Final Evaluation Report

98.The Evidence Base

Overview

Following on from the evaluation of Priority 5 of the Second Action Plan, which aimed to continue building the evidence base, this section of the report considers the existing evidence base for violence against women.

The First Action Plan aimed to ‘progress building the evidence base’ to better understand trends in sexual assault and domestic and family violence. The First Action Plan also sought to put in place the data collection and research mechanisms required to develop a rigorous evidence base for violence against women and their children. To achieve this ANROWS was established, the National Data Collection and Reporting Framework commenced and the government committed to conducting the PSS and the NCAS on a four-yearly rolling basis.

Evaluation of the First Action Plan found that while important gains had been made to improve the evidence base, the need to improve the consistency of data collection and application remained. For this reason, the Second Action Plan included ’continuing to build the evidence base’ as a national priority.

Now that we are nearing the completion of the Second Action Plan which prioritises ‘building the evidence base’ for violence against women and their children, it is timely to evaluate what the baseline is for measuring progress in reducing violence against women and their children.

There are three key pieces of evidence that are commonly used to measure the reduction in violence against women: the NCAS, the PSS and the NATSISS. With the exception of the NCAS which was published in 2014 the surveys are somewhat dated. Nonetheless, analysis of the historical data from these surveys enables development of a baseline for measuring progress in reducing violence against women and their children.

In addition to these surveys, data has also been analysed from the Experimental Family and Domestic Violence Statistics from the ABS Recorded Crime – Victims Catalogue (4510.0) which was consolidated under the Second Action Plan and provides data about the incidence of assaults, homicides and sexual assault recorded in each jurisdiction. Publicly available data for family protection orders and police incidents has also been included where this is available.

Feedback from service providers via the survey and consultations has revealed that there is significant concern around the lack of performance indicators for the National Plan as well as concern about the consistency and completeness of the data used to develop the evidence base. For this reason, we have examined the primary sources of data in the evidence base for the Second Action Plan and presented them under four themes: prevalence of violence; awareness and understanding of domestic violence; behavioural patterns and attitudes to violence; and awareness of mechanisms of response to violence and preparedness to intervene.

These themes in the current data sets loosely align with the high level indicators of change under the National Plan which are:

• 
  reduced prevalence of domestic violence and sexual assault;
  
• 
  increased proportion of women who feel safe in their communities;
  
• 
  reduced deaths related to domestic violence and sexual assault; and
  
• 
  reduced proportion of children exposed to their mother’s or carer’s experience of domestic violence.
  

The discussion below also considers the limitations of the current data in measuring change against these high level indicators for the National Plan, particularly in relation to the cohorts of women at high risk of violence.

3.1 Prevalence of Violence

There is no accepted or common method for estimating the prevalence or incidence of violence against women and their children. Violence against women and their children is also not limited to ‘domestic violence’ and involves both physical and sexual violence from people in non-intimate relationships with the victim. In this way, two of the high level indicators of the National Plan, which focus on ‘domestic violence’, measure only a subset of the violence against women. For this reason, we have evaluated the data available for measuring the prevalence of violence against women and their children without a particular focus on intimate partner violence.

In relation to the National Plan, the prevalence of violence against women and their children is usually measured using the results of the PSS. The PSS collects information about the nature and extent of violence experienced by men and women since the age of 15, including men’s and women’s experience of current and previous partner violence. The NATSISS also includes an indicator for measuring the prevalence of violence such as the proportion of respondents who have experienced violence in the 12 months prior to being surveyed.

Additionally information from the Stop the Violence Initiative and recent submissions to the Senate have been used to indicate the prevalence of violence in particular cohorts of women.

Crime data is also provided in this section of the report to indicate the incidence of violence and present the data which is most often used by the media to indicate the incidence of violence but also, in most instances, is incomparable and varies from State to State in its method of collection. Together this data presents a holistic picture of the data available to indicate the incidence of violence and the baseline against which to measure future changes in the incidence of violence against women and their children under the National Plan.

Prevalence of violence against women

The most recent PSS indicated that of the sample of women interviewed, 41 per cent of women had experienced violence since the age of 15, 34 per cent had experienced physical violence and 19 per cent had experienced sexual violence. It is estimated that more than half of the women who had experienced violence from a current partner had children in their care when the violence occurred, 31 per cent of these stated that children had seen or heard the violence.⁴³ Of the women who had experienced violence while living with their previous partner 48 per cent stated that children had seen or heard the violence⁴⁴. This is the only indicator of violence against children and is closely aligned with the fourth high level indicator for the National Plan: Reduced proportion of children exposed to their mother’s or carer’s experience of domestic violence.

The next PSS is due to be undertaken in 2016. As such, this evaluation of the Second Action Plan refers to the PSS undertaken in 2012, which was also considered in the evaluation of the First Action Plan. A similar data limitation is observed for the NATSISS which is due for release in 2016. The most recently published NATSISS dates back to 2008 and was included in the evaluation of the First Action Plan. While these two data sources are dated and not directly comparable, it is helpful to consider these indicators for the prevalence of violence together to show the varying sources of information used to indicate, and therefore measure, the prevalence of violence. We have presented a selection of key indicators of the prevalence of violence against women and their children in Figure 3.1 below.

Figure 3.1: The prevalence of violence against women in the 12 months prior to the
2012 PSS

Source: KPMG analysis of the 2012 PSS

Figure 3.2: The prevalence of violence against Aboriginal and Torres Strait Islander women in the 12 months prior to the NATSISS

Source: KPMG 2016, analysis of the NATSISS 2002 and 2008

99.Aboriginal and Torres Strait Islander Women

Figure 3.2 shows that according to the NATSISS Aboriginal and Torres Strait Islander women were much more likely to experience violence in the 12 months prior to being surveyed than women surveyed under the PSS: 23 per cent of Aboriginal and Torres Strait Islander women had experienced violence in the 12 months prior to the NATSISS as opposed to 5.3 per cent for the population consulted under the PSS.

The PSS does not interview women in discrete Aboriginal and Torres Strait Islander communities or women living in very remote areas. As such, the NATSISS is likely to provide a better indication of the violence against women in Aboriginal and Torres Strait Islander communities, albeit in a different year to the PSS. However it is not possible to make a direct comparison of the prevalence of violence in Aboriginal and Torres Strait Islander and non-Aboriginal and Torres Strait Islander populations using the NATSISS and the PSS because the methodologies for these two surveys are substantially different. For example, the NATSISS is specific to asking about physical violence in the 12 months prior to the survey, not lifetime experience or ‘experience since the age of 15’ as the PSS does.

In 2002, 24 per cent of Aboriginal and Torres Strait Islander women had experienced physical violence in the 12 months prior to the NATSISS survey. This indicates that little progress was made in reducing violence against Aboriginal and Torres Strait Islander women between 2002 and 2008 given that the same question in the 2008 NATSISS indicated a 1 per cent reduction in the proportion of Aboriginal and Torres Strait Islander women who had experienced physical violence in the 12 months prior to the survey (refer Figure 3.2 above).

100.Women with a disability

According to the PSS, women with a disability were also more likely to have experienced violence in the 12 months prior to the survey compared with the general population. The PSS estimated that in 2012, 7.6 per cent of women with a disability or long-term health condition had experienced violence in the 12 months prior. This is 2.2 per cent higher than the prevalence reported in the general population. Given that the PSS excludes women in institutional settings or those who require a facilitated interview, it is likely that this presents an underestimate of the violence against women with a disability. In 2013, the Women with Disabilities Australia led the Stop the Violence initiative which surveyed women with a disability who had presented at a disability service in the 12 months prior, to assess the level of violence and sexual abuse experienced by these women. This survey found that 22-25 per cent of women had experienced violence or sexual abuse in the 12 months prior to being surveyed. These results vary significantly from a similar analysis in the PSS and present a marked issue of consistency and reliability in measuring the prevalence of violence for women with a disability.

101.Women who identify as lesbian, gay, bisexual and transgender

A recent submission to the Senate Inquiry into Domestic violence and gender inequality from Flinders University, presented research regarding the prevalence of violence against women who identified as lesbian, gay, bisexual and transgender.⁴⁵ The survey results, which included both male and female respondents, indicated that 58 per cent of survey respondents had experienced some form of abuse from a partner or family members. Of those who had experienced abuse, 45 per cent had experienced physical abuse, 28 per cent had experienced sexual abuse and 34 per cent had experienced abuse specific to their lesbian, gay, bisexual and transgender identity. An earlier survey undertaken in Victoria in 2008 found that that just under a third of survey respondents had been involved in a same-sex relationship where they were subject to abuse by their partner and lesbian women were more likely than gay men to report having been in an abusive same-sex relationship (41 per cent and 28 per cent respectively).⁴⁶ Other survey research which focused on heterosexist violence and harassment indicates that approximately 22.5 per cent of women who identify as lesbian, gay, bisexual and transgender are victims of verbal abuse, 5.9 per cent are victims of physical abuse and 3.1 per cent are victims of sexual assault.⁴⁷ These studies are not directly comparable so it is not possible to make a clear determination on whether there has been a regression in attitudes toward people who identify as lesbian, gay, bisexual and transgender. However, the research does indicate that people, and specifically women, remain at high risk of physical and sexual violence.

102.

103.Victims of Family and Domestic Violence, Experimental Data

The data presented in Victims of Family and Domestic Violence Experimental Data 2014, which was released in 2015, draws from police crime recording systems in each State and is based on a Family and Domestic Violence (FDV) flag as recorded by police officers in 2014. The FDV flag is one of a number of indicators used by police to determine whether an offence is family and domestic violence-related. The experimental data set released by the ABS notes that not all family and domestic violence-related offences ends up being flagged on police recording systems and consequently the number of family and domestic violence-related offences may be understated.

The table below shows the number of victims of family and domestic violence-related assault from selected states as recorded by police in 2014. Data for Queensland and Victoria was not available because the interpretation and implementation of the National Crime Recording Standard Rule 2 for assault incidents varied from the standard for Victoria and Queensland.⁴⁸

Table 3.1: Victims of family and domestic violence related assault in each jurisdiction in 2014.

States and Territories	Victims of family and domestic violence related assault in 2014	Victims of family and domestic violence related assault per 1,000 people^49
Australian Capital Territory	615	2
New South Wales	28,780	4
Northern Territory	4,287	17
South Australia	5,691	3
Western Australia	14,603	6

Source: KPMG analysis of ABS data from Cat. 4510.0 Recorded Crime, supplementary release to Experimental Family and Domestic Violence Statistics.

The police data reaffirms that women are overwhelmingly the victims of family and domestic violence. The ABS experimental data set indicates that of the domestic and family violence-related assaults in each state and territory, the following ratios for female and male victims:

• 
  in the ACT three times as many female victims (465) as male victims (145);
  
• 
  in NSW twice as many female victims (19,488) as male victims (9,261) in NSW;
  
• 
  in the NT four times as many female victims (3,482) as male victims (807);
  
• 
  in SA four times as many female victims (4,534) as male victims (1,157); and
  
• 
  in WA three times as many female victims (10,648) as male victims (3,860).
  

Figure 3.3 below shows the variation in the incidence of family and domestic related assault per 1,000 people in each jurisdiction. This graph clearly shows that on a per capita basis the NT has a much higher incidence of family and domestic violence related assault than other jurisdictions with 17 incidents per 1,000 people compared with 4 per 1,000 people in NSW and 6 per 1,000 people in WA. A large proportion of Aboriginal and Torres Strait Islander women reported experiencing violence in the 12 months prior to the NATSISS than the general population surveyed for the PSS. The NT has a higher Aboriginal and Torres Strait Islander population than other jurisdictions which could explain the significantly higher incidents of family and domestic violence related assault.
Figure 3.3: Victims of family and domestic violence related assault in each jurisdiction in 2014

Source: KPMG analysis of ABS data, 2016

The numbers below show the number of victims of family and domestic violence-related sexual assault from selected states as recorded by police. In 2014, the number of victims of family and domestic violence-related sexual assault offences recorded by police was:

Table 3.2: Victims of family and domestic violence-related sexual assault from selected states as recorded by police in each jurisdiction in 2014

States and Territories	Victims of family and domestic violence related sexual assault in 2014	Victims of family and domestic violence related sexual assault per 1,000 people
Australian Capital Territory	36	0.09
New South Wales	800	0.11
Northern Territory	88	0.36
South Australia	249	0.15
Queensland	36	0.01
Victoria	1,321	0.23
Western Australia	544	0.21

Source: KPMG analysis of ABS data from Cat. 4510.0 Recorded Crime, supplementary release to Experimental Family and Domestic Violence Statistics.

Figure 3.4 below shows the variation in the incidence of family and domestic related sexual assault per 1,000 people in each jurisdiction. As with the results for assault in the figure above, this graph clearly shows that on a per capita basis the NT has a much higher incidence of family and domestic violence related sexual assault than other jurisdictions with 3.6 incidents per 10,000 people compared with 2.1 per 10,000 people in WA and 1 incident per 10,000 people in NSW.

Figure 3.4: Victims of family and domestic violence related sexual assault in each jurisdiction in 2014.

Source: ABS data from Cat. 4510.0 Recorded Crime, supplementary release to Experimental Family and Domestic Violence Statistics.

104.Longitudinal studies

The only reliable longitudinal studies for the prevalence of violence include: reported experiences of violence in the 12 months prior to the PSS; reported police incidents (data is publicly available for two jurisdictions) and court protection orders (to measures in jurisdictional police incidents data, (data publicly available for three jurisdictions).

Figure 3.6 below shows the changes over time in the prevalence of violence against women in the 12 months prior to being interviewed for the PSS. There appears to have been little change for sexual assault incidents. For physical assault, there was a 1.3 per cent reduction between 1996 and 2012 but this was not a statistically significant change in the proportion of women subjected to violence between 2005 and 2012. As such, this indicates that the prevalence of violence has changed very little and remains a cause for concern.

Figure 3.6: Changes over time in the prevalence of violence against women in the 12 months prior to the PSS

Source: KPMG analysis of the PSS

105.Family Violence Orders

Another source of longitudinal analysis for the prevalence of violence is the publicly available data for family violence orders.

Family violence orders are called different things in different states:

• 
  Protection Orders (QLD);
  
• 
  Apprehended Domestic Violence Order (NSW);
  
• 
  Intervention Orders (VIC & SA);
  
• 
  Violence Restraining Orders (WA);
  
• 
  Family Violence Order (TAS); and
  
• 
  Domestic Violence Order (ACT & NT).
  

Police in their respective jurisdictions may assist an aggrieved person in their application for an order. Dependent on the jurisdiction, police may have the responsibility to make applications for protection orders should they witness domestic violence. Usually this happens with the support of the aggrieved person, but where there is opposition by the aggrieved person, the police, in certain states, are obliged to pursue the application of a protection order. In some states, for example SA, the issue of a court protection order (or intervention order) mandates that the perpetrator must attend a behavioural change program. Consequently, court protection orders can function not only as a form of protection for the victim but also as an avenue for perpetrators to get assistance and support.

The number of family violence orders is recorded by the Magistrates Court in each respective jurisdiction. Data for QLD, VIC and TAS is publicly available but data for other jurisdictions was not readily accessible. The table below shows the steady increase in court protection orders for QLD and VIC between 2009-10 and 2014-15. TAS, on the other hand, has experienced a decrease in the number of court protection orders issued between 2009-10 and 2014-15. This reduction has however included a moderate increase from 486 finalised family violence orders in 2012-13 to 565 in 2013-14.

Table 3.3: Total number of family violence orders finalised by year and state

	2009-10	2010-11	2011-12	2012-13	2013-14	2014-15
QLD	17,641	17,307	17,274	18,709	20,148	21,721
VIC	22,991	24,925	27,668	29,214	29,978	N/A
TAS	693	590	602	486	565	532

Source: Magistrates Court for each respective state

The figure below shows that over the last six years VIC has had a consistently higher rate of family violence orders per 1,000 people than QLD and TAS. However, these figures do not take into consideration the number of Police Family Violence Orders (PFVOs) issued in Tasmania. Through the Family Violence Act 2004, TAS became the first jurisdiction to introduce police-issued family violence orders with Tasmania Police having the power to issue an immediate PFVO 24/7 without court involvement.

Figure 3.7: Number of family violence orders per 1,000 people for QLD, VIC and TAS over the period 2009-10 to 2014-15

Source: Annual reports for magistrates course for each state.

106.Police Incidents

The graph below shows the number of reported police incidents relating to family violence per 1,000 people for QLD and VIC over the last five years. Data for the other states and territories was not readily accessible and it is possible that QLD and VIC have different thresholds or regulations for reporting violence against women and their children. Regardless of the comparability of the data it is clear that police activity in response to family violence has increased over the last five years.

Figure 3.8: Number of police incidents per 1,000 people for QLD, VIC and TAS over the period 2009-10 to 2014-15

Source: QLD Police Service data and VIC Family Incident Reports 2009-10 to 2013-14

It not possible to say whether the increase in the number of court protection orders and police incidents relating to family violence in QLD and VIC indicates an increase in violence against women and their children or an increased awareness of the issue of violence against women and their children and therefore greater reporting of the violence as a result of a reduction in the stigma attached. In any case, the rise in court protection orders in two states, coupled with the increase in police incidents in QLD and the reportedly high incidents of ‘domestic and family violence’ matters attended by police, indicates that the vigour in reducing violence against women and their children through behavioural change and community awareness may be having an impact.

Prevalence of violence – issues with the evidence base

The biggest issue with measuring the incidence or prevalence of violence against women and their children is variation on what constitutes violence and the way in which it is measured. The PSS, the NATSISS and police data are all frequently used to indicate the prevalence of violence. Additionally, more recent sources are now beginning to be used to indicate the prevalence of violence for particular cohorts of women, for example data from the Stop the Violence initiative to indicate the prevalence of violence for women with a disability and other surveys of women who identify as LBGTI. The National Plan has had emphasis on high level population level indicators. However, gaps in evidence for diverse experiences have been recognised as a major concern. ANROWS’ Diversity Data Project and Our Watch’s Working with Culturally and Linguistically Diverse (CALD) Communities are among the key initiatives addressing these gaps. 

Going forward, it will also be important to consider violence in non-intimate relationships as part of the measurement of progress against National Outcomes in order to appropriately account for the needs of all women, and also all children, who are at risk of violence.

107.The PSS

The PSS provides a rich data set of interpersonal violence in Australia, based on a significant sample size (17,050 persons across 41,350 dwellings). Extensive quality control measures are used to ensure the robustness of the data but despite this there are limitations to the way that the data can be used, for example “examination of sub-populations can be difficult because the estimates quickly become too unreliable for general use”.⁵⁰ Many stakeholders indicated that despite the measures taken to ensure the robustness of sampling for the PSS this survey lacks an indicative representation of children, women with a disability, Aboriginal and Torres Strait Islander women and women from culturally and linguistically diverse backgrounds. Feedback from consultations indicated that there was significant concern about the exclusion of the following cohorts of women, who are often at high risk of violence:

• 
  people living in institutional settings from the survey;
  
• 
  people living in very remote areas;
  
• 
  Aboriginal and Torres Strait Islander discrete communities; and
  
• 
  residents who require the assistance of another person to communicate with the interviewer for example people with a native language other than English⁵¹.
  

The focus of the PSS on violence since the age of 15 also limits the representation of children in the primary piece of evidence for measuring the National Plan’s progress in reducing violence. Given that the National Plan targets women and their children, it is important to have a more complete understanding of the violence experienced and witnessed by children in the evidence base for the National Plan.

KPMG acknowledges that there are significant administrative and financial challenges to changing the sampling of the PSS to include greater representation of groups at high risk of violence, however measurement of the prevalence of violence and future evaluation of the progress made to reduce violence against women should account for these cohorts of women in other ways for example targeting data collection and survey.

108.The NATSISS

Although there have been major changes with each iteration of the NATSISS (in 2008, 2002 and 1994), there is still a concern that the NATSISS 2014 does not reflect the substantial change in Aboriginal affairs over the past twenty years. One critic has called the content of the NATSISS “problematically, overwhelmingly, focussed on problems… Aboriginal people are more than a social problem, and after 20 years NATSISS should be reflecting this.” ⁵² There are concerns around the lack of focus on achievement, aspiration or the benefits of being an Aboriginal person.

Workshop consultations also indicated that the current data and research used to indicate the prevalence of violence in Aboriginal and Torres Strait Islander communities was not representative of the actual issue. This perceived lack of co-design for data collection and research presents an opportunity for greater community input which could be explored with future data collection and any new survey design or amendments to the NATSISS.

The National Plan aims to measure success for National Outcome 3, Aboriginal and Torres Strait Islander communities are strengthened, through a reduction in the proportion of Aboriginal and Torres Strait Islander women who consider that family violence, assault and sexual assault are problems for their communities; and an increase in the proportion of Aboriginal and Torres Strait Islander women who are able to have their say within their communities. However, data is not currently collected or made available at a national level for these measures to be assessed.

It could also be beneficial to establish a long-term evaluation of progress against the second high level indicator of the National Plan: the proportion of women who feel safe in their communities. This is particularly relevant for Aboriginal and Torres Strait Islander communities because of the relatively higher risk of violence against women and also because action 9 of the Second Action Plan was to build community safety for women and children in Aboriginal and Torres Strait Islander communities. While the PSS might be a good proxy for measuring changes in community safety, there is no equivalent for Aboriginal and Torres Strait Islander communities, particularly when the PSS does not survey in discrete Aboriginal and Torres Strait Islander communities.

109.Women with a Disability

There is a designated hotline for people with a disability to report abuse. The National Disability Abuse and Neglect Hotline is funded by the Department of Social Services and delivered by WorkFocus Australia and provides referral to community services, accommodation and respite care services in response to reported cases of abuse and neglect. Types of abuse and neglect can include physical, sexual, psychological, legal and civil abuse, restraint and restrictive practices, or financial abuse. It can also include the withholding of care and support which exposes an individual to harm. Consultations with National Plan Partners indicate that it is not clear where the data collected for the National Disability Abuse and Neglect Hotline is collected or reported. It is possible that this data could be used to indicate the prevalence of violence against women with a disability. However, this would likely only collect data pertaining to people with the capacity to phone the hotline and possibly not those women who are being abused by their carer.

In 2013, the Stop the Violence initiative collected the most robust data to date regarding the prevalence of violence against women with a disability. However, the WWDA notes that the estimates provided are likely to still be an underestimate of the prevalence of violence for women with a disability because the survey was constrained to only consider women who had presented at a service in the 12 months prior (therefore excluding women in institutional settings), and only considered domestic violence and sexual abuse. Other forms of violence such as forced sterilisation and restraint and seclusion practices were excluded from the survey.

In November 2015, the Senate Community Affairs Reference Committee released a report to finalise their inquiry into Violence, abuse and neglect against people with disability in institutional and residential settings, including the gender and age related dimensions, and the particular situation of Aboriginal and Torres Strait Islander people with disability, and culturally and linguistically diverse people with disability. The report recommended that the Australian Government consider amending the National Plan to include institutional and disability accommodation settings, to ensure that people with disability are afforded the full range of rights protections that are available to people without disability⁵³. While the National Plan does not expressly refer to women and children in institutional and disability accommodation settings, they are included in the scope of the National Plan with respect to sexual violence. There are also actions under the Third Action Plan that specifically target women with disability.

110.Children

The National Plan aims to reduce violence against women and their children. However, some stakeholders indicated that the National Plan should look more broadly at the impact of violence on children and not just in regard to the mother-child relationship. If a child witnesses violence against a women in any circumstance it can have a lasting negative effect.

The only indicator of violence against children is the PSS measure of the number of women who have children in their care at the time of violence. A number of stakeholders, including.

Families Australia, indicated that this is a poor and incomplete measure for the reduction of violence against children generally and the impact of domestic violence on children specifically. Further to this the PSS and the NCAS only survey children older than 15 so the attitudes, understandings and experiences of children younger than 15 are not collected, analysed and used to support policy. Child protection orders and child abuse statistics are reported and do provide an indication of the prevalence of violence against children. Whilst it is not possible to discern exactly what proportion of child abuse and child protection orders relate to family violence (because there is no family related violence flag in this data) it could be beneficial to consider how this data could be used to inform policy and provide a strong evidence base for service development and delivery to children effected by violence.

111.National Crime data

The National Crime Recording Standard (NCRS) was designed to enable the recording of a crime in a comparable manner across all jurisdictions. However, Victoria and QLD do not adhere to certain rules in the NCRS which means that their data is not comparable to data from other jurisdictions. In particular, rule 2 of the NCRS requires that police officers record a report at 'face value' (that is, prior to further investigation) on their crime recording systems. In Victoria, the business processes around entering information on LEAP (Victoria Police's Crime Recording system) require some investigation of whether or not a criminal offence has occurred prior to recording the incident. Therefore, Victoria Police are not adhering to rule 2 and their data are not comparable with other jurisdictions.

In QLD, the legislative requirements mean that family and domestic violence-related assaults are considered a civil matter in many cases where they would be considered criminal in other jurisdictions. These would still be recorded as information reports on QPRIME (the QLD police crime recording system). However, unless a criminal offence is recorded, the victims are not included in the scope of the Recorded Crime Victims statistics. This means that due to the legislative environment and the structure of their crime recording system in QLD the number of criminal assaults recorded is much lower than, and not comparable to, data recorded in other jurisdictions.

Although a robust sampling method was used for the NATSISS, but exclusions in the in-scope population resulted in an under-coverage of an Aboriginal and Torres Strait Islander population in the survey. The coverage exclusions were applied to manage enumeration costs and resulted in an estimated under-coverage of approximately 6 per cent of Aboriginal and Torres Strait Islander persons in Australia.^54,55

In summary, key areas of concern are:

• 
  the poor representation of children and high risk groups of women in the primary data sources for example women living with a disability, Aboriginal and Torres Strait Islander women and women from culturally and linguistically diverse backgrounds;
  
• 
  the lack of clarity around the use of the evidence base in measuring the effectiveness of policy including the progress against National Plan outcomes; and
  
• 
  the lack of comparability in crime data between jurisdictions.