We were pleased to see confirmation last week that the Justice Data Lab (JDL) is now a permanent service offered by the Ministry of Justice. We made the case for this service—which allows organisations to measure the reoffending rates of a group of service users compared to a matched control group—back in 2012 when 84% of charities said they thought it would be useful. But when the pilot started in April 2013 there were fears that if not enough organisations used the service, it would be disbanded—consigned to history as one of those great ideas that just doesn’t get off the ground.
We knew that the publishing of the reports online could deter some organisations from using the service, but brave organisations have stepped forward and it’s opened up an interesting debate about learning, sharing and failure. We’ve supported the criminal justice sector in using the service and credit is due to the MoJ who have engaged with users and adapted the service to their needs, winning a number of awards in the process.
The MoJ has published a summary of the pilot, which details their findings and next steps. It’s great to reflect on the progress made so far, which includes:
- Metrics increased from just a yes/no on reoffending to includes frequency of offending and time to reoffend (and work is taking place to measure severity of offence too).
- A better explanation is being provided to organisations for why some offenders submitted are excluded from the final analysis.
- More up to date reoffending data now available—organisations can submit service users who ended an intervention up to June 2013*.
- Work is being carried out to link reoffending data to other data sets to enable better matching for the control group. If successful, this means organisations working with service users with multiple needs (substance misuse, homelessness etc) can now participate.
Additionally a user feedback survey (n=12 but covers 77% of the 124 published reports) shows that 83% found the Justice Data Lab service to be useful. Although survey respondents were biased towards those that had received a statistically significant result, two organisations that had an inconclusive result also said they found it useful. Even more importantly, 67% of respondents have made changes or improvements to their programme based on their JDL report.
While this is definitely a cause for celebration, our job is not done—we’re still working hard to convince other government departments to be as forward thinking as the MoJ. DWP have made the first initial step towards an employment data lab (we are still looking for organisations to pilot the service, so get in touch if you’re interested), we’re putting together a business case for education (watch out for our survey to be released soon!) and we’ve also been working on healthcare—although it’s proving quite tricky.
Most importantly we still need organisations to make use of the Justice Data Lab, which is still unbelievably free to use! With more government cuts ahead, it will be hard to justify a service which isn’t used, no matter how well intentioned it is—so if you’re working in criminal justice, get involved.
*The nature of the evaluating a 12 month reoffending rate means that data will always need to be retrospective to enable 12 months in which to commit an offence has passed, plus an additional six months for records to be processed through the courts and a further one month for administrative systems to be updated.