Executive summary

	The focus of our audit
	The NSW Police Force (NSWPF) spends nearly 80 per cent of its resources on its people, many of whom work shifts. Getting rosters right can improve law enforcement, the efficient use of public resources and officer welfare, health and safety.
	There have been some common trends in the types of rosters used for policing around the world. Traditionally, the predominant roster was three eight-hour shifts a day, seven days a week with similar numbers of staff rostered to work each shift. These rosters gave little or no consideration to when crimes were actually occurring.
	In the 1990s many police forces began to modify police rosters to better match resources to demand and to improve the working conditions of officers.
	The NSW Police Force implemented a flexible roster model service-wide in 1995. Under this model, shift length can vary between six and 12 hours and shifts can start and finish at different times. Flexible rosters can also provide more days off duty. This can help officers overcome the stresses of policing if they rest on their days off.
	Flexible rosters are intended to best match staffing numbers to workloads. Commands, in consultation with officers, can choose shift lengths and shift starting times that maximise capacity when crime is occurring or when police need to be visible.
	In this audit we wanted to find out whether the current police rosters, including practices around the use of 12-hour shifts, support efficient and effective policing.
	Specifically we examined: whether rosters provide for the flexible deployment of staff whether rosters meet the needs of the community and victims of crime what impact rosters have had on officers and the NSW Police Force.
	The audit focuses on general duties police. That is, police who wear uniforms, work out of local police stations and represent the largest group of officers that routinely work around the clock.
	Audit opinion
	Today, few elements of the 1995 flexible roster model are in practice in the NSW Police Force. Current rosters are neither consistent with the flexible roster model nor represent best practice. As a consequence, both the community and police are missing out on some of the benefits that the Force saw this model providing.
	Instead, general duties police work predominantly 12-hour shifts. These shifts have common starting times and finish times. And in most commands, the majority of these shifts are worked in blocks of four, followed by at least four days off.
	Some commands deploy more officers on busier nights, but generally a similar number of officers are rostered on duty regardless of demand. Although a number of factors affect performance, the rigid work pattern increases the risk the Force may be unable to attend calls quickly, provide good customer service and protect the well being of its officers.
	Overall, the Force is not meeting its target for responding to urgent calls for assistance. Today, one in four complaints is about customer service. Victims of crime report issues including difficulties in contacting an officer and frustration at having to wait for an officer to return to duty to find out about their case. The timeliness of investigations is also at risk when follow-up is delayed. Seventy per cent of commanders surveyed indicated that rosters did not support continuity of investigations or victims’ needs.
	We are also concerned that the current roster pattern may not be in the best interest of the health and safety of officers. Sick and injury leave is higher in NSW than other states and the Force is yet to assess the extent to which rosters may contribute to this.
	The problem is not entirely the use of 12-hour shifts. Twelve-hour shifts are not inherently bad. Rather, it is the practice of block rostering that seems to be the main problem.
	The NSW Police Force and the Police Association of NSW are both publicly committed to the flexible roster model. According to the Association, flexible rosters can provide greater coverage during periods of peak demand resulting in better services to citizens. As this is the case, the Force should implement flexible rosters as envisaged in 1995.
	The Force also needs to do more to protect its general duties officers from fatigue in line with contemporary practice for managing shiftwork.
	Key findings
What are flexible rosters and do police use them?	A flexible roster is one that varies shifts patterns and lengths to optimise policing, efficient resource use and the welfare, health and safety of officers.
	In 1995, the NSWPF replaced the traditional eight-hour roster with flexible rosters. Varied shift patterns were meant to meet fluctuations in demand and reduce the impact of shiftwork on officers.
	We found aspects of flexible rosters in practice. Some commands vary the start time and length of some shifts. Several commands roster extra officers on busier evenings and overlap shifts to increase capacity.

A more flexible roster – see page 18

But there is much inflexibility. Most shifts are 12 hours long, start around the same time and deploy a similar number of officers day and night.

A less flexible roster – see page 19

	Typically, commands roster officers for blocks of four days on duty (two day and two night shifts) followed by four or more days off duty.
	Until recently, the NSWPF has not done much to promote flexibility and commands report that it is difficult to change entrenched roster practices. The NSWPF has now developed a roster training package and is introducing a tool to assess how commands manage and deploy officers.
	The NSWPF is also reviewing its roster policy and practice to achieve more flexible rosters. The NSWPF plans to complete this review by the end of 2007.
Are rosters good for the community and victims of crime?	Effective rosters should help officers respond rapidly to calls, manage cases and support victims. The rosters used by NSWPF do not always do this.
	The number of urgent calls for assistance has increased by 33 per cent over the last four years to more than 110,000 calls. In 2006-07, police attended 80 per cent of these calls within 12 minutes 25 seconds, failing to meet its target response time of 80 per cent within 10 minutes.
	Many variables affect response times. Rosters are one of these variables yet commands do not review response time to identify if the roster contributes to delays. Nor do they use response time performance as a test of whether the roster is efficient and effective.
	The current block roster gives officers only two day-time shifts every eight days. These day shifts are the most effective time for following up victims, witnesses, and other police who work standard (non-shift) hours. The limited opportunity for follow-up under the current roster may affect the timeliness of investigation and victim support.
	The NSWPF has procedures in place for case management and victim support and has set minimum standards for the review of open cases. Even so, compliance with these standards is variable. Over 25 per cent of complaints to police were about customer service in 2006-07, compared to 15 per cent in 2002-03. Common complaints were that it was difficult to contact the officer handling their case and that calls were not returned.
	Of the commanders we surveyed: § 70 per cent said current rosters did not support the continuity of investigation. This rose to 86 per cent for commanders whose officers predominantly worked 12-hour block rosters § 66 per cent said current rosters did not enable officers to manage their caseloads effectively § 70 per cent said the rosters did not facilitate victim support.
Are rosters good for officers and the NSWPF?	A good roster meets the operational needs of the business, minimises the risks of shiftwork to staff and allows employees to balance their work and private lives.
	The 12-hour block roster is popular with general duties police. It provides more time off and by concentrating working hours allows officers to have longer breaks from work. Officers report that these rest periods allow them to recover from the stresses associated with policing.
	Yet research indicates that rosters can negatively affect employee health and welfare and there are specific risks associated with shift work that need to be managed.
	NSWPF officers take more sick and injury leave than officers in forces that have different roster arrangements. The NSWPF is currently analysing how its rosters affect officer welfare.
	The NSWPF does not have a fatigue management policy, but it has parameters that limit the number of 12-hour shifts worked to three shifts in a week. These parameters are not adhered to. Officers typically work four 12-hour shifts in four days and we found some officers working six 12-hour shifts in seven days.
	Fatigue may be reduced by the long rest periods between shifts but this also makes it easier for officers to have a second job. If this happens, then the opportunity for rest may be lost. One in ten police officers has approval to work a second job but others may work without approval on their rest days.
	The NSWPF does not provide guidance to commands assessing requests for secondary employment in terms of fatigue management. There is no limit on the maximum hours that can be worked. We found one case where an officer had approval to work for up to 30 hours a week on his rest days.
	The NSWPF is currently reviewing the impact of rosters on officer health and welfare and is developing a fatigue management policy.
	Recommendations
Achieve compliance with current requirements	We recommend that the NSWPF ensure commands comply with the current roster parameters regarding the number of 12-hour shifts that officers can work (page 35).
Review current requirements	We recommend that the NSWPF review the rostering parameters and, by July 2008, develop best practice principles that are consistent with contemporary practices for managing the risks to the health and welfare of officers arising from shiftwork (pages 22 and 35).
Incorporate best roster practices	We recommend that the NSWPF require commands, as part of the best practice roster principles, to:
	§ review patterns of response times, use different shift lengths and stagger start times to better meet the demand for police (pages 26 and 27) § provide officers sufficient opportunity to follow-up investigations, witnesses and victims at customer-friendly times (page 29) § review the roster’s ability to support the needs of the community and victims by monitoring such indicators as complaints, community satisfaction and the timeliness of case follow-up (page 29) § monitor the effect of rosters on such organisational indicators as retention, injury and sick leave, overtime, shift penalties and annual leave (page 40).
Better manage fatigue	We recommend that the NSWPF by July 2008 introduce a fatigue management policy that meets best practice principles and includes:
	§ training for commanders and officers in how to identify and manage fatigue (page 37) § a minimum amount of rest taken before resuming duty (page 37).
Help commands to improve rosters	§ We recommend that the NSWPF by July 2008: § provide commands with a methodology to assess local rosters against best practice principles and identify gaps (page 22) § support commands in their negotiations with officers to change rosters to meet best practice principles (page 22).
	We recommend that by July 2009, the NSWPF have rosters in place that are consistent with best practice principles (page 22).
	Response from the NSW Police Force
	I refer to your letter dated 2 November 2007 regarding the final Performance Audit on Police rostering.
	The New South Wales Police Force welcomes the submission of this report into police rostering. As an organisation, we accept that improvements can be made in the way we manage our human resource work hours through rostering.
	The organisation acknowledges the audit opinion, in terms of the Local Area Commands surveyed, that current rosters are not consistent with the current agreed parameters for flexible rostering.
	There are 4 main areas of interest to the audit report:
1. Local Management of Rosters	A local roster is designed to ensure that police are deployed both proactively and responsively. Within a local roster, a Command will task police to activities such as the following: § First response § Investigative work § Station duties, such as custody or exhibit management § Court attendances § Training
	Thus the number of police rostered on a given shift is based on the requirements for many policing activities, including responding to calls for assistance. It is an oversimplification to correlate the number required to be deployed for responding to calls for assistance with the total number of police required to be rostered on shift, and the patterns for this rostering.
	Each Local Area Command is required to enter into a First Response Policing Agreement (FRPA) with their staff.  The number of police assigned to first response on a given rostered shift will be based on historical trends of calls for assistance, which may then be adjusted seasonally or based in intelligence.  The first response deployment also enables capacity for a high visibility police presence.  Based on current crime activity officers are proactively tasked with crime intervention, crime detection and crime prevention.
	Tasking of those officers within the FRPA is conducted at weekly Command crime reviews, however each Command also reviews crime occurring each 24 hours and there is the ability to respond to this crime immediately by changing the pro-active taskings.
	Best roster principles, when applied, ensure a FRPA allows for the rostering of police for shifts of various durations with staggered finishing times to meet changing demands for service.  Additionally, Commanders often roster extra crews, particularly at night, to ensure the safety and occupational health and wellbeing of staff.
	The New South Wales Police Force is currently providing training for roster officers to assist them in the population of rosters and adherence to the agreed parameters for flexible rostering.
2. Flexible rostering parameters	Much has recently been said of police working 12-hour shifts.  I note the Auditor General’s report does not criticise 12-hour shifts per se, but recommends greater flexibility within FRPAs.  I agree, however I also support the notion that 12-hour shifts underpin FRPAs.  It appears that 12-hour shifts, with corresponding regular breaks for officers, contribute to the attractiveness of working within the New South Wales Police Force.  Attending to outstanding work in a timely manner is the basis of good case management and victim support within a Command, regardless of the roster cycle.
	The agreed parameters for flexible rostering allow, for those officers who work mainly 12-hour shifts, a maximum of three 12-hour shifts in any seven day period, with four 12-hour shifts during the final week of any six week roster period.  I note the findings of the Audit that Commands appear to have moved to a pattern of four 12-hour shifts.  To manage any identified fatigue risk associated with this, I propose to firstly identify the extent of this risk and then implement appropriate risk control measures, which may be the acceptance of the recommendation to ensure compliance with the current flexible rostering parameters, or another series of controls which are to be identified as a part of our fatigue management project described below.
3. Investigation Case Management and Customer Service	The audit report stated that current rosters limit the time officers have to follow-up cases and victims, suggesting improved rostering practices would provide for more effective case management and victim support. It is acknowledged that the effective rostering of police will enhance their ability to conduct investigations and service customers.  However, it is my view that the enhancement and reinforcement of processes should ensure that case information is readily accessible by co-workers who can assist customers should the contact officer be unavailable for any reason (such as court duty, annual leave, training, strike force or other duties), thus reducing the reliance on a single contact officer.
	The merits of maintaining an effectively managed case management system, already established within COPS (Computerised Operational Policing System), will impact greatly on the quality of investigations and customer service.  It is imperative that the workload of each officer is closely monitored by supervisors by utilising the system within COPS.  Furthermore, the role of ‘assisting officer’, which can be assigned within the case management system, should be spread across shifts to allow a team based approach for the completion of tasks.  Finally, the recording of investigative milestones and related documentation must be current and accessible to allow case continuation in the absence of the officer in charge.
4. Fatigue Management	The establishment of the Safety Command in 2006 has given NSW Police the capability to identify and respond through contemporary methodologies to health and safety issues including fatigue management.
	A project to address the specific issues related to fatigue management has commenced.  The basis of the New South Wales Police Fatigue Project is founded upon contemporary Safety risk management principles.
	The New South Wales Police Fatigue Project will encompass the following stages:
	a) Stage 1 (expected completion February 2008)
	Examination of the current flexible rostering agreement and other related policies (eg secondary employment) in terms of contemporary fatigue management principles
	Examination of a sample of current Police rosters in terms of contemporary fatigue management principles
	Completion of risk assessment of the sample rosters in relation to fatigue
	Introduction of basic risk control strategies including a communication strategy to increase staff awareness of fatigue and personal risk assessments
	Preparation for phase 2 including a stakeholder forum in the near future.
	The main goal of stage 1 is to identify the current safety risks to the New South Wales Police Force employees from current rostering practices; and identify attainable risk control measures to counter this in the short term.
	b) Stage 2 (expected completion July 2008)
	Creation of a fatigue management policy
	Inclusion of contemporary fatigue practices into future flexible rostering agreements
	Organisational education in fatigue management practices, including how to identify and manage fatigue
	Creation of an organisational toolkit to be utilised by managers
	It should be noted that a sample of 16 Local Area Command rosters are being utilised in stage 1.  This sample will include the four Local Area Commands that were chosen for the Audit Office report.
	The Centre for Sleep Research, University of Adelaide has been engaged to undertake the scientific analysis of rosters in stage 1.  Professor Drew Dawson, who is internationally recognised for his contribution in the area of fatigue management, will oversee this analysis.
	The New South Wales Police Force has gained the support of the Police Association of New South Wales (PANSW) on this project, and will work in collaboration with the PANSW during the entirety of the project.
	I anticipate that this Auditor General’s report will greatly assist the NSW Police Force in its endeavours to improve rostering practices through the work of its recently established Safety Command whilst also ensuring improved service delivery to the people of New South Wales.
	(Signed) A P Scipione APM Commissioner of Police Date:  21 November 2007