Giving prisons more freedom is the right move – but the MoJ needs to sharpen its scrutiny

Written by Eleonora Harwich, Reform on 28 April 2016 in Opinion
Opinion

The government's vision of autonomous prisons held to account for delivering against defined outcomes is welcome – but it will require a step change in the way performance is measured

David Cameron's recent landmark prison reform speech focused on the welcome principles of autonomy, accountability and transparency. These are core themes which have run through the public service reform agendas of both this and the coalition governments. Much of the focus since the speech has been on the new “reform prisons”, modelled on academy schools and run by innovative prison governors. But the more radical announcement was the commitment to introduce prison league tables.

As the prime minister argued, “any modern public service has to be able to demonstrate its value”, yet “we have no idea” which the best performing prisons are. To create the promised league tables, the government will have to overcome several hurdles. 

Firstly, no official prison performance mechanism takes into account reoffending rates. Both the prime minister and justice secretary Michael Gove have spoken of the opportunity prison presents to turn lives around. To, in Cameron’s words, “correct some earlier – often catastrophic – state failure”. This is best captured in a measure of reoffending. Success against this measure also means fewer victims and less demand on criminal justice services – outcomes that citizens care about. If public services should be measured on what matters, evaluations of the prison estate are falling woefully short.


Michael Gove brings in senior education figures to lead prison and probation watchdogs
Civil service to "ban the box" to help rehabilitate ex-convicts, David Cameron announces
Ministry of Justice to hand more power to prison governors as David Cameron vows “wholesale” reform


The prison rating system (produced by the National Offender Management Service) comes closest by including a proxy for reoffending. Their "reoffending" metric is based on the provision of purposeful activities or accredited programmes. This is on the basis that there is some evidence of the efficacy of such programmes in reducing offending, but that prisons cannot be held directly accountable for the reoffending rates of their inmates on release.

"The availability and quality of prison data is inadequate, hindering the analysis of prison performance"

Whilst it is true that the causal relationship between a prison and the offending behaviour of its inmates is hard to establish, government should understand the progress a prison makes in reducing reoffending. In new analysis undertaken by Reform, the reoffending rates of comparable prisons are assessed and found to vary significantly – suggesting that the prison itself can have an effect. As such, Reform recommends that the Ministry of Justice therefore creates a "reoffending progress measure" where predicted reoffending rates for a prison cohort are compared to proven ones. Measures of this type would allow for a fairer comparison.

Secondly, the availability and quality of prison data is inadequate, hindering the analysis of prison performance. For example, a full value for money assessment considers how effectively money is spent to produce the outcomes sought. Yet broken down expenditure data for prisons is not readily available, and the Ministry of Justice does not even hold this data for private prisons. There is therefore no easy way of knowing how much a given prison is spending on, for example, training and industry courses or different types of staff. There is also no way of comparing how prisons spend their budgets. This lack of transparency prevents lessons being learnt from institutions that are making "smart" expenditure decisions. 

Thirdly, the standard metrics used by the Ministry of Justice are too narrowly focused on what can be easily measured. This has hampered the Ministry’s ability to think more innovatively about metrics that could more accurately capture what goes on both within and outside a prison’s walls. Current data only allows for an approximate understanding of a prisoner’s experience. For example, there is no publicly available data on the average number of hours a prisoner spends outside his cell, or how much contact a prisoner has with his family. In addition, data on employment and accommodation is based on offenders self-reporting this information at the point of release. This fails to capture the type of employment and accommodation, nor its sustainability.

Despite these issues, Reform’s new report – Unlocking Prison Performance – experiments with current data to assess how efficient and effective adult male prisons (category B and C) are. The analysis shows a wide variation in performance among prisons across the indicators used for the analysis. While few prisons perform well across both the efficiency and effectiveness measures, lessons can clearly be drawn from the performance of comparable prisons within each of the categories. With continuing pressure on public finances, narrowing this performance gap would deliver better value for money for taxpayers. It would also mean better outcomes for prisoners.

Cameron’s vision of autonomous prisons held to account for delivering against defined outcomes is welcome. High performing prisons should be identified and learnt from, poor performing prisons should be subject to intervention. This will require a step change in the way performance is measured – and a dramatic improvement in data quality. The prime minister has put the prison estate on a challenging path, but the prize for success is improved lives, safer communities and lower costs.
 

About the author

Eleonora Harwich is co-author of Unlocking Prison Performance, published today by the public service think tank Reform

Share this page

Further reading in our policy hubs

Add new comment

Contact the author

The contact details for the Civil Service World editorial team are available on our About Us page.

Related Articles

Related Sponsored Articles