Understanding the rise in violent crime and answering the right questions – before reaching for solutions

Violent crime is now rising in many parts of England and Wales with significant upticks in urban areas in the last two years that reflect a genuine increase in offences, rather than changes in recording practices or more reporting by victims. In response, politicians (and some senior police officers) have leapt to conclusions about what is responsible, and what should be done. But the truth of where we are is harder to acknowledge.

Though we have encountered spikes in violence before – and today’s absolute numbers for youth violence are still lower for a city like London than they were a decade ago (and much lower on a per capita basis because the capital’s population has swelled in the meantime) – there is a growing degree of public anxiety and a genuine problem to be tackled. Unfortunately there is also a casual attitude to the evidence.

Wanted: Evidence and Good Data

Good policy starts with good research, borne of good data, and the truth is for the current problem of urban violence we do not yet have it. There is a perception that this time, the problem is not hard to explain, and that government cuts are responsible (though the exact mechanism is hardly ever articulated). There is then the inevitable politicisation of a public safety problem before the causes of that issue are even properly understood, let alone accepted so widely that the solutions easily present themselves.

The harm and heartache from teenage stabbings and gang murders creates enormous emotional pressure to do something – and for the State, responding to an upsurge in violence is both necessary and invaluable, but rarely happens quickly enough. But a response to a public safety problem needs to be grounded in a proper understanding of that problem, or we risk wasting resources on interventions that will make no difference. And it is clear in the last year that even those closest to the issue – the police and the mayors of the cities affected – do not really know why this is happening, in this way, now. Officials will not admit it, but they are flummoxed.

For assault and homicide, the crime numbers continue to go in the wrong direction, and this is never a profitable scenario for any politician. But it behoves policy-makers and responsible journalists to avoid jumping to conclusions when the only common feature of many of these serious crimes is how hard to predict they actually are. Individual cases attract media attention, and on their own they usually turn out to be too complex for other agencies to be fairly held responsible for not stopping them, but together these incidents do form patterns that we should be able to see, understand, and respond to.

That requires exhaustive analysis, mountains of good, recent (usable) data, and academic and practitioner insights to interpret it. Without that, we will get lost in a partisan argument over how to spend additional money, without first reaching a position of consensus on what the problem actually looks like. The Project for Modern Democracy is contributing to this discussion by convening a number of senior practitioners, policy experts and academics to explore what we know, and what we need to know in order to define the problem and begin to craft solutions.

At a parliamentary roundtable on 10 September, hosted by our chair, the former policing minister, Rt Hon Nick Herbert MP, we will seek to understand the current picture with the former Commissioner of the Metropolis, Lord (Sir) Bernard Hogan-Howe QPM, and the Stanford University academic, Professor Keith Humphreys. Both will present on the issue of urban violence and offer some possible explanations, and practical responses.

A whole set of contextual questions need answering, and up until now, none of these issues has been properly addressed in public debate or media reporting.

The police’s role

Given the scale of their resource, and the evidence of their direct impact on crime rates, the role of the police is critical. However it is not a simple question of whether funding reductions (and by extension, fewer officers on duty), have caused crime to rise. Indeed, those who jump to this conclusion (often the same people who find it hard to credit the police when crime rates are falling), often overlook more complicated changes to police deployment and tactics. If it was simply a matter of police resources, violent crime increases would track reductions in officer numbers (or frontline presence), which have been more marked outside of London – but they do not.

So we need to look behind headline numbers.

Does the leadership of the Met have a grip on the local picture and a structured process for holding commanders to account for local outcomes? Is the Met still performance managing through CompStat style meetings to nip problems in the bud? Have local Integrated Offender Management (IOM) teams been able to continue their important work? Has the well-known Trident Gangs Command been adequately resourced? Is the intelligence picture and the matrix of known gang members proving effective, or are risk scores no longer aligning with those who are now committing random acts of violence?

If the overuse of stop and search was a contributing factor of the riots in 2011, can it also now be behind a rise in violence among offenders of a similar age? Some have argued that the pendulum may have swung too far so the tactic is now too infrequent to be effective, with its use in London having been cut by 19.8% since 2015.

Conversely, if the answer to violence is more stop and search, then that implies that the cause of violence is weak deterrence creating a culture of impunity that encourages young people to arm themselves, and therefore to become more readily capable of serious violence and also more likely to be a victim of it. But is this proposition falsifiable, and do we even care to find out?

Stop and search is an important tactic, but it is not a credible policy response on its own because when overused it is a very blunt tool with well documented side-effects for community trust. However, if detailed local data showed a compelling correlation between the volume and concentration of stop and search – including with other behaviour like the stashing of weapons – then the debate could be more informed.

Are policing resources in boroughs now out of kilter with the demand on the ground? Should more be shifted into response teams or neighbourhood units, and taken from central borough units addressing signal offences like hate crime for example? If a city’s police cannot afford to service both demands, then they have to prioritise – and with political cover – to respond to the public’s primary concern. Has the deployment and staffing at a local level been adjusted to respond to more stabbings, shootings, and street homicides, or has the Met simply skimmed officers into a newly created squad at the centre? In an organisation as large as the Met Police, is a unit of 150 officers – smaller than the force’s communications and Freedom of Information department – going to make a real impact?

Factors beyond policing

Beyond policing, have wider changes around enforcement, offender management and judicial restrictions been negatively associated with rising violence?

Have budget cuts to Youth Offending Teams played a part, and if so, how? Are YOTs now less capable of monitoring the clients on their books, or have the interventions that might address offending been scaled back, to the detriment of those who needed them?

Is the gang structure in London changing, with more young members and new modes of enforcement and retaliation? Are gang feuds triggered more frequently by social media, or by other factors? Are aggressive outbursts often predicted by an altercation online, that social media analytics might be able to predict in advance, or are they sometimes also triggered by psychosis?

Do changes in mental health and addiction treatment services provide some explanation? How many of the offenders and victims of recent violent assaults fit these categories? Is the more mundane explanation of most modern urban violence – alcohol – a consistent feature of those who stab, as well as those who punch? What about street cannabis, and the drug markets that are underpinned by it?

There is some evidence that school truants are disproportionately responsible for youth crime (and also that shifting the school day back a couple of hours can reduce volume crime), so what role have these pupils played in the latest uptick in violence? Are the offenders typically enrolled in mainstream education or attending Pupil Referral Units, and how many are frequently absent?

What role, if any, has the (laudable) aim to keep teenagers out of the justice system (and reduce first time arrests) played in the weakening of a deterrent message, or more practically, in keeping out in the community more young people who are on a criminal career trajectory and who yet may not be adequately supervised? Are we paying a price in street violence for a dramatic reduction in the youth custody population over the last decade (and particularly after 2015)? Which community sentences are working, and how can we make courts and the sentences they pass, more effective at mitigating violence, rather than simply requiring curfews or rehabilitation courses?

Are our violent and disorderly prisons themselves now playing a part in the upsurge in street violence, with drug feuds and organised gang activity leaking out of jails into the community itself, and how is this worse than what has always happened to some extent? How many were recently released from prison, and what did their post-release plan amount to?

Are shifting drug markets to blame? We know there are new psychoactive substances, high potency street cannabis, and now even fentanyl products making up a lucrative trade that continues to evade suppression, while fueling turf wars between organised drug dealers and their suppliers. Given a big drop in the number of drug offences – largely a reflection of fewer stops and police arrests – is the wider market now becoming more violent for being less proactively policed?

What has the breakdown of an effective probation service led to, and have staff been unable to provide adequate supervision of the right people? How many (and what share) of the perpetrators of serious youth violence were in the charge of a Community Rehabilitation Company at the time of their offence?

There are lessons from the August Riots in 2011 – including around how to respond to gangs, and the importance of a swift court response to crime – that have been arguably forgotten, or at least, set aside as exceptional and unrelated to business as usual. If the causes of that earlier eruption of rioting were social, and caused in part by cuts to youth services as some claimed at the time, why did general levels of violence take another five years to begin increasing? What, if any, of the same factors are at play now?

Answering these and other questions is the first priority.

Using data to gain new insights

What is published in terms of the data map is better than many countries, but still remains inadequate. Where the rates of violence are worst, we would hope that MOPAC and the Met Police in London (but also PCCs in other parts of the country), are working with much more granular datasets that encompass more than just crime and policing information.

The unique advantage of an elected Mayor is to utilise their position to coral other agencies to collaborate on a shared problem, and that cannot begin to happen without a clear picture, which is impossible to construct with just one agency’s data. City Hall in London is the best positioned – they have education, demographic, environmental and transport data that could be integrated with the Met Police’s data. This might provide new insights into the drivers of violence, or allow agencies to disregard possible explanations – for instance, the contribution of absentee pupils to violence, or those with mental health conditions.

Until 2016, the Met Police was trialing a sophisticated predictive analytics tool supplied by Palantir (and used in the LAPD), but discontinued its use in favour of developing their own in-house platform. Has this been implemented and if so, is it working to predict the likely locations of violence? The earlier trial showed promise insofar as volume acquisitive crime was concerned, and other cities around the world have proved that other products (from firms like StreetScan) can do the same, with a high degree of accuracy, for violence (assumed to be more random, dynamic and therefore less predictable). If such tools are not already deployed against this problem in London, then they need to be – and it is a capability that the Mayor’s Office should be supporting.

Much of this detail will not be readily available for analysts to crunch, and will take time and resources to compile. But software products can now provide quick insights across blended datasets and reveal patterns and associations that humans cannot see. If it means examining every homicide (and near homicide) case file for relevant social and other data not available on the Police National Computer, and digitising every feature of an offender and victim so their case can be analysed, then that will be a time investment that pays off.

New ideas for a familiar problem

Urban violence is not alien to us, and many cities around the world have periodic spikes in violence that are unrelated to changes in public spending or local police practice, or justice system policies. Nor are they so radically new – as some have implied – that they represent a wholly unfamiliar challenge. But the issue is urgent and serious, and despite the rising homicide total and the political heat around this issue leading up to the next local elections, we are still in need of much more light, and a better informed debate.

Many previous crime spikes have seen some of the same knee-jerk explanations that feature in today’s discussion, with even the Met Commissioner resorting to the simplistic suggestion that a new teenage music trend in the black community is partly responsible for glorifying violence. We can do better than that.

So where are the new ideas? Where is the new data on offenders, victims and locations? What do we know that we do not know? For each city the factors will be complex and the local context will be key.

Government Ministers cannot answer these questions. Instead, Police & Crime Commissioners and Mayors should be leading a public discussion about the causes of urban violence and what tools they have to address it, in a way that is honest, and humble. And where they want to try new approaches, these should be backed up by evidence, or at least ideas that have been validated elsewhere and might be worth a try. That might mean adopting the Scottish public health approach to enforcement and treatment, or it might mean reviving London’s Operation Shield pilot from 2014-15 of targeted deterrence using gang call-ins, first pioneered by David Kennedy in Boston.

Keeping this issue internal, and leaving the Home Office and local agencies to wrestle with the possible causes and solutions, is short-sighted, anti-democratic, and also unnecessary. It only fuels a perception that the police and others are complacent, or that they are the opposite, and because they think they know more than anyone else they do not need help from outside, external policy input, or further performance scrutiny.

What the justice system can do

Many solutions may exist outside of the conventional justice system toolkit – for example legalising and regulating the market in cannabis, or changing school opening hours, or tracking some internet use – and they deserve to be explored. But many of those policies will be highly contested and take years to devise and implement, and the crisis of young people dying is here and now. So in keeping with our vision for our justice and security programme, our research is focused on what State agencies can do to make a difference now, rather than the wider set of social and economic factors that might provide longer-term solutions.

Where the police and justice agencies are able to make an impact, this may involve rediscovering, or reinvigorating, a familiar practice or a previously used tactic, rather than reinventing the wheel, or importing a whole new idea. Equally, it may call for a radical departure that deserves to be tried out – even if the full implications for system-wide adoption are too hard to work through now.

The upcoming event P4MD is hosting is designed to clarify the questions we need to answer – not to provide a list of definitive answers. Our proposition is simple: there is much we do not yet know, but if we understand today’s problem, and marshal the evidence of what works to reduce violence, then we can decide how the justice system needs to respond to it, given the part that it is able to play. The rest is for others to contribute – be they schools, community groups, health agencies, or local government – based on a shared picture of the problem, clear and trusted metrics to baseline and measure progress, and a large amount of relevant, recent data. We will not fix this problem by retreating to old explanations, or simply trading anecdotes and assumptions, however well-intentioned.