Talking tough (but intelligent) on crime

New research, undertaken for the Ministry of Justice by Helen Bewley at NIESR, shows that increasing the punitive element of community orders, as proposed yesterday by the Prime Minister, has the potential to reduce re-offending. But there are potential pitfalls: reform needs to be cautious and based on the evidence of what works in reducing re-offending.  

[Helen is one of the leading experts in the complex statistical techniques required to establish the causal connection between particular policy interventions (like community orders) and economic and social outcomes (like reoffending). This post describes the research and its implications for policy, but he techniques and data used here have implications for policy analysis going well beyond this particular topic] 

Talking tough (but intelligent..) on crime

Speaking yesterday, David Cameron called for ‘proportionate, meaningful punishment’, with a combination of tougher sentences and a greater emphasis on the rehabilitation of offenders. This approach was clear in the measures recently proposed by the government to reform community sentences, published under the previous Justice Secretary, Kenneth Clarke. The main thrust of the proposals is to ensure that all community orders include a clear punitive element, except where there is a high risk that this would be counter-productive. This pledge was recently reiterated by the current Justice Secretary, Chris Grayling, in his speech to the Conservative Party Conference.

Will it work? 

Should the government implement the existing proposals, in future most community orders will include unpaid work, a curfew, exclusion or prohibited activity, a fine, or other new forms of punishment, alongside rehabilitative elements. This compares with the current situation where only around two-fifths of community orders include a requirement to do unpaid work - by far the most common punitive requirement. While there may be other motivations for these proposals - punishing offenders may have popular appeal, but the government’s consultation also mentioned crime reduction; reform and rehabilitation of offenders; protection of the public; and reparation - the key test is surely whether it reduces the likelihood of re-offending

New research undertaken for the Ministry of Justice by Helen Bewley at NIESR, and published today, gives grounds for cautious optimism. It suggests that, if implemented, the proposals to increase the punitive element of community orders could potentially reduce re-offending. Indeed, adding a punitive element to some packages of requirements can actually reduce the number of crimes committed in the two-year period following the initial offence. Adding a punitive requirement (unpaid work or a curfew) to a supervision requirement reduced the number of re-offences committed over the two years following the start of the order by 7.5 per cent. This positive effect also worked the other way; adding a supervision requirement to a punitive requirement reduced the number of re-offences committed by 8.7 per cent.

This analysis is based on a detailed analysis of a very rich dataset on offender characteristics and outcomes held by the Ministry of Justice. NIESR's research compared re-offending rates for offenders given different combinations of requirements, but with the same characteristics. These characteristics capture both the combination of requirements to which the offender is assigned and many other characteristics which are known to be related to re-offending behaviour.More broadly, the impact of imposing a punitive element on a wider circle of offenders sentenced to a community order is uncertain; there is no guarantee that punitive requirements will have a similar impact on re-offending behaviour for offenders with different characteristics. Also, increasing the number of requirements placed on offenders has resource implications. The effectiveness of requirements may fall if resources are not increased proportionately. Finally, impacts were only considered over a two-year period in the NIESR research, so longer-term effects are not yet known.


Custody versus community orders

 Past research has demonstrated that community orders are more effective than short custodial sentences in reducing re-offending. If the government was to move away from the use of community sentences towards custody, there is a danger that re-offending rates will rise. NIESR’s research did not raise any serious grounds for concern in the proposals to enhance the effectiveness of community sentences, but a more draconian approach which involved exchanging community orders for custodial sentences would run greater risk of ultimately being counter-productive.

Grounds for caution

The evidence suggests that the government’s proposals for the reform of community sentences are consistent with the rhetoric of ‘tough but intelligent’. However, it also shows that it is important to ensure that punitive requirements are used in combination with other requirements which can be used to enhance their effectiveness, such as supervision and programme requirements.