The Scottish Government

« Previous | Contents | Next »

Listen

Section 5
MEASURING, LEARNING AND ACTING

MEASURING, LEARNING AND ACTING

5.1 We need to be much smarter about the way we measure reoffending, the things which have an impact on it and the effectiveness of the different things we do to reduce it. Both policy makers and service deliverers need to learn from and act on this information in a systematic way. We need to use targets intelligently as a way of driving change.

5.2 So an effective system needs a robust framework for quality assurance, performance management and outcome measures. Overall, the data collected should measure whether the new system of offender management is successful in achieving its main objectives and its impact on priorities. Those at the front-line who collect and input information should be able to see how better information connects with improving service delivery, and be motivated to support data integrity. Managers should be able to access data which is as current as possible. Clear and consistent definitions should guide data collection.

Measuring the things which impact on reoffending

5.3 This Strategy sets specific high-level outcomes for offenders, the community and the system, where success in delivery will contribute towards a reduction in reoffending and reduce its impact on communities. We need to be able to tell how successful we are in achieving each of these outcomes.

5.4 We would understand more about the effectiveness of work with offenders if more of the data available came from studies which tracked individuals through the system, rather than from statistical "snapshots". But there will be significant practical challenges in moving in that direction.

5.5 So we need to continue to evaluate our own system, to improve our ability to track the impact of interventions on individuals and to learn from international research. We need to develop the use of common assessment tools and programmes, to make it easier to produce sample sizes from which robust conclusions can be drawn about effectiveness. We need to share the outcomes of research with sentencers and others. There needs to be ways in which those working at the front-line can feedback into research programmes their own views on effective practice. More could be done to seek the views of offenders themselves on which interventions have made the most difference to them.

Measuring reoffending

5.6 For reoffending we have to rely on the proxy measure of reconviction. The measure we use now simply identifies what percentage of offenders are reconvicted of an offence within 2 years of release (prisoners) or 2 years of start of sentence (offenders on community sentences). This does not tell us anything about changes in the seriousness or frequency of reoffending. It also includes reconvictions for offences which happened in the past but have not previously been considered by the courts - and so overstates the level of actual reoffending.

A better measure would:

• Measure the rate of reoffending - i.e. not simply whether someone has or has not reoffended within two years, but how quickly and how often. We already publish figures for this, but it needs to become part of our fundamental measure of performance.
• Measure the seriousness of reoffending - i.e. not simply whether someone has committed an offence, but whether or not the seriousness of their offending is changing. We are not able to tell this from the figures gathered at the moment.
• Make like-for-like comparisons easier. Certain factors - such as age and number of previous convictions - increase the likelihood of reoffending. We need to be able to take this into account in comparing the outcomes of the different sentences and interventions and making comparisons over time. We can learn here from work done in other jurisdictions, including work by the Home Office.
• Exclude reconvictions for offences committed in the past. Summary justice reform is already helping - "rolling up charges" will be far more common in future. But it would be better if the measures we used only ever picked up convictions for new offences.
• Recognise that reconviction figures only capture reoffending which results in a reconviction. We can only measure what we know about. But even allowing for this unavoidable limitation on the data, we need to understand better whether reconviction data may understate the level of reoffending for some offences or types of offender more than others.

Examining how agencies are performing on the ground

5.7 Inspectorates, particularly HMCIP and SWIA, will continue to play a central role in providing assurances about how services are being delivered. Under the Management of Offenders etc. (Scotland) Act 2005, the relevant inspectorates have a new duty of co-operation. Audit Scotland also has an important role to play.

Learning and acting

5.8 Every organisation with responsibility for delivering offender management should be able to demonstrate that it has effective performance management and quality assurance systems in place and that these are used to manage and develop services.

Targets

5.9 No single organisation can reduce reoffending on its own, but many have a contribution to make. This needs to be recognised in setting targets. Equally, not everything we may want to measure in order to improve our understanding needs to have a target associated with it. We will choose targets carefully, involving the NAB and stakeholders in decisions.

5.10 The 2004 Spending Review set a target of a 2% reduction in reoffending by 2008, based on headline reconviction rates. With the help of the NAB, a suite of new targets will be set which will provide a more meaningful incentive to organisations. These will:

• be stretching but realistic
• set reductions in the seriousness and frequency of reoffending as a collective goal
• set improvements in the offender, community and system outcomes as goals, to which a number of organisations will contribute
• set a number of organisation-specific targets which will be agreed with the organisations, consistent with their own performance measurement and management needs.

In the period to April 2008

The NAB will

• Monitor performance against the existing target for a reduction in reoffending.
• Advise on what the priorities should be in developing a smarter set of measurements.
• Advise on how targets should be set for the system in the longer term and what these should be.
• Examine performance data, inspection reports and relevant research, and advise on their policy implications.

The Executive will

• Support work to develop smarter performance measures. As a first step, we will establish a short-life technical group to assess the practical challenges in developing these measures: this will report to the NAB in Autumn 2006.
• With the advice of the NAB and the involvement of stakeholders, set new targets for a reduction in reoffending.
• Continue to develop links with the Home Office and with researchers in other jurisdictions, to learn from work they are doing on achieving better measures of reoffending and system effectiveness.
• Support investment in IT systems which will assist with the collection and analysis of performance data.
• Specify a format for reporting performance information in CJA reports to be submitted for 2007-08 and involve CJAs themselves in deciding the format.
• As part of the 21st Century Social Work Review, develop the new Performance Improvement Framework for CJSW, working closely with stakeholders.
• Continue to develop our understanding of how Scotland compares with other jurisdictions.
• Continue to support a range of research activity.
• Communicate the outcomes of research widely.

CJAs will

• Require local authorities, and through them their contractors, to have in place performance management and quality assurance arrangements which support continuous improvement and quality enhancement as a condition of receiving grant. From 2007-08, where any authorities in an area are unable to demonstrate robust performance management and quality assurance systems, CJAs will require an action plan to be put in place to address this and monitor progress on that plan.
• Undertake detailed analysis of key performance data available for the area, provided to them by the various agencies. They will examine their own area's performance in comparison with that of other CJAs as information becomes available, and work with other CJAs to undertake this in an efficient way, avoiding duplication of tasks. But they will not be expected to undertake detailed direct performance management of services in their areas, or to undertake service inspections. This will remain the role of bodies such as SWIA.
• Include in their area plans a clear strategy for taking forward the statutory duty CJAs have to ensure the sharing of good practice.
• Obtain information on performance in relation to the area plans and inform local authorities, SPS and statutory partners, in order to include information on this in their 2007-08 reports.

Local authorities will

• Ensure they operate robust performance management and quality assurance systems. This is to support continuous improvement and quality enhancement, consistent with the new Performance Improvement Framework being developed under 21st Century Social Work Review (for all social work services, including CJSW).
• Support the collection of robust comparable data.
• Assist the CJA in carrying out its performance monitoring and reporting functions.
• Support the work of the Accreditation Panel and RMA, to assist with the longer term ability to make comparisons about what works.

SPS will

• Provide robust performance information to CJAs.
• Support the collection of robust comparable data.
• Assist the CJA in carrying out its performance monitoring and reporting functions.
• Support the work of the Accreditation Panel and RMA, to assist with the longer term ability to make comparisons about what works.

Partner Bodies will

• Be asked to examine how their performance management systems can provide reassurances about their particular contribution to delivery of the area plan.
• Support the collection of robust comparable data.
• Provide information to assist the CJA in carrying out its performance monitoring and reporting functions where relevant.
• Where they are delivering offender management services under contract to local authorities, follow the list of activities for local authorities above.

Inspectorates and other independent and advisory bodies will support these changes:

• Through the provision of timely information on and analysis of current practice.
• Relevant inspectorates will continue to provide advice on the state of offender management services in prisons and in the community through routine and special inspections.
• Inspectorates will develop their plans for closer co-operation under the new provisions of the Management of Offenders etc. (Scotland) Act 2005.

« Previous | Contents | Next »