Youth diversionary activities: what the latest data reveals

It is the difference between a young person entering formal prosecution at an early stage and receiving structured prevention support instead. Research highlighted by North East Lincolnshire Council indicates that those formally prosecuted during the early stages of criminal or anti-social behaviour are six times more likely to re-offend than those supported through prevention and early help.

That finding changes the policy question. The issue is not whether enforcement is necessary. It is where enforcement sits within the sequence of intervention, and whether formal prosecution is being used before less damaging and more proportionate options have been exhausted. Recent local results from Nunsthorpe, the Youth Engagement Alliance, the JEFF project and Safer Streets activity suggest that diversion is becoming a measurable preventative framework rather than a general statement of intent.

The shift from prosecution to prevention

Youth crime prevention programmes in Grimsby and elsewhere in the borough operate within a practical constraint: early behaviour is often unstable, inconsistent and responsive to intervention. A young person involved in low-level anti-social behaviour may be displaying several overlapping problems, including poor school engagement, family stress, unmanaged trauma, substance use, social isolation or a lack of structured activity.

Formal prosecution addresses the immediate incident. It does not necessarily address the conditions that produced it.

This distinction is central to reducing youth reoffending in North East Lincolnshire. A prosecution can establish accountability, but at an early stage it may also strengthen contact with the criminal justice system without changing the underlying pattern. The local six-times figure does not mean that every prosecuted young person will re-offend, or that diversion prevents all future offending. It indicates a substantial difference in observed risk between early formal prosecution and prevention or early-help routes.

That is a policy signal with direct implications for resource allocation.

A diversionary model can include supervised activities, youth work, targeted mentoring, education support, restorative processes, family intervention and low-intensity mental health assistance. These measures are not interchangeable. Their value depends on whether they are available early enough, whether referral routes are clear, and whether organisations share information without creating unnecessary administrative delay.

The strongest model is therefore not a single programme. It is a connected local system.

Early prosecution may resolve the incident on paper while leaving the cause untouched. Diversion works when it changes the sequence before offending becomes established.

National data provides a broader indication of this direction. For the year ending December 2025, 43% of Youth Justice Service outcomes were resolved through diversionary measures without formal youth cautions or court involvement. The figure does not establish that every diversionary outcome is effective, nor does it provide a North East Lincolnshire-specific re-offending rate. It does show that non-court responses are now a significant component of youth justice practice rather than an exceptional alternative.

What diversion is designed to change

The impact of diversionary schemes on anti-social behaviour is best understood across several timeframes:

• Immediate disruption: removing a young person from an escalating situation and providing a supervised alternative.
• Short-term stabilisation: creating regular contact with trusted adults, structured activities and practical support.
• Medium-term prevention: improving school, family or health engagement and reducing exposure to settings associated with repeated incidents.
• Long-term reduction: lowering the likelihood that low-level offending develops into a persistent pattern requiring more intensive criminal justice intervention.

These outcomes require different metrics. A reduction in reported incidents can indicate a local improvement, but it does not by itself demonstrate individual behaviour change. Attendance figures show reach, but not necessarily unique participants or long-term outcomes. Referral numbers demonstrate system activity, but not the quality of support.

A serious assessment therefore needs more than one indicator.

Nunsthorpe: measuring a multi-agency intervention

The Clear, Hold, Build initiative on Grimsby’s Nunsthorpe estate offers the most concrete local example of a coordinated approach. The initiative was launched in September 2023, with the period from July 2024 to June 2025 providing a defined basis for comparison.

During that period, reported anti-social behaviour incidents fell by 35.1%. Public order offences fell by 53.6%.

Those reductions are significant, but they should be interpreted accurately. The available figures show change in reported incidents and offences within the intervention area and period. They do not prove that youth diversion alone caused the reduction. Clear, Hold, Build is a multi-agency initiative. Its effects may reflect the combined contribution of policing, environmental measures, enforcement, community engagement, local services and youth activity.

That is not a weakness in the evidence. It is a description of how neighbourhood safety interventions operate in practice.

The term “multi-agency” can become administrative shorthand, but in this case it points to a specific cause-and-effect structure. “Clear” addresses immediate criminality and disorder. “Hold” sustains attention and stability in the area. “Build” develops longer-term capacity, relationships and prevention. The three components are mutually dependent.

If enforcement clears a location without maintaining a presence, activity can return. If a neighbourhood is held through policing but no alternatives are built for young people, the intervention remains reactive. If positive activities are provided without a credible response to serious disorder, participation may not translate into wider safety gains.

The Nunsthorpe results in context

The difference between the two reductions is also relevant. Public order offences declined more sharply than reported anti-social behaviour incidents. This may indicate that the intervention had a stronger effect on particular forms of visible disorder, or that reporting patterns and offence classification changed differently across the period. Without a full breakdown of locations, incident types and repeat involvement, it would be premature to make a narrower claim.

The available data supports a restrained conclusion: coordinated intervention was associated with a marked reduction in local disorder, and the results justify continuing to examine the preventive component rather than treating enforcement as the complete solution.

For local policy, this has two consequences. First, neighbourhood initiatives should define their baseline and measurement period before launch. Second, evaluation should separate outputs from outcomes. The number of sessions delivered is an output. A sustained reduction in repeat incidents is an outcome. Both matter, but they answer different questions.

The scale of youth engagement networks

The Youth Engagement Alliance and the JEFF project provide a different type of evidence. Across 2024–2025, the network of voluntary, community and social enterprise organisations recorded more than 11,000 attendees across its youth positive activity provision.

The number demonstrates reach. It does not represent 11,000 unique children and young people. It may include repeat attendance, which is not a defect. Regular attendance can be the mechanism through which trust, routine and behaviour change develop. But the distinction must remain explicit when assessing performance.

A single visit shows access. Repeated participation may show retention. Neither alone proves reduced offending.

This is where youth engagement strategies in North East Lincolnshire need more developed metrics. A credible local dashboard would distinguish between:

• total attendances;
• unique participants, where data protection and operational arrangements allow;
• repeat participation over defined periods;
• referrals from schools, police, youth justice and community organisations;
• completion of agreed intervention plans;
• changes in reported anti-social behaviour linked to intervention areas;
• participant progression into education, training, employment or further support;
• subsequent contact with the youth justice system.

Not every provider will collect every measure. Smaller voluntary organisations may not have the capacity to operate a complex evaluation system. The answer is not to impose a bureaucratic burden that reduces frontline provision. It is to establish a small shared measurement framework, with consistent definitions and proportionate reporting.

The local network’s size also matters for coverage. A programme can have strong attendance while remaining inaccessible to young people in the highest-risk locations, at the times when incidents are most likely to occur. Conversely, a modest project may be strategically important if it operates in a gap left by larger services.

Participation data should therefore be read geographically and temporally, not only as a borough-wide total.

The JEFF project and the role of mental health support

The JEFF project—Journey to Enrichment, Fulfilment and Friendship—is jointly managed by the North East Lincolnshire Youth Justice Service and Compass Go. It is commissioned through the ICB and NHS England and provides early, low-intensity mental health and trauma-informed support.

This is an important part of the preventative framework because mental health provision is often treated as separate from crime prevention. In practice, the boundary is less clear. A young person’s offending behaviour may coexist with trauma, anxiety, emotional dysregulation or difficulty forming stable relationships. Addressing those factors does not excuse the behaviour. It can improve the probability that other interventions will work.

The relevant question is not whether mental health support is a substitute for accountability. It is whether accountability is more effective when a young person can access support that addresses the circumstances affecting their conduct.

Trauma-informed practice is particularly relevant to diversion because it changes the operational response. Staff are trained to consider how previous experiences may affect behaviour, trust and engagement. This does not mean removing boundaries. It means applying them in a way that reduces unnecessary escalation and makes continued participation more likely.

There are several practical advantages:

1. Earlier identification of need. Low-intensity support can be offered before a situation reaches a threshold requiring specialist clinical intervention.

2. Improved engagement. A youth worker or community organisation may be a more acceptable first point of contact than a formal service.

3. Better continuity. Support can be integrated with positive activities rather than delivered as a disconnected referral.

4. Reduced escalation risk. Staff who understand trauma-related responses may be less likely to interpret withdrawal, anger or distrust as simple non-compliance.

5. More proportionate intervention. Young people can receive support matched to the level of need rather than being moved immediately into a more formal pathway.

The limits are equally clear. Early low-intensity support is not a replacement for specialist treatment where risk or clinical need is higher. Nor can it establish a direct causal link between attendance and reduced offending without follow-up data.

The value of JEFF is therefore partly systemic. It creates an intermediate layer between universal youth activity and formal criminal justice or clinical intervention. That layer is often where prevention either succeeds or fails.

The Haven Centre and Safer Streets activity

The Haven Centre on Osborne Street opened on 6 December 2023. During Safer Streets 5 operations, it recorded more than 1,500 attendances by young people between March and September.

Again, attendance is not the same as unique participation. The available figure should not be read as evidence that 1,500 individual young people used the centre. It is a measure of recorded visits. That distinction is essential, but it does not make the figure uninformative.

Repeated visits may indicate that the venue has become a functioning point of contact. A static facility can provide several things that dispersed activities cannot: predictable opening arrangements, a known location, staff continuity and a visible alternative to public spaces where anti-social behaviour is more likely to occur.

The timing of provision is a practical issue. Youth services that operate only during conventional daytime hours may have limited relevance to evening disorder. Safer Streets activity is intended to address public safety during periods and in places where risk is more concentrated. That requires coordination between venue schedules, outreach work, local policing information and transport patterns.

A venue also depends on its surrounding environment. If access is difficult, if young people do not consider the space legitimate, or if the programme is not aligned with local interests, the building becomes an asset without sufficient utilisation. The attendance data suggests that the Haven Centre has achieved a meaningful level of engagement during the period recorded. The next analytical question is whether that engagement is sustained, targeted and connected to wider outcomes.

From attendance to prevention

A robust assessment of a youth facility should examine at least five stages:

1. Access: Can young people reach the facility safely and at relevant times?

2. Engagement: Do they attend once, occasionally or regularly?

3. Retention: Are participants returning over a period long enough for relationships to develop?

4. Progression: Are they moving into education, training, mentoring, health support or other constructive routes?

5. Safety impact: Are there changes in incidents, repeat contacts or reported disorder in the surrounding area?

This sequence prevents a common analytical error: treating a high activity count as a completed prevention outcome.

The opposite error is also possible. A programme may produce few visible incidents of success because its primary effect is avoidance. A young person who does not enter the criminal justice system does not create a case record demonstrating that diversion worked. This makes prevention more difficult to measure than enforcement. It also makes consistent baseline data particularly important.

What the evidence says about anti-social behaviour

Anti-social behaviour is not one uniform category. It includes conduct with different causes, levels of harm and intervention requirements. A group gathering in a public place, repeated nuisance behaviour, criminal damage, threatening conduct and public disorder should not be analysed as though they represent the same problem.

This matters for programme design. Youth diversionary activities may be effective for some forms of low-level or emerging behaviour while having limited relevance to serious violence or organised criminal exploitation. The intervention has to match the risk.

A useful local framework distinguishes between three broad levels:

• Emerging risk: early warning signs, low-level disorder or school and community concerns where universal or targeted activity may be sufficient.
• Established pattern: repeated incidents, poor engagement or multiple agency contacts requiring coordinated case management.
• Serious or acute risk: violence, weapon carrying, exploitation or immediate safeguarding concerns requiring specialist enforcement and protection.

Diversion sits across the first two levels, but it cannot replace decisive safeguarding and enforcement at the third. The strategic challenge is preventing young people from moving from emerging risk into an established pattern, while ensuring that serious harm is not minimised under the language of prevention.

This is another reason the six-times re-offending figure should be used carefully. It supports early intervention. It does not support a blanket policy of diversion regardless of behaviour, harm or risk.

Resource allocation: where the data points

The evidence suggests that North East Lincolnshire’s preventative capacity should be assessed as a network rather than a collection of disconnected projects. The Youth Engagement Alliance, JEFF, The Haven Centre, Clear, Hold, Build and Safer Streets activity address different points in the local system.

The resource allocation questions are therefore operational:

• Are services available in the neighbourhoods where incidents are concentrated?
• Do opening hours correspond with the periods of highest demand?
• Can a young person move from a police or school referral into support without a prolonged delay?
• Are youth workers, mental health staff and youth justice practitioners able to coordinate?
• Is there continuity when a short-term funding stream ends?
• Are repeat participants identified and offered a suitable next step?
• Can the borough distinguish between activity volume and individual outcomes?

The planned £1.5 million annual running cost for Horizon Youth Zone in Grimsby illustrates the scale of investment required to sustain major youth provision. The figure should not be treated as a direct cost-benefit calculation for diversion. It is a planned operating cost, not a quantified saving from reduced offending. Its relevance is that large facilities require long-term funding, governance and utilisation planning.

Short-term projects can produce useful evidence, particularly when they operate in defined locations. But if services are repeatedly commissioned for limited periods, the system risks losing staff, relationships and institutional knowledge between funding cycles. Prevention is cumulative. A disruption in provision may not appear immediately in incident data, but it can reduce the quality of future engagement.

The limitations of current evidence

The local data is sufficient to support a stronger case for prevention, but not to answer every question about effectiveness.

The main limitations are structural:

• The six-times re-offending comparison is presented as a broad finding linked to early prosecution and prevention or early help; it does not provide a complete local cohort analysis.
• Nunsthorpe figures show reductions in reported anti-social behaviour and public order offences, but the available information does not isolate the effect of each component of Clear, Hold, Build.
• Attendance totals do not establish the number of unique participants.
• Long-term re-offending outcomes for North East Lincolnshire diversion participants are not available in the supplied evidence.
• National diversion figures cannot be transferred directly to local performance.
• Changes in reporting behaviour, policing activity, population movement and offence recording may influence incident totals.

These limitations should lead to better measurement, not weaker ambition. A prevention programme that cannot be evaluated perfectly can still be preferable to a formal intervention associated with a much higher risk of re-offending. The correct response is to improve the evidence base while continuing proportionate delivery.

The most useful next step would be a longitudinal local dataset linking intervention entry, attendance, participation intensity, referral source and subsequent justice or safeguarding contact. That work would need clear governance and privacy controls. It would also need to avoid reducing young people to risk scores. Metrics should inform decisions, not replace professional judgement.

A more precise definition of success

For North East Lincolnshire, success should not be defined as the complete absence of youth anti-social behaviour. That is not a credible operational target. Youth behaviour, public disorder and vulnerability are influenced by factors outside the control of any single initiative.

A more realistic definition includes:

• fewer repeat incidents involving the same individuals or locations;
• earlier contact with support before formal prosecution becomes necessary;
• sustained participation in positive activities;
• improved access to mental health and trauma-informed assistance;
• reduced escalation from low-level disorder into more serious offending;
• stronger coordination between statutory and voluntary organisations;
• safer use of public spaces, supported by both enforcement and community provision.

This definition also changes how success is communicated. A fall in incidents is important. So is the number of young people who remain engaged with support. A programme should be judged on whether it changes risk and opportunity over time, not simply whether it produces a high number of sessions.

The strongest evidence is not the largest attendance figure. It is a consistent reduction in escalation, repeat involvement and reliance on formal prosecution.

The likely direction of local policy

The data points towards a preventative system with three defining features.

First, early intervention will remain the preferred response for lower-level and emerging risk where safeguarding allows. The six-times re-offending comparison provides a strong rationale for avoiding unnecessary formal prosecution at the beginning of a young person’s contact with services.

Second, neighbourhood initiatives will increasingly be judged through combined metrics. Nunsthorpe demonstrates the value of measuring both anti-social behaviour and specific offence categories. Future assessments are likely to require clearer separation between enforcement outputs, engagement activity and longer-term behaviour change.

Third, mental health and trauma-informed provision will become more integrated with youth crime prevention. Projects such as JEFF occupy an important position between universal activity and formal justice or clinical intervention. Their effectiveness will depend on referral quality, continuity and the ability to provide a route onward when low-intensity support is insufficient.

The main risk is fragmentation. Multiple programmes can generate impressive totals while leaving gaps between them. A young person may attend an activity but never receive follow-up support. A police referral may not connect efficiently to a community provider. A successful venue may lose its workforce when a funding period closes. These are system failures, not necessarily failures of individual projects.

The preventative framework will therefore need sustained coordination, common definitions and a more mature approach to outcome measurement.

Conclusion

The latest evidence from North East Lincolnshire supports a clear but bounded conclusion: youth diversionary activities are most effective when they intervene early, connect practical engagement with mental health support, and operate as part of a multi-agency neighbourhood strategy.

The strongest figures are substantial. Early formal prosecution is associated with a six-times higher risk of re-offending than prevention and early help. On Nunsthorpe, reported anti-social behaviour fell by 35.1% and public order offences by 53.6% during the measured period under Clear, Hold, Build. More than 11,000 attendances were recorded across local youth engagement networks in 2024–2025, while The Haven Centre recorded more than 1,500 attendances during Safer Streets 5 operations.

None of those numbers should be overstated. They do not prove that diversion eliminates anti-social behaviour, that every attendee is a separate individual, or that one programme explains a borough-wide trend. They do show that prevention now has measurable operational evidence behind it.

The next phase is less about expanding activity for its own sake and more about improving the link between activity, engagement and outcome. If North East Lincolnshire can maintain local access, integrate trauma-informed support and track repeat involvement over time, diversion will remain a credible mechanism for reducing youth reoffending and limiting the escalation of anti-social behaviour.

The strategic direction is established. The future performance of the model will depend on whether its measurement and resource allocation are as disciplined as its underlying premise.