Qld) s 177. This provision was amended on 22 October 2015 with the maximum penalty increasing to 240 penalty units or five years' imprisonment if previously convicted of a contravention offence within the preceding five years or otherwise 120 penalty units or three years' imprisonment. Our data does not cover court decisions made post 22 October 2015. 5 Where defendants had more than one order within the year, the latest order outcome was counted. 6 Note that studies conducted by Jeffries (2011, 2012) depending on different jurisdictions and courts studied on the role of indigeneity on sentencing have made different conclusions. In another study, they examined whether domestic versus non-domestic violence offending is sentenced differently. Their analysis of sentencing NSW cases heard between 2009 and 2012 concluded that, in domestic violence cases, both older and younger Indigenous males were treated more harshly in sentencing, net of other sentencing factors, than others. Their very tentative finding in this study was that the influence of Indigenous status on sentencing is conditioned by other offender attributes (Bond and Jeffries 2014: 867).