Secondary Data Analysis
The aim of this research is to examine which children and young people are diverted from the criminal justice system in London and the relationship between their use and reoffending.
Diversion is an approach aimed at preventing reoffending by providing alternatives to formal criminal justice proceedings, typically used with first-time and less serious offending. It has been shown to be effective. However, there is limited evidence in England and Wales about their use and the relationship to reoffending amongst young people here.
What are the main research questions?
This project explored the characteristics of children and young people (CYP) diverted from the criminal justice system in London, the variation in diversion for CYP across London and whether there’s a relationship between receiving a diversion and reoffending.
The main questions this project set out to answer, are:
What did the analysis involve?
The analysis looks at patterns across all offending, violent offences and knife-related offences. It uses data from 2015 to 2022 from the London Metropolitan Police Service (MPS) on reported crimes, stop and searches, and arrests involving CYP. The total dataset includes records of 264,932 CYP who interacted with the police.
Please refer to the analytical protocol and final report for more details.
Key findings
| Seriousness and frequency of offending explain some but not all of the differences in which CYP are diverted | |
| The majority of CYP in London receiving a formal disposal are now diverted v | Between 2015 and 2022, the number of CYP in contact with the police fell by 17%. The proportion receiving a diversion or escalation (e.g. sentenced or charged) also fell – down from 63% to 48%, driven by an increase in victims being unwilling to prosecute. Of CYP diverted or escalated, the majority are now diverted (54% in 2022). Escalations fell from 31% in 2020 to 22% in 2022. |
| CYP involved in serious and prolific offending are diverted less often, but there are outliers | 45% of boys are diverted vs 67% of girls. 88% of 10-year-olds are diverted vs 32% of 17-year-olds. 55% of White CYP are diverted vs 42% of Black CYP. 51% of CYP with no prior arrests are diverted vs 41% with four prior arrests. 35% of CYP with five or more prior arrests are diverted, yet 48% of lower-harm offences led to a diversion vs 38% of moderate-harm offences and 17% of high-harm offences. |
| Black CYP are diverted less often, even when controlling for the seriousness of offending | After controlling for the seriousness and frequency of offending, Black CYP are still diverted less often than White CYP. These factors, plus their broader characteristics (e.g. age and gender), explain around half the difference in diversion rates with White CYP (falling from 16%pts to 9%pts1). |
| There are also unexplained differences in the rate of diversion across London boroughs | |
| CYP in outer London have the highest rates of diversion | The raw difference between the highest and lowest rates of diversion across boroughs was 24%pts (66% compared to 41%). Outer London boroughs, particularly Bromley (66%), Bexley (65%) and Kingston Upon Thames (65%), have the highest diversion rates. Inner London boroughs, such as Haringey (43%), Lambeth (42%) and Hackney (41%), show the lowest. |
| Controlling for the profile of who offends and the type of offending explains some but not all of these differences | After controlling for individual- and offence-level factors, the spread between the highest and lowest boroughs falls from 26%pts2 to 22%pts. Bromley (61%), Bexley (60%), Hammersmith and Fulham (59%), Merton (58%) and Kingston Upon Thames (58%) had significantly higher diversion rates compared to Newham (49%) – the borough with the median rate of diversion. Harrow (39%), Tower Hamlets (42%) and Greenwich (42%) were significantly lower. |
| Children and young people diverted are less likely to reoffend | |
| Controlling for individual and offence level factors, diversion is associated with lower reoffending | Raw differences show diverted CYP are less likely to be arrested than those escalated: 3% vs 8.% after six months and 4% vs 12% after 12 months. Controlling for individual- and offence-level factors, diverted CYP are still less likely to be arrested: 4%pts at six months and 5%pts at 12 months. |
| Diversion may be more beneficial in reducing violent reoffending and for Black CYP | When controlling for individual- and offence-level factors, the reduction in arrests appears greater for violent offences compared to any offence with –8%pts at 12 months (vs 5%pts for any offence). Diversion is associated with a 6%pt reduction in arrests at six months for Black CYPs compared to 5%pts for White CYP. |
| The apparent reduction in reoffending may be due to study limitations | It’s not possible to fully control for all the factors that explain why diverted CYP are less likely to reoffend. There may be contextual factors not captured in the data but considered at the time a decision is taken over whether to divert. |
| CYP involved in knife offending are less likely to be diverted than those involved in other types of offending. Diversion for knife offences is associated with reduced reoffending | |
| Most CYP involved in knife offences have no prior arrests | CYP involved in knife offences fell by 32% between 2015 and 2021. Most CYP (68%) had no previous arrests for any offence. 13% had three or more previous arrests for knife offences. Older CYP are overrepresented (70% aged 15-17), as are Black CYP (51%) – twice as likely relative to Black CYPs’ share of the population. |
| CYP involved in knife crime are less likely to be diverted compared to those involved in any offending | CYP involved in knife offences are less likely to be diverted, accounting for 17% of outcomes on average between 2015 and 2022 compared to 26% for all offences. They were more likely to be escalated (58% 2015–2022 vs 29% for all offences). Black CYP have the lowest rate of diversion compared to escalated outcomes for knife offences across ethnic groups (17.2% vs 35% for White and Asian CYP). |
| Diversion for knife offences is associated with lower reoffending | CYP diverted for knife-related offences are less likely to be arrested (when controlling for individual- and offence-level factors) by 9%pts at six months. At 12 months, the reduction in reoffending was 8%pts, although this wasn’t significant. |
Interpretation and implications
This research finds that the number of CYP receiving any formal disposal (diversion or escalation) has fallen in London since 2015 due to a growing number of victims being unwilling to prosecute. Among formal disposals, diversion is being used more frequently, particularly post-Covid-19. On the whole, when used, diversions are being targeted at first-time and less serious offending, yet there remain unexplained differences in who receives them. A minority of persistent offenders and those involved in more serious and violent crimes are diverted. Black children are significantly less likely to receive them, even after controlling for the types of offences they’ve been involved with. And there are areas of London that experience higher or lower diversion rates than we might otherwise expect.
The analysis finds tentative evidence that receiving a diversion may reduce the likelihood of reoffending, particularly violent reoffending and for Black CYP. This study reinforces the findings from the wider literature on the potential benefits of diversion and extends it by using recent data on a large UK population. However, more work is needed to establish whether the relationship between diversion and reoffending is causal. It’s important to bear in mind that this analysis was limited by the data that was available. It’s not possible to observe all the contextual and offence-level information that would have been considered when deciding whether to divert someone. These findings are also specific to the application of diversions in London. More work is needed on how diversions are used in other parts of the country.
Please refer to the final report for more detail on these findings, how to interpret them, and their implications.