Introduction

It would be true to suggest that advancing crime prevention knowledge not only involves understanding whether specific interventions ‘work’ to reduce crime, but also how crime prevention is delivered (Homel and Brown 2017). Since the flagship Morgan Report of 1991, it has often been asserted that the development of crime prevention strategies and their delivery requires partnership approaches or inter-agency working (Tierney 2001; Tilley 2002; Berry et al. 2009; Rosenbaum and Schuck 2012; Homel and Brown 2017). Partnership approaches to crime prevention have been developed globally, and in England and Wales, the 1998 Crime and Disorder Act made such inter-agency working a statutory requirement (Homel and Brown 2017). While a body of literature has outlined the challenges with partnership working, this paper primarily focuses on (1) the development of inter-agency approaches to the reduction in violence across England and Wales and (2) how prepared areas are for the Serious Violence Duty (SVD) as legislated for in the recent Police, Crime, Sentencing and Courts Act (2022).

While the challenges of partnership approaches to crime prevention are well documented (Tierney 2001; Tilley 2002; Berry et al. 2009; Homel and Brown 2017), it has been recognised that partnerships are required for the successful delivery of public health approaches to violence reduction (Craston et al. 2020). Rosenbaum and Schuck (2012) point to the efficacy of community partnerships in reducing gang violence in the USA, and the successful approach to violence reduction in Scotland (see Scottish Violence Reduction Unit, nd) has been held as an exemplar of what can be achieved through public health approaches. While much violence reduction delivery in England and Wales has been piecemeal and completed in addition to the substantial requirements of the 1998 Crime and Disorder Act, it was the Serious Violence Strategy (Home Office 2018) that laid the foundations for much of the current activity across England and Wales. The strategy not only set out an approach for tackling violence through interventions (such as early interventions, community responses and law enforcement), but also presented an opportunity to rethink how violence is tackled (see Youth Endowment Fund 2021) and build a body of knowledge around what works in this space. The Strategy emphasises a need for further partnership working (Home Office 2018, p. 9) and made a commitment to fund interventions that would be implemented through Violence Reduction Units (VRU) in the eighteen police areas worst affected by violence.Footnote 1The expectation was that the VRU’s would tackle ‘the root causes’ of serious violence through inter-agency working and public health focused interventions. Public health approaches consider violence to be a treatable disease and require that strategies consider the role of the individual, their relationships, community and wider societal factors in both the production and methods to prevent violence (Krug et al. 2002). As Grimshaw and Ford (2018, p. 15) identify, the risk factors for violence are ‘crosscutting’ and with this in mind, any comprehensive approach to violence reduction would require input from a number of agencies.

In addition to the developments outlined above, future inter-agency working will largely be driven by the SVD (see Home Office 2019, 2021). The duty, as outlined in the Police, Crime, Sentencing and Courts Act (2022),Footnote 2 places a new statutory duty on a number of public sector agencies to tackle and prevent serious violence and is part of a range of government initiatives aimed to develop whole system multi-agency approaches to violence prevention (see Bath and O’ Moore 2018, p. 5). The essence of the SVD is in making the prevention of serious violence a legal requirement with duty holders including the police, local authorities, youth offending teams, probation, fire and rescue and health authorities.Footnote 3 As Bath and O’ Moore (2019, p. 5) assert this ‘will help create the conditions for collaboration and regular communication to share data and intelligence to understand and tackle the root causes of serious violence’.

This current plethora of activity makes it a good time to explore inter-agency approaches to violence reduction as areas prepare for the Serious Violence Duty. The findings focus on the East Midlands region of the United Kingdom, which comprises five police force areas (Derbyshire, Leicestershire, Lincolnshire, Northamptonshire and Nottinghamshire) that serve 8% (n = 4.8 million) of the population of England and Wales.Footnote 4 The fieldwork explored the development and delivery of violence reduction strategies through the lens of Bath and O’ Moore’s (2019) 5c’s approach (see below). Bath & O’ Moore not only build upon the idea of the centrality of partnerships to successful crime prevention delivery, but develop a ‘whole-system multi-agency’ approach to violence prevention which provides a framework for our data collection and analysis. This allows us to bring together some important findings that reveal much about current partnership arrangements, workings and preparedness for the SVD. As two of the five areas have Home Office funded VRU’s, this also represents an opportunity to compare the experience of those areas with VRU’s to those without. Ultimately, this has implications for the delivery of public health approaches to violence reduction.

Methodology and analytical framework

The data presented in this paper were collected from each of the five policing areas across the East Midlands region of the United Kingdom through fieldwork completed with relevant stakeholders in the autumn of 2021. Detailed interviews were conducted with thirteen participants responsible for violence reduction, ranging from police officers and analysts, VRU and Police and Crime Commissioner staff to council and health representatives. Due to restrictions in meeting face-to-face imposed by the COVID-19 pandemic, all interviews were conducted virtually using the Microsoft Teams platform. Due to the broad range of subject knowledge covered by the interview questions, each area’s representatives were interviewed jointly. This allowed the most comprehensive answers to be obtained as partnership knowledge was aggregated. The interviews took place following granting of ethical approval, and all participants provided their informed consent. The interviews were structured around the five core elements of the whole-system multi-agency approach to serious violence prevention (the 5c’s) as outlined by Bath and O’ Moore (2019) and focused upon.

  1. 1.

    Collaboration: how areas bring partners and system leaders together;

  2. 2.

    Co-production: multi-agency strategy and intervention development, co-location and establishing collaborative working;

  3. 3.

    Co-operation in data and intelligence sharing: understanding what data are collected across relevant agencies and challenges with this;

  4. 4.

    Counter-narratives: identifying risk and protective factors in local populations and working with communities to produce alternatives to criminal activitie;

  5. 5.

    Community consensus approach: how areas recognise that communities are key assets, knowledge providers and able to provide solutions.

In addition to these detailed interviews, further conversations and feedback sessions were run with each of the areas between late 2021 and early 2022 to ensure the findings were accurate.

While the interview method was able to reveal much about the development of the 5c’s approach across each area, there are two primary limitations that need to be observed. First, as data were only collected across five police areas, they cannot be generalised to be considered representative of all community safety partnership, VRU or police areas in England and Wales. Second, perspectives are only gained from thirteen interviewees, most of whom represent police/OPCCs, so care has to be taken in claiming the findings represent those of all potential partners.

Findings

A requirement of the SVD is for local authorities to prepare and implement a strategy for exercising their functions to prevent and reduce serious violence in their area (Part 1, Chapter 1 Sect. 8: Police, Crime, Sentencing and Courts Act).Footnote 5 How far advanced areas were in relation to the development of formalised violence reduction strategies was largely dependent on what funding had been secured to raise serious violence as a key priority within the area. At present, the VRU areas were the most advanced here. As a condition of VRU funding, both VRUs had a Strategic Needs Assessment (SNA) in place that set a plan for the delivery of violence reduction interventions. Representatives from both VRU areas spoke about developing plans that were focused on public health approaches and they integrated system level theory of change into the SNAs. While the non-VRU areas had often developed plans that were focused on specific local authority areas where serious violence was identified as a problem, these often faced significant challenges with partnership arrangements (see below) and did not include a fully integrated public health approach. The disparity in intervention activity across the VRU areas as compared to the non-VRU areas is evidenced by an audit of intervention activity completed alongside the interviews. This revealed that there were 47 interventions being implemented across the region to reduce violence, with 62% (n = 29) of these being implemented in the two VRU locations.

The reminder of the findings is presented below through the lens of Bath & O’ Moore’s 5c’s framework.

Collaboration with partners

Collaboration is noted by Bath and O’ Moore (2019, p. 21) as ‘bringing partners together from a broad range of functions who have a shared goal of tackling and preventing violence’. Indeed, the sentiments expressed by Bath and O’ Moore (2019) seem to reiterate the notion of partnership working that has been well ingrained in the crime prevention literature for over two decades (see Berry et al. 2009; Homel and Brown 2017). These sentiments were also considered important by our interviewees. All areas had done a great deal to build partnerships—much of which is a legacy of the requirements of the 1998 Crime & Disorder Act. However, there was a distinction between the VRU and non-VRU areas. The VRU areas had developed robust governance arrangements in relation to violence reduction, with strong Office of Police and Crime Commissioner (OPCC) buy-in, strategic boards and violence delivery groups in place (indeed, there is already close involvement of the partners required to comply with the SVD legislation). A constant theme from all respondents was the desire to move away from ‘silo working’ and ‘to bring all the right people together’ (Interview 1), though current arrangements presented some significant challenges for the non-VRU areas. The non-VRU areas are currently working to the legislation outlined in the 1998 Crime and Disorder Act (Sect. 17), which focuses on crime and disorder, rather than serious violence. This means that violence prevention can often be side-lined over other priorities. One non-VRU area noted that as ‘serious violence’ is not a ‘heading’ (Interview 4) in their current community safety plans, only aspects of serious violence get prioritised. It was felt that this can be a hindrance to developing a public health approach to violence—which bringing partners together under a SVD or a VRU banner could possibly remedy. While it was acknowledged that the SVD might help overcome some of these issues, the development of collaboration was also said to be hindered by other issues—especially budgetary constraints and competing priorities of agencies. This resonates with findings of previous research which has noted how resource limitations and getting agencies to all pull together in the same direction can be a challenge (see for example, Rosenbaum and Schuck 2012).

Co-production of strategy and interventions

Co-production includes developing approaches that are informed by all partners. As Bath and O’ Moore (2019, p. 28) suggest this ‘should be informed by the multi-agency perspectives of all partners’. However, it was clear in the interviews that the areas are at different stages of development here and many of the issues observed in relation to collaboration surface again when considering co-production. While all areas noted that some strong local working relationships had developed, the non-VRU areas were most likely to suggest that co-production was hindered by a lack of strategy and a lack of accountability, which again could potentially be overcome through the SVD and/or the development of a VRU. Indeed, one non-VRU area commented on the willingness of some partners to engage, but how the area had been ‘plagued’ by a lack of strategic direction and letting partners know ‘exactly what it is they needed to do’ (Interview 4). Another non-VRU area also commented that co-location is problematic where there is no VRU as the VRU banner enables and fosters the view that collaboration and co-location are required.

While Homel and Brown (2017) note how a lack of integration between partners can be a challenge, one VRU suggested what was required was not just ‘multi-agency’ integration, but the correct multi-disciplinary knowledge. While previous research often advocates ‘multi-agency working’, public health approaches require ‘knowledge that is multi-disciplinary’ (Interview 3). Generally, there was a view that the SVD should help remove silo working, but in order to achieve this, it would be necessary to have partners working together to tackle a range of risk factors (and build protective factors) required for a public health approach. Indeed, it was suggested part of the challenge going forward will be in identifying how partners can ‘pool together’ (Interview 3) the required ‘multi-disciplinary’ knowledge to build effective public health focused strategies.

Co-operation in data and intelligence sharing

If multi-disciplinary knowledge is to be produced, it is evident that data and information sharing is required (Bath and O’ Moore 2019). Respondents spoke about (1) what data were required; (2) issues with data sharing; and (3) achieving partnership data ‘buy-in’ for data sharing. While most areas were clear about what data they would like to have access to, one non-VRU admitted that they still had a depth of learning to go through to understand what data were held by partners that could help them develop a strategy. Others held a better understanding, with one VRU area even being able to develop a multi-agency data dashboard for monitoring and evaluation purposes and another non-VRU area giving an example of a similar data hub they were developing (also with the help of partners).

The challenges of partnership data sharing are nothing new and have been observed in previous research. For example, writing in relation to the formation of crime and disorder partners in 1998, Tilley (2002, p. 32) notes the requirement for partnerships to have reliable data that can be shared across agencies and routinely analysed. However, Homel and Brown (2017, p. 551) observe the challenges that such partnerships faced from the outset with data sharing, which led to a lack of available data for audits. A key issue that surfaced in our research was that data sharing is not only required to develop strategy, it is also required for day-to-day operational reasons and evaluation. Therefore, when one key partner is unable to share data, this can become a significant issue. For example, one VRU noted ‘constant challenges’ (Interview 1) in accessing health datasets (which are needed for information on survivors of knife crime). One non-VRU area did mention that they had few issues with data sharing as they had often referred to the legislation set out in Sect. 115 of the Crime and Disorder Act (1998) which gives partners the power to share data for the purpose of crime and disorder reduction. Further to this, it was also suggested that issues with data sharing could be overcome if (1) there was a dedicated person centrally in the VRU tasked with this and (2) if partners are able to clearly understand why the data are required.

Counter-narratives

The whole systems approach requires that there ‘has to be a commitment to creating environments that nurture protective factors’ (Bath and O’ Moore 2019, p. 38). This involves moving away from traditional reactive criminal justice approaches to a public health focused one that builds protective factors within communities. Such approaches emphasise tackling root causes of violence that involve working with communities and those at risk of violence (as perpetrators or survivors). Of course, there is nothing new about engaging with communities to develop crime prevention strategies. As Tilley (2002, p. 31) notes, a key principle of crime reduction is in acknowledging ‘members of communities know something about their own crime and disorder problems’. All areas were mindful of identifying risk factors and working with communities to develop protective factors against violence. Indeed, one VRU area had several community ambassadors working across a city to help connect with local communities. The other VRU area spoke about their strong efforts to engage with communities through events and trying to integrate community involvement. One non-VRU area spoke about their efforts to do community mapping to try to reach the ‘correct people’ (Interview 5).

Overall, there was a clear message from all areas that new approaches to violence reduction were required, but that the development of such counter-narratives can be challenging. Indeed, four out of the five fieldwork locations had large urban centres, with concentrations of public space violence where development of counter-narratives was thought necessary. In these areas, it was generally noted that gathering knowledge can be resource intensive and knowing who to engage with and how to engage with them is a challenge. Despite this, there was clear evidence that counter-narratives (or moves away from traditional CJS approaches) were being developed. Of the 47 interventions that were being implemented, the majority (62%: n = 29) involved working with a range of partners (such as schools and young people’s services) to provide interventions and activities for those thought to be at risk of involvement in violence or providing universal delivery to those with no known involvement (via schools and social media for example). A further (38%: n = 18) involved working with people who were in the criminal justice system (such as providing mentoring and support in custody settings or working with Integrated Offender management Teams).

Community consensus approach

Community consensus builds on counter-narratives and requires that approaches must be ‘with and for local communities’ (Bath and O’ Moore 2019, p. 46). There was a clear sense across all-areas that approaches to serious violence reduction needed to empower communities and help them to produce solutions. While the notion of co-production and community census is not new, how to achieve community consensus provides a set of challenges. As one respondent said, ‘you don’t do to communities, you have to take them with you’ (Interview 5), but then also observed that partnerships needed to understand and work together with communities to generate trust and provide tools for change. Indeed, Rosenbaum and Schuck (2012, p. 240) note that effective partnerships require a common unifying purpose and respect across agencies that link into communities. The interviewees also pointed to the need to have a common purpose with communities, though the fieldwork areas were at different stages in building consensus. Indeed, building consensus involves collaborating with the correct people—the vulnerable, those at risk and tapping into community networks and local social capital. In order to do this, one VRU had a separate community engagement and participation strategy with a communications plan (which links into the police communications team). The other VRU had invested in a communication and campaigns officer to engage and build networks. However, the non-VRU areas suggested they lacked the resource to be able to engage with communities, but generally spoke about the need to develop better communications with key vulnerable communities in their locations. Indeed, across all areas, there was a general view that partnerships had to develop a shared understanding of community needs to help them to build stronger links with communities, but concerns were raised about effective ways of actually doing this. In line with Crawford (1998), there were worries that targeting specific communities can lead to the same ‘community voices’ being heard. Indeed, there is a risk this can lead to partnership working being what Crawford (1998, p. 263) refers more of a ‘club good’ rather than a ‘social good’ for the community at large.

Preparing for the serious violence duty: discussion, implications and conclusions

Several authors have noted how the development of local crime prevention partnerships builds on a policy of responsibilisation, where provision for crime prevention is largely moved from the central state to the local level (Anderson and Wahlgren 2022). Shearing (as cited in Crawford 1998, p. 252) observed this move in relation to the creation of Crime and Disorder Partnerships in England/Wales and noted how the central state retained a level of control, referring to this as ‘state rule at a distance’. The development of VRUs follows this pattern. There is a degree of local autonomy, though the policy approach as set out in the Serious Violence Strategy and the conditions of funding mean that the state can largely retain control over what VRUs do. The requirements of the SVD can be viewed as an extension of this control from the current VRUs to all community safety partnerships, where there will be a statutory duty for a range of agencies to come together to address serious violence.

Most scholars would support the view that crime prevention requires locally developed strategies (see for example, Tilley 2002; Rosenbaum and Schuck 2012). The evidence from our fieldwork would also concur with this view. Indeed, the development of public health approaches requires close working across partnerships and local communities to co-produce and achieve consensus about what needs to be done to tackle violence. The SVD might be considered a route to the delivery of these partnership approaches and, ultimately, the development of a more holistic public health approach. Indeed, the SVD was generally welcomed across all areas in this study as it was seen as having the potential to facilitate the change required to deliver violence reduction. The SVD could help to generate stronger inter-agency ownership of serious violence issues, accountability for serious violence and change the perception held by some that this is just a police responsibility. It also has the potential to embed VRU expectations and a public health approach nationally.

Although the SVD seems to underline the centrality of partnerships to the delivery of violence reduction, there are three key areas of concern as we move towards its implementation. Firstly, it is apparent that, the current arrangements for violence reduction have allowed for a two-tiered approach to develop. When viewed through the lens of the 5c’s approach, our research identifies that—due to a number of factors—the VRU and non-VRU areas are at different stages in the development and delivery of the process. The VRU areas have started to develop a violence reduction culture. They have better developed strategies that aim to address violence as a public health issue; have developing governance structures in place to monitor progress and more integrated partnership collaboration (and are delivering more interventions). Secondly, the current arrangements have created division in available resourcing and funding for violence reduction activities. While the VRU areas were funded on the basis of the extent of serious violence problems in their areas, it needs to be borne in mind that serious violence also concentrates in non-VRU funded locations across the region, and at present, there is a risk that provision for violence reduction in these areas could be lacking in the future. In interview, one non-VRU area expressed concerns over additional work: ‘What we are being asked to do is quite a lot on top of the day job. I’ve not got staff for knife crime, I’ve not got the budget, winging it is effectively what I am doing’ (Interview 5). Thirdly, many of the challenges associated with partnership arrangements that were observed after the implementation of the 1998 Crime and Disorder Act still appear to exist today. As we move towards the SVD, in the non-VRU areas, some worries were also expressed around ‘historically…. poor partnership buy in’ (Interview 4) and how these might be addressed.

While previous research has highlighted some of the issues in making local partnerships central to crime prevention delivery, developing a public health approach also generates further challenges. It also needs to be borne in mind that the VRU’s are operating with a slightly different remit to previous Community Safety Partnerships (as legislated for in the 1998 Crime and Disorder Act). They are being tasked with implementing a public health approach, which not only represents a shift from the crime reduction remit of Community Safety Partnerships, but an expectation to impact on a broad range of public health outcomes. This requires developing counter-narratives and building community consensus. In a process evaluation of the 18 VRU areas, Craston et al. (2020) noted there are considerable differences in understanding what a public health approach is. Our evidence suggests all areas (especially the VRUs) are making strides to develop interventions that tackle the root causes of violence. However, there are still uncertainties around how to co-produce with communities and build community consensus. Crawford (1998, p. 89) notes that community consensus can, if care is not taken, start echo the moral voice of key interest groups rather than the concerned citizen. This is something that all agencies delivering the SVD will need to be mindful of or public health approaches will lack credibility in the communities they aim to help.

The Serious Violence Duty does, however, present an opportunity to build a national body of evidence about the development of partnership structures designed to implement public health approaches to violence reduction and also to build a wider body of evidence around what works in preventing violence. The VRUs included in the fieldwork for this study had a clear package of evaluation work related to their funding requirements. However, for the non-VRU areas, the lack of a central ‘violence reduction identity’ (Interview 2) sometimes meant evaluation could be a little more piecemeal and haphazard. Another non-VRU area also raised concerns about a rigid ‘hard-nosed’ (Interview 5) performance culture linked to police targets. They noted how positive outcomes from violence reduction might focus on harms and helping the vulnerable in ways not always easily measured through police targets, but that possibly fit better with a more public health focused approach. Indeed, a potential drawback of ‘state rule at a distance’ is that the desire for short-term performance measures of success and lack of funding might result in missed opportunities for robust evaluation of new public health focused interventions. Therefore, the SVD might be viewed as an opportunity for the state to facilitate the development of a vast body of robust evidence through funding evaluation and disseminating findings. This would not only help to build on the existing evidence base, but ensure the Serious Violence Duty yields long-term benefits for crime prevention practice for years to come.