Measuring police performance in domestic and family violence

Foreword | As police are often the first point of contact in domestic and family violence incidents they are in a unique position to respond to, intervene in, and be proactive about, preventing family violence. With indications that the volume of recorded violence has increased and the complexity of family violence matters has increased police workloads, police face many challenges in responding in an effective and timely manner to reported incidents. The recent release of a national policing strategy to combat family violence recognises these challenges and reflects a commitment to improve responses and the sharing of information by police and other partner agencies. To know whether there are changes in practice and in outcomes over time, there is a need to identify ways to measure police performance. This paper reports on a number of performance indicators that could be used by police to measure their effectiveness in responding to and reducing family violence. Importantly, the indicators are based on working with police from one jurisdiction to identify what is most relevant and accessible in operational data, as the aim is to have indicators that are both practical and useful. Monitoring these indicators over time will allow police to identify where they are making an impact and how their intervention strategies may be improved.

Judy PuttGeneral Manager, Research

Australian governments, both federal and state/territory, have in recent years recognised the enormous impact domestic and family violence has on victims and the community. In 2005 the Australian Bureau of Statistics (ABS) estimated that approximately 160,000 women in Australia aged 18 years and over had experienced violence with a current partner since the age of 15 years (ABS 2006: 34). This represents about two percent of Australian women aged 18 years and over. It was also estimated that over one million women had experienced violence from a previous partner since the age of 15 (ABS 2006: 35), representing 15 percent of women aged 18 years and over. Ten percent of men and ten percent of women aged 18 years and over in Australia had experienced physical abuse before the age of 15. The vast majority of these incidents for both males and females had been perpetrated by parents or step-parents, or by another relative (ABS 2006: 41). In May 2008 the National Council to Reduce Violence Against Women and Children was established by the Australian Government in recognition of the continuing need to address the issue of violence against women.

Australian community-based police work at the coalface of domestic and family violence incidents, although it is acknowledged that most incidents are not reported to police (for example Holder & Caruana 2006: 57). This means that police are in a unique position to not only obtain and provide information about reported domestic and family violence, but to have an impact on such violence through their response, intervention and prevention strategies. As police are often the first point of contact, it is vital that the police response is timely, appropriate and effective. This is important for ensuring victims receive the care and support that they need, minimising further harm and encouraging a willingness to report to and cooperate with police. It is also important for ensuring offenders are dealt with appropriately and quickly, minimising their capacity to further re-offend and delivering and reinforcing the messages that offending behaviour will not be tolerated and that police will act swiftly.

A report by the Crime and Misconduct Commission (CMC) highlighted increases in police workload relating to family violence due to more time spent on cases involving a domestic partnership, and expanded legislative definitions of domestic violence to include a range of relationships (CMC 2005). New legislation to introduce a broader range of behaviours and relationships in domestic and family violence is also currently under consideration in Victoria and the ACT. Partly in response to an overarching national drive and partly due to a growing appreciation of the time and resources involved in policing family violence, the need for measures to assess effectiveness of police performance in family violence has been identified as a key objective by police commissioners nationally. The Australasian policing strategy for preventing and reducing family violence was launched by police commissioners across Australia in November 2008. The strategy outlines priorities for action to improve information and intelligence sharing between police, as well as between partner agencies. Identifying good practice and achieving consistency in how police respond to family violence are key components of the strategy.

Benefits of measuring police performance

Like all areas of government spending, police services are accountable to the public and are increasingly called upon to show evidence of their performance. As summarised by Moore & Braga (2004: 3): 'Policing managers, too, have needed a measure of police performance - partly to meet external demands for accountability, and partly to establish a form of accountability inside their organisations that could focus attention on achieving valuable results rather than simply reliably executing established policies and procedures'. In the UK, police have an overarching framework which guides police performance (Home Office 2008) and in the US much work has been undertaken in the area of measuring police performance over the past thirty years (see O'Neill, Needle & Galvin 1980; Moore 2002).

Why is a focus on police performance in the area of family violence important? Members of the public may ask: How well are police responding to family violence in my neighbourhood? How are they dealing with offenders? Is it worth reporting to police? For policy makers, the question is: What policies could be changed to assist and improve police performance in family violence, and how? Finally, police need to know: Are our family violence policing strategies effective in reducing victimisation and protecting victims? How do we know? Is our response to victims appropriate? How can we encourage victims to come forward and report? To answer these questions, specific performance measures need to be identified and monitored.

Policing practices in family violence

The policing of family violence has undergone a fundamental shift in recent years. Since the beginning of the 1980s, when police did not consider policing of family and domestic violence to be part of their job (Newbold & Cross 2008), to recent times when legislative changes and pro-arrest mandates across many western countries including Australia (NSW Ombudsman 2006), New Zealand (Newbold & Cross 2008), the US (Friday et al. 2006) and the UK (Applegate 2006) have forced changes in practice in the policing of family violence incidents.

In Victoria, the Code of practice for the investigation of family violence was introduced in 2004. It focuses on improved collection of evidence, investigation and laying charges where appropriate. The introduction of this code appears to have increased the number of charges laid in family violence incidents and the proportion of family violence incidents in which charges were laid (See Figure 1). Just under half of all offences recorded in family violence incidents in 2006-07 in Victoria were assault.

In the context of the entire criminal justice system, there is little doubt that a collaborative, coordinated, interagency approach to addressing family violence is best practice. While police may be the first point of contact in many incidents they cannot combat and prevent family violence alone. While not always running smoothly (see Giacomazzi & Smithey 2001), the experience from North America is that 'a police unit in and of itself cannot break the cycle of violence; it needs to be able to work in union with the rest of the criminal justice system to have its greatest impact' (Friday et al. 2006: 63).

A good example, which draws heavily on programs developed overseas, is the ACT Family Violence Intervention Program (FVIP) which is a proactive, multi-agency approach to family violence in the ACT. The ACT has a pro-arrest, pro-charge policy on domestic and family violence and such cases are fast tracked through the courts. The FVIP encourages sharing of information between agencies about family violence, resulting in a publication in 2006 with data and trends across a number of agencies in the ACT (Holder & Caruana 2006). The report showed substantial increases in the arrest rate for domestic violence and the volume of family violence defendants coming before the courts (Holder & Caruana 2006: 2-3). In addition, the percentage of successful convictions increased from 76 percent in 2003-04 to 85 percent in 2007-08 (ACT DPP annual reports).

While the responsibility for primary prevention of family violence does not and cannot rest with police alone, there is increasing pressure on police to monitor and improve their performance in this area.

However, little attention has so far been given to how police performance might be measured or how their effectiveness in responding to and impacting on family violence might be gauged. In 2006, the NSW Ombudsman produced a comprehensive report looking at improving police practice in domestic violence. This report noted that there were currently no standards by which to effectively measure performance in policing of family violence (NSW Ombudsman 2006: i).

Figure 1: Family violence incidents attended by Victoria Police 2006-07

Family violence incidents attended by Victoria Police 2006-07

Source: Victoria Police (2007)

Figure 2: Performance measurement in policing family violence

 Performance measurement in policing family violence

Developing performance measures

In early 2008, the Australian Institute of Criminology (AIC) was engaged by Australian Capital Territory Policing (ACT Policing) to conduct a review of their family violence data collection processes. This involved identifying information needs of FVIP stakeholders, reviewing how family violence data are recorded in police systems and identifying the most efficient methods of data extraction. The project also identified family violence police workload indicators and examined how police performance might best be measured. The performance measures in Figure 2 were generated as part of that project and were informed in two ways; through Australian and international literature and through consultation with ACT Policing.

Reduction in repeat victimisation

The primary reason to improve the response of police to family violence is to decrease the frequency and severity of violence experienced by victims and to ensure their safety. A reduction in the number of victims who continue to experience family violence after police are called to an incident may be an indicator that previous intervention has had an impact in reducing offending. It may also reflect proactive behaviour on the part of victims to remove the risk of further violence (for example, through protection orders that police may have helped initiate). This indicator should be tracked over time to monitor whether the proportions of domestic and family violence victims who continue to be victimised are decreasing. This would require a unique victim identifier to be available to police statisticians. The specific indicator would be calls for assistance to police by or on behalf of a person who has previously required police assistance relating to family violence. While overall levels of victimisation in the community would be of interest, many family violence incidents are not reported to police, and police performance should only be assessed for those incidents of which they are aware.

Reduction in repeat attendances

Many domestic and family violence incidents are likely to be perpetrated by the same offender and/or experienced by the same victim and/or occur at the same address. That is, it is likely that a substantial proportion of domestic and family violence incidents involve repeat offenders, victims and locations. This suggests that placing an intervention focus on repeat offenders and locations may have a greater impact in reducing incidents than treating all incidents in the same manner. There is currently little information available about the number of repeat domestic and family violence incidents police attend. A reduction in repeat attendances to the same address could be one indicator that previous police intervention has had a positive impact. This performance measure would need to be tracked over time to monitor whether the proportions of households with repeat attendances are decreasing, and could be measured using existing statistical data held by police.

Reduction in repeat offending

This measure aims to record the number of calls to police for assistance in relation to a specific alleged offender. A reduction in the number of offenders who continue to perpetrate family violence after police are called to incidents would be an indicator that previous police intervention has had an impact on further offending behaviour. This performance measure should be monitored over time to assess whether the number of repeat offenders is decreasing and, again, existing police data could be used. It is particularly important that this indicator be tracked over time, as desisting from further offending may not occur until formal procedures have had time to affect behaviour. Buzawa and colleagues (2000: 18) described an evaluation of a US proactive domestic violence program where 'recidivism rates remained quite high especially within the first month after arrest, but before the formal court processing'. Tracking offenders over time would allow for formal processes to run their course and for repeat offending to be properly measured.

Accurate identification and recording of incidents

Information about family violence incidents that police attend is only as good as the information entered by operational police into their recording systems. If information is entered incorrectly, or relevant information is left out, this will affect the quality and accuracy of the data that can be extracted from police systems about incidents. Accurate recording of domestic violence incidents is highlighted as essential to good practice policing in the UK (Home Office 2006).

While almost all jurisdictions in Australia have a provision for recording whether an incident was domestic/family violence related (an FV/DV flag), the accuracy of this flag may be questionable due to continually changing legislative definitions of such incidents and the need for operational police to keep up to date with these changes. The review of ACT Policing data found that how operational police interpreted and recorded an incident was not always consistent with the legislation. However, as an example of good practice, one outcome of the review was to recommend that ACT Police monitor the accuracy of the family violence flag over time to determine whether correct identification of incidents as family violence improves. Although this practice may be resource intensive due to the manual examination of incident descriptions, the measure could be achieved by a periodic review of the assignment of the flag.

Increased number of offenders charged and successfully prosecuted

Given the pro-arrest and pro-charge policy operating in many jurisdictions for family violence offences, it would be expected that the numbers of offenders being charged and prosecuted should increase over time. In Victoria the numbers and proportions of charges being laid has substantially increased. In the ACT, Holder & Caruana (2006) showed that the number of family violence matters being prosecuted also increased markedly. Between 1998-99 and 2005-06 there was a 464% increase in the family violence matters prosecuted in the ACT by the Office of the Director of Public Prosecutions. Between 2003-04 and 2007-08 successful prosecutions of defendants on family violence charges increased from 76 percent to 85 percent.

Increasing prosecution of family violence through the courts has two benefits. The first is the opportunity for the offender to be involved in treatment and increased supervision, and the second is the deterrent effect, both for the individual (specific deterrence) and for others involved in, or on the path to family violence (general deterrence).

Although successful prosecution is not solely the responsibility of police, they are responsible for the initial decision to charge, the types of charges laid and the effective collection of evidence to support the charges. An increase in successful prosecutions would be an indicator (albeit a proxy measure) of police effectiveness in relation to charging and evidence collection. Data to provide this measure may be held already by the police, or may be sought from the Office of the Director of Public Prosecutions depending on the jurisdiction.

Increased arrests and laying of charges for DVO and violence order breaches

The objective of protection orders made by the court is to prevent further violence and/or abuse: to protect people from future assaults, threats of violence, property damage, stalking, harassment and offensive behaviour. They protect the applicant (the person who applies for the order) by ordering the respondent (the person against whom the order is made) from doing certain things. In family violence cases the applicant will generally apply for a Domestic Violence Order, or DVO, (in the ACT), an Apprehended Violence Order (in NSW) or similar order elsewhere, which can be granted for up to two years.

A breach of a violence order is an offence that renders the offender liable to arrest by police. In 2003-04 in the ACT, around 12 percent of all charges laid by police for family violence incidents involved a breach of DVO (Taylor 2006). The ABS 2005 Personal safety survey (ABS 2006) found that, where a violence order had been issued against a previous partner since the age of 15 years, 42% of women had continued to experience violence. Improving community confidence that police will act when a breach occurs is crucial for ensuring the safety of victims, encouraging reporting of breaches and for reinforcing to offenders that breaching a violence order is a serious offence for which they will be arrested. One means of measuring this is to identify the number of incidents involving breaches of DVOs and monitoring over time how many of these incidents result in the offender being arrested and charged. The data needed to measure this outcome are already held by police.

It should be noted that police can only act on breaches reported to them. Often the offender will have left the scene before the police arrive and it may be one person's word against another that a breach occurred. The wider issue of the unknown volume of breaches which are not reported to police is also an area that needs attention. It would be hoped that, as a byproduct of increasing the arrest and charging rates of offenders who breach violence orders, confidence in, and rates of reporting to police may also increase.

Ensuring police are adequately informed about previous attendance and criminal histories before arriving at an incident

For police to respond appropriately to family violence incidents, it is essential that the attending officers are aware of any previous police attendance at that address for family violence, and whether the offender may be violent. While police communications often provide a summary of criminal history and previous attendance to patrols dispatched to an incident, it is unclear how much information is provided, and how much of that information is received and understood by the patrol officers before they arrive at the incident. For example, a patrol that takes 20 minutes to reach an incident is likely to be in possession of more information prior to arrival than a patrol that takes five minute to arrive. Officers in the latter scenario who have not had time to digest the information provided will be less prepared for what they might find and how they should respond than officers in the former scenario.

In terms of measuring performance, regular monitoring of the adequacy of information provided to patrol officers by communications areas is necessary. Communications areas need to identify and provide patrol officers with key information quickly and accurately. Improving performance in this area will increase the actual and perceived safety of officers entering family violence situations.

Improved willingness of victims to call and/or cooperate with police, and increased victim satisfaction with police response

It is also necessary to consider how victims of crime interact with police and how this interaction might be improved. Family violence is heavily under-reported to police and there are many reasons that victims may be unwilling to report. If someone else reports the violence or even if the victim reports, the victim may not wish the offender to be arrested and may not cooperate further with police. Two key challenges facing police are how to increase the willingness of victims to report family violence and how to encourage their cooperation in a pro-arrest and pro-charge environment.

As far as can be determined, state and territory police in Australia do not routinely collect data on victim satisfaction with police response specifically in family violence incidents (although data on community satisfaction with police are published annually in the Report on government services (SCRGSP 2008)). This is an area which has been acknowledged as being important for future research in policing family violence (Hanmer & Griffiths 2000).

The police response to victims who have experienced family violence will be crucial to whether they report any further victimisation (Smeenk & Malsch 2005). Anecdotal evidence suggests that victims who perceive that the police response was poor (did not believe victim, did not act, responded inappropriately or did not locate or arrest offender etc.) are less likely to report again. Evidence from a UK study found victims who were subsequently victimised were more likely to contact police again if their previous experience had occurred in a manner which had been consistent with their preferences (Hickman 2003). Buzawa and colleagues (1999) argued that other measures of the impact of a domestic violence intervention, including levels of victim satisfaction, may tell more than recidivism rates alone. Holder (2001) presents examples from a number of studies which show that victim satisfaction levels increased after reforms in policing family violence. While reforms currently underway in Australia make baseline measures of victim satisfaction difficult to obtain, it is nevertheless an important indicator to track over time. Annual surveys of those who have been in contact with police over the previous year would contribute to addressing the current knowledge gap.


Policing of family violence is an important issue. As Australian society has become less tolerant of domestic and family violence, the policing of it has become the subject of greater interest. Recent changes in legislation across the states and territories have led to a widening of the definition of family violence, in terms of relationships and behaviours. While this is a positive step in that it situates family violence squarely as a priority issue, it also presents pressures and challenges for police. Police are, in many cases, the front line response to family violence and this places them in the unique position of being an obvious source for accurate data on reported family violence incidents. It also places them in the spotlight in terms of how they respond to family violence incidents and whether their intervention and proactive strategies are effective in responding to and reducing family violence.

To meet these challenges, performance measures that will inform police, policy makers and the community about the effectiveness of policing strategies are needed. It is only through establishing baseline measures and monitoring performance over time that gaps in performance can be assessed and improvements to future policing strategies identified. For internal purposes alone, police need to know how much, and in what way family violence contributes to their workload - given resource constraints, the impact on and response by police needs to be managed as effectively and efficiently as possible. This requires measurement and continual evaluation of key performance indicators about family violence. The Australasian policing strategy for preventing and reducing family violence is an acknowledgement by all police commissioners that performance in this area needs to be improved and is an important first step in moving this agenda forward.

A number of useful performance measures have been identified in this paper. Performance, however, is a multi-faceted concept as it covers a range of areas, not just one. For this reason the performance measures identified in this paper are not intended to be exhaustive and should be considered in conjunction with each other, rather than viewed in isolation. Identifying a range of performance measures and the need to monitor them will ensure that the policing of family violence is viewed as more than simply responding to late night disturbance complaints.


All URLs were correct at 20 October 2008

About the authors

Dr Natalie Taylor is a senior research analyst and manager of the Justice and Crime Analysis team at the AIC.

Kiah Rollings is a research analyst at the AIC.