Reframing urban violence

The Project for Modern Democracy convened a seminar on urban violence last week, in which leading policy makers and senior practitioners explored the potential drivers of rising crime and sought to address some of the questions we posed earlier this month.

Defining the problem

The premise for the event was a sense that the political discussion around current spike in urban violence in England is at risk of becoming muddled, if it isn’t already. Participants agreed that though the category of ‘urban violence’ could cover many types of crime, the most urgent issue was the serious violence affecting young people, predominantly in urban settings, and often involving guns and knives in the public realm.

In trying to understand the violence, there has been a tendency to present the policy decisions of the past – cuts to social services, or a dramatic reduction in rates of stop and search – as the primary causal drivers and to reach instinctively for well-worn policy responses as a result. However, as Lord Hogan Howe said, if the analysis of the problem is flawed and we choose the incorrect response, there is a risk of amplification, not just misallocation of scarce resources.

The justified sense of moral outrage around the spate in stabbings risks tipping into a moral panic. But if burgeoning public outrage nudges the police into a premature tactical response, or one based on impartial, incomplete or misleading data, the eventual outcome is unlikely to be effective. The roundtable was an attempt to bring the debate back to the causes, to explore the available relevant data, and chart areas where more information is required to truly understand the drivers.

Firstly, operational changes in the criminal justice system have resulted in improvements in the recording of violent crime in recent years. A more proactive Police Inspectorate, the advent of PCCs and increased mayoral involvement in policing has also spurred better data collection and recording as attempts are made to respond appropriately to local crime.

However this does not mean that the increases are not real. And significant gaps remain. We lack a typology of violent crime, and so cannot begin to disaggregate the problem properly. To begin with an ostensibly simple question – and one that should be central to any narrative of current urban violence – how many people are stabbed? The number of people wounded is recorded, but of course not all wounds are stabbings. We measure how many people carry offensive weapons, but not all of those weapons are used. When we record incidents of serious attempted violence, we still do not differentiate stabbings from other kinds of attempted violence.

Marshalling evidence

It is not only policing data that is limited. Looking to data sets from other emergency services, hospital admissions and other health reports only give a description of the injury without necessarily recording that it was a stabbing, and also rely primarily on the victim’s own testimony. If knives are commonly used in in gang-related violence, victims – who may be gang members themselves or may fear retribution – are not likely to divulge the source of their injury. Advances in emergency treatment for stabbings, air ambulance response times, and crime scene medical triage have also occurred since the previous spike, further complicating comparisons over time. However it is also possible that the under-performance of the London Ambulance Service may have its own part to play in the difference between serious woundings that lead to death and those that do not.

Historically, attempts to mesh health and crime data have been limited. While there are obvious concerns around patient confidentiality, there is no consistency between hospitals on data sharing with law enforcement agencies. Cross-comparing the available data does paint a picture of increasing serious violence, though only in the broadest of brush strokes: there has been a recent coterminous rise of approximately 18 per cent in reported incidents of violence and a 14 per cent in hospital admissions for injuries sustained through violence, with a rise in injuries sustained through serious violence within that, but the data remains uneven.

Given the broad nature of the current picture, what other similarly broad drivers could be at work? What has changed since the last spike in urban violence in 2007-2008? There is a well-documented causal relationship between economic performance and crime, both in terms of intensity and the sorts of offences committed. It was believed economic booms drive more socialising and higher rates of substance use, and as a result more violence, where recessions lead to more theft and other acquisitive crime. Yet the 2008 crash and subsequent recession did not follow this pattern.

Secondly, the population of London has increased significantly since 2008, but this increase is not demographically or geographically uniform. London is getting younger, and the population balance is shifting to the East. Those boroughs that have seen the biggest increases in numbers of young people – particularly young men – have seen above average increases in violence.

London’s booming population forms the backdrop to all planning in City Hall, so why is this not being factored into our understanding of the phenomena of urban violence? The population explosion has other ramifications for the data gathering and analysis necessary to build up a useful understanding of urban violence. Many more people are now living undocumented in sub-par accommodation in so-called ‘beds in sheds’, or are living ‘off the radar’ of the criminal justice system as a result of exclusions from school, illegal migration, or other factors.

Risk markers and victim profiles

Going beyond the possible macro-drivers, the discussion shifted to ways of capturing a micro-profile of those involved in urban gang-related violence, whether as victims or perpetrators. Alice Miles from the Office of the Childrens’ Commissioner revealed a worrying trend of pupils being directly recruited into gangs from pupil referral units, and suggested that similar societal factors that drive gang membership also drive exclusions from mainstream education.

Rory Geoghegan from the Centre for Social Justice think-tank pointed to the inadequacy of data differentiation for victims of gang related violence in London, particularly of the Home Office Homicide Index. He suggested that in future, when a young person dies or is seriously injured, there should be something akin to a Serious Case Review to better understand the contact points.

Roger Grimshaw from the Centre for Crime and Justice Studies emphasised the history of sustained emotional trauma and abuse and particular forms of destitution that so often is associated with gang membership, and goes some way to explain why they are willing to take these extreme risks. Blair Gibbs then suggested that this longer term health and social data – including family data – could perhaps be drawn upon alongside other markers of risk (such as exclusion from full time education, domestic abuse, or family structure) to build a better portrait of children at risk – an exercise that is already being explored by the Childrens’ Commissioner.

The policing dimension

Discussing deprivation-related drivers behind gang recruitment and violence invariably led to a discussion of cuts to government services enacted since the previous spike. Changes to policing tactics – whether driven by financial imperatives or otherwise – clearly had some part to play in the eyes of some invitees, though a dearth of data meant that the precise mechanism remained hard to pin down.

Rory Geoghagan noted that arrests and prosecutions in possession and trafficking cases had both declined, and suggested that the street-level drugs trade played a central role in the uptick in violence. Lord Hogan Howe noted that cuts to police numbers have a disproportionate impact on neighbourhood policing, as often they struggle to quantify and define their impact in comparison to other police units but are essential in providing intelligence to disrupt street-level drug dealing and identify at-risk young people.

He also suggested that types of crime involving resource-intensive specialised investigation, such as historic sexual offences and cyber crime, are also now more prevalent, which has had an effect on demands on wider investigative resources for traditional ‘street crime’.

Significant recent changes in the street level drug trade have also weakened the assumed causal lever between police cuts and rising rates of violent crime. It was noted during the discussion that there are now lower barriers to entry as a result of technological change: widely available smartphones and encrypted messaging services allow for one-man drug dealer operations. Is this generating sporadic disorganised conflict over territory?

Means and ends, rewards and sanctions.

Penelope Gibbs of Transform Justice made a wider point about the effectiveness of any criminal justice response to rising violent crime: that effective deterrence is rooted not in the severity of any eventual punishment, but in the likelihood of being caught. Assuming that reduced use of stop-and-search and lower arrest rates for drug possession and trafficking mean that there is less perceived risk of being caught, could this be a potential driver of urban violence? Equally, there is no evidence that the blunt sanction of imprisonment for unlawful possession is working to deter knife-carrying.

There was a consensus that certainty of detection and sanction, not severity, was critical in any policy response by the criminal justice system. Professor Keith Humphreys delved deeper into this question in his presentation on tackling substance-related violence, while also drawing out questions around the ability of criminal justice sanction to address the fundamental neurological profiles of offending behaviour such as impulse control and decision-making, especially in those with a history of addiction.

Professor Humphreys gave a compelling overview of a South Dakota sobriety pilot, later rolled out across the United States, which used swift, certain, (but modest) sanctions to change behaviour, helping to reduce drink-driving, but also other behaviour like domestic violence. It was suggested that our traditional criminal justice model was especially inept at responding to youth violence, with inconsistent sanctioning of gang members after repeated warnings and the long response timeframes of the court system were both considerable hurdles for an effective criminal justice response to urban violence. Similarly, Nick Hardwick emphasised the importance of consistency and certainty for through the gate services for offenders leaving prison.

New thinking about sanctions

The success of 24/7 sobriety as a radically new type of criminal justice response is one that P4MD is continuing to examine as part of our research programme. Even if current youth violence involving knives is not often drink-related, the principles of how to change behaviour seem just as relevant. Could an intervention with a similar logic to 24/7 sobriety – one grounded in neuroscience and focused on outcomes that ensures personally and socially damaging behaviour is addressed – be formulated in response to urban violence, and particularly to knife crime? If a mandatory jail term for knife possession – which is rare, inconsistent, applied late, and usually does not seem proportionate when it eventually comes – fails to work, are there other options we can develop for the courts which are swift and certain and more targeted?

If we want to change social norms around knife-carrying, do we need to revisit anonymous reporting channels that might actually be familiar to and used by teenagers? Do family members need a trusted way to raise the alarm without the fear that the result will be a prison sentence and a criminal record for their son or brother? Behavioural science and the evidence of what works in court-mandated sanctions should be able to provide some options here, and P4MD will be doing further work in this area in the coming months.

Innovation at a local level

But as attempts to roll out compulsory sobriety sentencing akin to North Dakota, first in London and then elsewhere have demonstrated, there are organisational and systemic barriers preventing the easy transmission and implementation of novel policy solutions in the context of the centralised English state. Nick Herbert raised the issue of the ability of local mayors and PCCs to drive policy innovation, or to marshal other public service bodies (plus their data and other resources) to collaborate with police, something that was central to the success of the Glasgow serious violence strategy.

Will Tanner raised the question of whether there had there been a deterioration of relations, or an unpicking of local links, between these services and the police, given how critical local coordinated action and good leadership was to effective responses on the ground. Lord Hogan Howe gave an account of some of the difficulties that had arisen with a previous attempt at importing a Ceasefire-style response (a strategic framework for targeted deterrence originally developed in Boston before being replicated elsewhere). The convening power of the Mayor and Commissioner in London had not been sufficient to overcome the reluctance to engage from many of London’s 32 boroughs during the rollout of the Operation Shield pilot in 2014. Blair Gibbs did suggest that current crime levels and the public mood around urban violence might now mitigate some of those political difficulties, or at least provide a greater impetus for local authorities to overcome them, but it would need mayoral leadership.

Local efforts to corral disparate data to identify at-risk young people in Manchester and Bristol were also underway, under the aegis of the Mayor’s Office, and the Avon & Somerset PCC, with funding from the Home Office Transformation Fund respectively. However, Rob Allen noted the importance of a genuine, properly resourced response once these people are identified. This was his dispute with the Home Office’s Serious Violence Strategy which was right to identify the importance of prevention, but did not square the circle for how resources could be brought forward from the ‘unproductive’ back end of the system’. Allen pointed to the possibility of devolved funding models, or ‘making people eat their own smoke’ through criminal justice reinvestment in things like child services at a local level.

The wide-ranging discussion helped identify some of the more persistent and troubling blind spots in our view of urban violence and the levers available to respond to it, and how these could be addressed. What also became apparent over the course of the conversation were some of the limits of a response by the criminal justice system to the phenomena of urban violence, and the importance of other agencies and organisations – schools, youth services, hospitals and communities themselves – in both building a cohesive and detailed picture of what is actually going on with urban violence, and in deciding on an appropriate and effective response.