Buy art at www.brucehoulder.art and support Fighting Knife Crime London
By Professor Stan Gilmour, Oxon Advisory
I write in connection with Ending Coercive Offending (ECO), a Community Interest Company built on a principle that should not need stating but plainly still does: no child should be criminalised as a result of their own exploitation.
Youth crime has declined significantly since 2019, yet NRM (National Referral Mechanism) referrals have risen year on year, with criminal exploitation, most commonly county lines activity, accounting for around half of all child referrals by 2024. Ministry of Justice and Home Office data confirms both movements: fewer child offenders, and more safeguarding referrals for children identified as potential victims of modern slavery. Yet the UK publishes no linked data on how many of those victims are subsequently charged or prosecuted, making it impossible to assess whether statutory duties to protect them are being upheld. ECO exists to close that gap, which is precisely why the government’s announcement of 11 February 2026 deserves careful examination.
The announcement and what it gets right
Every child caught carrying a knife in England and Wales will receive a mandatory, tailored intervention plan through Youth Justice Services, backed by a £320 million commitment and multi-year stability for Turnaround. The Policing Minister called it “an immediate, mandatory intervention, no excuses.” For families devastated by knife crime, that carries real weight. For practitioners who have argued that children carry knives out of fear or coercion rather than inclination, the investment signals something promising. But there is a design feature that deserves scrutiny.
How mandatory support becomes a criminal pathway
The government is explicit: these plans will be mandatory, with “no child able to avoid getting the support they need.” Failure to engage will have “serious consequences, such as possible criminal charges and time behind bars.” That means a child can face prosecution not for carrying a knife but for non-compliance with a welfare plan, creating a new criminal exposure by administrative design.
Many children who carry knives do so because someone with power over them has instructed or coerced them to. County lines operations rely on exactly this dynamic, and NRM data increasingly confirms the knife-carrier may be a victim of modern slavery. Disengagement from statutory services is rarely defiance; it is a survival response shaped by exploitation, trauma, and rational distrust of institutions that have failed to protect them. Mandating compliance under threat of prosecution ignores all of this.
We have seen this before
This pattern has repeated itself across thirty years of youth justice policy. The Anti-Social Behaviour Order, introduced under the Crime and Disorder Act 1998, is the most instructive precedent. ASBOs were civil orders framed as welfare tools, but breach was a criminal offence: five years’ imprisonment for adults, two years’ detention for children. Ministry of Justice sentencing data showed that children received longer average custodial sentences for ASBO breaches than adults did. The Youth Justice Board Chair warned from the mid-2000s that such imprisonment was rising sharply, and a 2005 National Association of Probation Officers assessment found people being imprisoned for behaviour not in itself imprisonable.
Those most likely to breach were children with mental health problems, learning disabilities, neurodevelopmental conditions, and chaotic home lives: precisely those the welfare framing was meant to help. Use varied widely across authorities and fell disproportionately on disadvantaged young people. ASBOs were abolished in 2014, the coalition government citing high breach rates, poor evidence of effectiveness, and a 2012 survey in which only 8% of the public believed they had worked.
Knife Crime Prevention Orders, introduced under the Offensive Weapons Act 2019, replicated this architecture. Civil orders preventative in stated intent, they imposed positive requirements including mentoring, counselling, and educational attendance, and could be made against children as young as 12. Breach carried up to two years’ imprisonment, imposed on the civil standard of proof, meaning a child could face criminal penalties on evidence insufficient to secure a conviction for the original act.
The pilot evaluation was damning. A youth offending service manager interviewed by UCL researchers captured the circularity: “If you’re saying to a child ‘don’t carry a knife’, well that’s already against the law. So what are we hoping to achieve here? Often, a child’s got a sentence but then you’re giving another order that is exactly the same. If they breach one part, they breach two things, which is treated more severely. These are children.” Practitioners reported no evidence base for such orders, requirements routinely going unmet for want of funding, and disproportionate use against young Black men and boys. Of 138 KCPOs issued during the pilot, the MPS Strategic Insight Unit found 44% involved under-18s.
Scholars have framed this pattern using Adam Crawford’s analysis of how welfare language can be appropriated to legitimise extensions of state power that bypass due process. Hendry (2022) in the British Journal of Criminology characterised KCPOs as producing the stealth criminalisation of disadvantaged children, with inadvertent breach structurally probable given who receives orders. Jo Phoenix (2009) described the underlying dynamic as “repressive welfarism”: punishment positioned as care where genuine welfare has been dismantled. Marshall’s 2023 research in Children and Society found the same mechanism in current exploitation responses, with practitioners using criminalisation as a safeguarding tool because it was often the only route to intensive support under austerity. As one officer put it: “It does sometimes take an offence to happen to be able to get that support, which is really wrong.” The February 2026 announcement risks recreating this structure, with non-compliance now generating the criminal exposure.
The JUSTICE organisation’s 2023 review of more than thirty behavioural control orders found recurring failures: inadequate procedural safeguards, disproportionate impact on people with mental health conditions and learning disabilities, and a legal aid landscape too depleted to mount effective challenges. Children breach at higher rates than adults, and positive requirements go unmet because the services needed are unfunded, and the criminal consequence follows regardless. The government’s own Home Office framework guidance for KCPOs acknowledged that punitive sanctions have little impact on reoffending and described enforcement as a blunt instrument, even as those sanctions stayed in the legislation. The February 2026 announcement reproduces this contradiction at scale.
The Turnaround evidence problem
The government cites Turnaround’s 90% non-reoffending rate as the evidence base. That figure is real, but its context is everything: Turnaround works because it is voluntary and relationship-driven. Lifting a single outcome metric from a voluntary model and applying it to a coercive compliance framework is not evidence-based policy. Research consistently shows that coercive interventions with children embedded in exploitation, fear, or neurodevelopmental difficulty tend to deepen disengagement from statutory services rather than resolve it.
Neurodevelopmental need and the compliance trap
A significant proportion of children who carry knives have unmet neurodevelopmental needs, including traumatic brain injury, ADHD, and related conditions that affect capacity to engage with formal plans. Research shows how rarely these are identified by the systems these children encounter. A child with an acquired brain injury may simply be incapable of meeting the terms of a mandatory plan; under the announced framework, the consequence is criminal charges. That is not welfare; it is the justice system reaching further into a child’s life through the language of care, and it is the institutional pattern from which trajectories into serious harm are built.
What accountability would look like
None of this is an argument against investing in Youth Justice Services or early intervention. Both are overdue. The real question is whether the coercive architecture threaded through this scheme will produce the outcomes it promises, or whether it will add a new layer to the accountability gap ECO was built to close.
At minimum, the government should publish linked data on every child referred under this scheme: how many are identified as exploitation victims, how many fail to engage, how many face prosecution for non-compliance, and how many were already in the NRM. Without it, we cannot distinguish a scheme that protects children from one that simply processes them through a different gate.
ECO’s founding purpose is clear: no child should be criminalised as a result of their own exploitation. The February announcement could be a step toward that goal. The coercion embedded in its design means it could also move us in precisely the opposite direction. That is worth saying plainly, and worth watching closely.
Professor Stan Gilmour is a criminologist and Global Strategic Consultant at Oxon Advisory and Research and Implementation Advisor at ECO, with 30 years of policing experience spanning homicide investigation, counterterrorism, and violence reduction. He is a contributor to global prevention in criminal justice and publishes the Doing Violence Reduction newsletter.
