A multi-force study spanning more than 62,000 criminal incidents across 13 English police forces over four years has found that police drug diversion schemes are associated with reduced reoffending, while flagging implementation fidelity as the variable most likely to determine whether a scheme delivers or disappoints, according to research published by PMC/NCBI in November 2023.

The study — a realist evaluation coproduced by academic, policing, health, and service teams — is among the most systematically grounded assessments of diversion practice in England to date. Realist evaluation, for readers unfamiliar with the framework, moves beyond asking "does it work?" to ask "what works, for whom, in which conditions, and why?" That design choice matters here, because drug diversion is not a single intervention. It ranges from out-of-court disposals such as cannabis warnings and conditional cautions, to structured treatment referrals and educational programmes. Collapsing them into one category, as earlier outcome studies often did, obscured more than it revealed.

What the Evidence Shows

The headline finding — reduced reoffending — is consistent with earlier UCL-based research and with analysis by Blais et al., published via the White Rose repository, which concluded that police-based diversion measures show promising results for health, wellbeing, and social circumstances, in addition to recidivism outcomes. The White Rose work also found that fidelity to intended scheme design correlates with effectiveness — a finding now reinforced at much larger scale by the 13-force dataset.

That convergence across independent research streams is analytically significant. It suggests the reoffending signal is not an artefact of a single methodology or study population, but holds across different force areas, delivery models, and evaluation approaches.

The fidelity question, however, is not a minor implementation footnote. In the diversion literature, low-fidelity delivery — officers applying schemes inconsistently, failing to complete referrals, or using diversion tools as informal decriminalisation rather than structured health pathways — has been documented as a primary mechanism by which promising interventions fail in practice. The 13-force study's scale gives that finding empirical weight that smaller pilots could not.

The Stop and Search Problem

Running alongside these findings is a structural integrity concern. In 2021, HMICFRS — His Majesty's Inspectorate of Constabulary and Fire & Rescue Services — found that nearly one in five stop and search operations for drugs had no reasonable grounds. That figure bears directly on diversion because the lawfulness of the initial encounter determines whether any subsequent disposal is procedurally sound. Diversion that flows from an unlawful stop does not merely raise human rights concerns; it undermines the programme's legitimacy with the communities it is intended to serve — typically the same communities most exposed to discretionary policing.

This is not a new tension in British drug enforcement. What the current research does is sharpen it: if diversion is expanding as an alternative to prosecution, the quality of the upstream contact — the stop, the search, the grounds recorded — becomes more consequential, not less.

Youth Outcomes and the Charge Escalation Risk

Kent researchers have documented an association between previous charges and higher reoffending rates among young people, a finding that strengthens the case for pre-charge diversion as a specific priority. Formal charging carries a labelling effect that can entrench criminal identity and constrain employment, housing, and social support — precisely the factors protective against reoffending. Diversion that genuinely redirects a young person away from that pathway has compounding downstream value.

The policy logic here is straightforward. Earlier formal contact with the criminal justice system predicts worse long-term trajectories. Diversion interrupts that contact before it crystallises into a record. But the Kent evidence also implicitly demands that diversion schemes are resourced well enough to make the health and social referral meaningful — a warm handover to a functioning service, not a leaflet and a caution.

What Comes Next

The 13-force study's coproduced design — drawing in policing, health, and service partners from the outset — is itself a methodological signal about what good evaluation looks like in this field. It allows researchers to examine not just outcomes but the mechanisms generating them, which is the information practitioners and commissioners actually need.

The outstanding questions are operational. Which scheme elements drive the reoffending reduction: the diversion decision itself, the quality of the referral, or the service on the other end? Do effects hold equally across demographic groups, or does the well-documented racial skew in stop and search create differential access to diversion's benefits? And at what threshold of implementation fidelity does a scheme stop being clinically meaningful and become, in effect, an unmonitored caution with extra paperwork?

Those questions are answerable with the data that now exists. Whether commissioners and police and crime panels have the appetite to commission that analysis is a different matter.