Even when supported by evidence of effectiveness, interventions can sometimes fail due to focusing too heavily on replicating a successful intervention and not on the factors needed to implement that approach (Ekblom 2010; Knutsson & Clarke 2006). This has been termed ‘implementation failure’ (see Ekblom 2010, 2002; Tilley 2009). Implementation failure refers to problems with translating an idea into practice, resulting in either no interventions being delivered, the desired interventions not being delivered, or the interventions being implemented so poorly that the intended results are not produced (Tilley 2009). The risk of implementation failure increases with the number and complexity of the interventions being delivered, the number of agencies that are involved, the presence of separate lines of accountability, personnel changes, the absence of support and the volatility and changeability of the context in which the intervention is being delivered (Knutsson & Clarke 2006; Tilley 2009). Implementation failure has proven to be a significant issue impacting upon the effectiveness of crime prevention programs and initiatives, both in Australia and overseas (Homel 2009a; Sutton, Cherney & White 2008; Tilley 2009, 2005).
A common theme across many of the strategies reviewed as part of this project was the use of situational approaches to crime prevention. These are supported by a strong theoretical framework and a range of established techniques. The situational crime prevention literature also provides strong guidance around how to implement effective crime prevention strategies. It is also a common approach adopted with some success by local government in New South Wales and other jurisdictions (Morgan & Homel 2011; NSW DAGJ 2012). In addition, the majority of strategies involved community-based organisations, such as local government, often working in partnership with police, private businesses or other community-based organisations. Therefore, the interventions identified are suitable for local government implementation and adaptation.
As with other practitioners, local government are frequently encouraged to adhere to a problem-solving process in the development, implementation and review of crime prevention initiatives (Cherney 2006; Henderson 2002; Homel 2009a; Laycock 2005; Tilley & Laycock 2002). Various models have been developed to guide this process (Cherney 2006; Ekblom 2010; Tilley 2009). All of these models involve some combination of problem analysis, strategy selection, implementation, partnership working and review (Cherney 2006). Problem solving involves a systematic analysis of current and emerging crime problems, their causes and risk factors. Once these problems are identified and understood, an appropriate response can then be identified and developed based on evidence regarding the effectiveness of different approaches and a consideration of the circumstances to which it will be adapted. The process then involves identifying the key parties that need to participate and mobilising them for action to implement the response. The response is then subject to regular review and evaluation, and feedback on implementation and the effectiveness of the strategy is used to inform improvements (either to the current or future strategies).
In order to develop factsheets and handbooks to assist local government to select and then implement suitable interventions to address local crime priorities (based upon this problem-solving process), the AIC collected information on the characteristics of successful strategies and on the requirements for implementation. This review aims to contribute to the knowledge base on the requirements for successful implementation, as well as providing an assessment on the effectiveness of different intervention types. This will help to build the capacity of practitioners to select an appropriate intervention, adapt that intervention to the local context and minimise the problems commonly associated with implementation.
Since the strategies reviewed for this research often involved a suite of interventions delivered in combination and relied on similar approaches to crime prevention, it is possible to identify common requirements for successful implementation. There may be unique factors relevant to the different intervention types, but there are many requirements that were common across all interventions. It was therefore possible to identify a number of common factors among those strategies that were successfully implemented and reviewed as part of the current project, which fall into the following categories:
- a thorough and systematic analysis of a range of data sources to identify significant crime problems and to understand their causes and risk factors;
- community engagement and consultation in the development of the strategy (including but not limited to residents, business operators, local service providers etc);
- strong interagency partnerships, led by a driver responsible for maintaining project momentum and implementation; and
- availability of appropriate expertise, technology and resources.
The application of these generic principles to the preferred intervention types is described in detail in the relevant factsheets and handbooks, along with a number of examples to demonstrate how these principles should be applied in practice.
Analysis of the problem
An important dimension of the evidence-based approach to crime prevention is the selection of a response based on an understanding of the problem being addressed and its underlying causes (Clarke & Eck 2003). Determining how, where and when to intervene requires an understanding of the nature of the crime problem being addressed (Hirschfield 2005).
Unless there is a clear definition of the crime problems to be addressed, with evidence for them, it will not be possible to work out what to introduce or how to implement it (Laycock 2005: 572).
The effectiveness of a particular intervention is highly dependent upon how appropriate that intervention is for the crime problem being addressed (Hirschfield 2005). Experience from problem-oriented policing projects has shown that crime prevention projects may fail because the targeted problem was inaccurately identified or inadequately analysed, leading to the selection of a response that does not address the actual problem or its causes (Scott 2006). There needs to be a clear rationale for the proposed intervention, based on an understanding of the problem being addressed and its causes (Hough & Tilley 1998). This requires the systematic identification and analysis of crime problems (Laycock 2005). Comprehensive analysis of the crime and contextual factors will help to inform an understanding of the problem and of environmental factors that may help facilitate or inhibit the implementation of the chosen solution and influence its overall effectiveness (Hirschfield 2005). In some instances, the most obvious and appropriate solution will emerge through the analysis of the problem (Clarke & Eck 2003; Goldstein & Scott 2001). In others, a range of potential solutions may be identified. Further analysis will help determine whether the proposed interventions are appropriate for the problem being addressed.
The majority of the interventions reviewed as part of this research involved situational approaches to crime prevention. A key feature of situational crime prevention is that it works most effectively when it is targeted at a specific crime problem in a specific context. However, this principle applies irrespective of the problem being addressed or the approach to prevention. For example, a common element of effective comprehensive responses to gang-related crime in the United States is that there was an extensive research process undertaken before developing the scheme in order to understand the key issues associated with gang-related violence in the local community (Braga et al. 2001; Skogan et al. 2008; Tita et al. 2010). This involved wide-ranging consultations with police, community leaders and service providers. The result of this process was the identification of a discrete and manageable problem. Similarly, in the case of residential burglary, effective strategies were those that had been able to identify high-risk households (based on local crime data and previous victimisation), factors that contributed to this high risk and access points for burglary offenders, so as to inform the development and implementation of appropriate responses (Bowers, Johnson & Hirschfield 2004; Ekblom, Law & Sutton 1996; Forrester, Chatterton & Pease 1988; Forrester et al. 1990). Recorded crime data from police were frequently used to identify the locations of recent burglaries, the extent of repeat victimisation, common access points for offenders, the types of premises that were targeted and the types of property that were stolen. Strategies were then developed to address these specific risk factors. Conversely, a meta-evaluation of burglary prevention strategies across three countries, undertaken by Grove (2011), concluded that many unsuccessful strategies had failed to accurately diagnose the problem and tailor the interventions adopted accordingly.
Overall, the studies reviewed as part this project showed that effective strategies were targeted at relatively small or well-defined areas with high rates of offending. The areas that were targeted had been identified as having a specific crime problem or high crime rate, either through local crime statistics and/or on the basis of concerns raised by the community. This requires access to information to identify crime hotspots and targets, as well as information about the characteristics of offences (such as when offences are most common) to inform a more targeted approach. Various sources of data were used in the development of the strategies reviewed for this project, including:
- recorded crime data, which is a valuable source of information about crime trends and temporal patterns (ie by month, day of week and time of day), the types of locations that are targeted, people or households at risk of becoming a victim, offenders apprehended by police and the types of property that is stolen; (mention limitations—ie data is only that which is reported).
- hotspot maps, which provide a visual representation of the locations within each neighbourhood with the highest concentration of recorded offences, as well as the specific locations where these offences occur;
- surveys or interviews with victims of crime provided useful information about the level and nature of victimisation, risk factors and the types of measures that were already in place (and may not be working);
- surveys of the wider community were used to assess the degree of concern among residents about the prevalence of offending in their neighbourhood, possible explanations for this offending, high-risk locations, perceptions of safety and the level of support for different types of prevention strategies;
- surveys or interviews with offenders, while less common due to the challenges associated with their implementation, have been used to develop a better understanding of the motivation of offenders and the reasons they target specific locations or victims and the techniques used to conduct the offences; and
- consultation with relevant local stakeholders (potentially as part of a working group or project committee, see below) was often used to seek useful information regarding other organisations’ experience and understanding of offending in the local community, the possible causes of offending and information on local initiatives trialled in the past.
Effective strategies frequently relied upon a combination of data sources to provide a more in-depth understanding of the problem being addressed. This is in recognition of the fact that relying on individual data sources provides a limited understanding of a problem. For example, police-recorded crime data is limited to offences that are reported to police (which can vary considerably between different offence types) and restricted to the type of information that may be recorded and made available to assist with planning interventions. As well as requiring high-quality data, this also requires someone with the ability to analyse and accurately interpret the data that has been collected.
Further, effective strategies often had a process in place to monitor the impact of the strategy as it was being implemented and throughout the life of the project. Regular data collection and analysis enabled project managers to monitor progress and identify new issues as they emerge. Given the concerns regarding the impact of situational measures over the long term, it is vital that key indicators continue to be monitored so that interventions can be modified in response to changing crime rates and patterns.
Community engagement, involvement and commitment
Community engagement is a key feature of effective crime prevention (Camina 2004; Mistry 2007). Community engagement, involvement and commitment to the project were also important factors in the successful implementation of the strategies described in this report. Experience has shown that interventions involving the community are more likely to be more effective when members of the community are enthusiastic and supportive of the initiative. Experience has also shown that it is important to begin working with the community as early as possible, involving them in both the design and implementation of a strategy.
It is therefore necessary to establish appropriate consultation mechanisms at the commencement of the project to seek input from members of the community, business operators and local service providers (and others) into the development of strategies that are likely to require their involvement (or at least compliance) and that will impact upon them. This was particularly true of projects targeting young people, such as mentoring or diversionary schemes.
Stakeholder commitment is also important and requires the establishment of appropriate partnership arrangements. There are a range of stakeholders who make a valuable contribution to the development, implementation and evaluation of crime prevention initiatives. In Australia, contemporary crime prevention has generally embraced the value of interagency partnerships, collaborative policy development and program delivery, in recognition that the causes of crime are wide ranging, complex and frequently require a coordinated response (Homel 2009). Partnerships are an effective mechanism for the delivery of integrated solutions, comprising closely linked and coordinated interventions that can achieve shared outcomes. However, experience has shown that establishing and maintaining effective partnerships can be challenging (Morgan 2011; Morgan & Homel 2011) and that while partnerships offer numerous benefits, they also present significant operational challenges (Knutsson & Clarke 2006; Laycock & Tilley 1995).
The degree of partnership working between two agencies can vary across a continuum ranging from networking through to collaboration, depending on the need, purpose and willingness of two agencies to work together (Morgan 2011; VicHealth 2006). Research has identified several important features of effective crime prevention partnerships:
- a clear mission and agreement on the objectives of the partnership;
- good knowledge and understanding of one another’s roles, responsibilities and motivation for being involved in the partnership;
- a high level of trust between partner agencies;
- similar organisational perspectives, objectives, performance indicators and cultures;
- members who work well together, respect one another and are committed to ensuring the partnership succeeds;
- strong leadership at senior levels to exercise some level of influence and ‘champions’ for the project working at the local level;
- the capacity of agency representatives to commit resources to enable partnerships to function and to address barriers to implementation as they arise;
- clear lines of accountability within the partnership and its parent agencies through performance management processes;
- division between strategic management and the management of operational and implementation issues, with clear lines of communication and accountability;
- partnership structures that are relatively small, businesslike, with a clear process for making decisions and a focus on problem solving;
- adequately resourced, including ensuring that staff have enough time away from agency core business to provide input to the partnership;
- data sharing policies and protocols; and
- continuity in partner representation and participation, and documentation of processes and decision making (Gilling 2005; Rosenbaum 2002; Scott 2006).
These characteristics can act as a checklist against which policymakers and practitioners are able to assess whether the necessary conditions exist for a proposed solution involving multi-agency collaboration. The review of projects as part of this research has also highlighted a number of practical lessons for developing and maintaining effective stakeholder partnerships. This is particularly important given that the majority of strategies involve multiple components and related interventions and will therefore require local government to work with other stakeholders.
Where projects involve multiple interventions requiring input from a range of different stakeholders, a committee with representatives from the various parties should be established early to oversee the development, implementation and ongoing review of the project. This will help to ensure strong partnerships between key agencies and provides a process whereby all parties can be held accountable for delivering different aspects of the program. It is also a forum for sharing ideas regarding potential solutions to new problems as they arise. This was a common feature of many of the projects reviewed. Ideally, this group should be led by a dedicated project coordinator. Membership stability among agency representatives involved in governance or management committees is also important (Anderson & Tresidder 2008; Morgan & Homel 2011).
Similarly, multicomponent strategies should be supported by a comprehensive implementation plan that describes the key stages in project delivery and the interrelationships between different but complementary interventions. Experience has shown that having a clear strategy is important in ensuring that a project can be implemented as it was originally intended. It can also help to outline relevant roles and responsibilities. Progress against the plan can then be documented and monitored by the stakeholder committee to ensure that key information and lessons are passed on to new staff.
There are occasions where it can be difficult to engage or maintain the involvement of key stakeholders. This was evident in a number of the studies reviewed and when it occurred, often had a detrimental impact on the implementation of the project. There is a range of different reasons that this might occur. Stakeholders may be unable to take action to prevent crime (eg due to a lack of resources), they may perceive it to be the responsibility of someone else, they might think the cost outweighs any potential benefits and in some extreme cases, may actually benefit from the problem that is being addressed (Clarke & Eck 2003). Scott and Goldstein (2005) have identified a number of ways police can shift ownership to other agencies as part of problem-oriented policing projects, some of which may be appropriate to local government. Overcoming strong resistance may require legislation mandating action to be taken or withdrawing (or threatening to withdraw) particular services or support for the other agency. Generating positive media coverage about a project (or negative media coverage about a problem) may foster support for a project or shame another party into taking action (Scott 2006; Scott & Goldstein 2005). Other strategies can include:
- pushing for the establishment of a new organisation that can assume responsibility for a particular activity;
- approaching alternative organisations who might be able to help address the problem;
- making more targeted and/or formal requests for assistance where informal approaches have been unsuccessful, such as producing evidence that might help convince other agencies of the need for their involvement or by approaching senior management (Scott & Goldstein 2005).
The suitability of these responses will depend on who the partner is, what role they perform in the project, how important that role is and what the potential negative consequences of some of these options could be.
Expertise, technology and resources
Projects often require specific expertise. For example, as part of an access control and awareness campaign to reduce residential burglary, suitably qualified personnel such as police are required to undertake security audits. Further, in order to provide assistance with the installation of security devices, qualified personnel may need to be engaged to improve the physical security of doors, windows and fencing, as well as to install entry systems and security lighting. Similarly, strategies that involve making improvements to street lighting require professionals such as lighting engineers. They will have an understanding of lighting design and technology and be aware of issues that may reduce the effectiveness of an intervention (eg the need to ensure lighting is bright enough and to prevent obstructions that will limit its impact). Further, trained personnel will ensure that lighting equipment is hardwearing, vandalism-proof or resistant and (where necessary) is maintained over time. This will help to ensure that the appropriate technology is employed.
Recruitment strategies should aim to attract appropriate project staff with the skills, experience and knowledge relevant to the position. Strategies also need to be developed to ensure the longer term stability and sustainability of projects, particularly in the event of staff turnover.
It is also important that projects are adequately resourced. Interventions should be selected on the basis that there are adequate available resources to enable them to be implemented in accordance with their design. This includes resources to support the purchase and installation of equipment, employing qualified personnel, preparing relevant materials and promoting the project and (in the case of CPTED) changes to buildings and/or landscaping. It may be possible to adapt interventions to meet funding constraints, but this needs to take into consideration the potential impact upon the effectiveness of the intervention.
Selecting a response
In this review, a number of intervention types have been identified that are supported by evidence of effectiveness and the requirements for their implementation. However, there are a number of factors that need to be considered in deciding what to do when selecting a response to an identified crime problem. In some instances, the most obvious and appropriate solution will emerge through the analysis of the problem (Brown & Scott 2007; Clarke & Eck 2003; Goldstein & Scott 2001). In others, a range of potential solutions may be identified (even within the one type of intervention, such as access control).
There are a number of important factors that need to be considered in determining which of the preferred interventions is the most appropriate and most likely to be effective for a local government area with an identified crime problem. Brown and Scott (2007) have identified 10 questions to ask about planned interventions as part of the planning stage:
- What is the change mechanism?
- What evidence is there that the intervention has worked before?
- How difficult will it be to implement the intervention?
- Does the intervention rely on external partners’ actions?
- Are regulatory or high-level policy changes required to implement the intervention?
- How will the intervention interact with other interventions being implemented in the same area/with the same group?
- What will be the stakeholders’ reactions to the intervention?
- Will any negative consequences accrue from the intervention?
- How long will it take for the intervention to show results?
- Can the impact be measured?
The answers to each question will help inform the process of deciding what to do to address an identified crime problem, and which of the intervention types should be implemented by local government with the support of the NSW CPD.