Select Committee on Public Accounts Forty-Ninth Report

2 Measuring the performance of prisons

The performance of PFI prisons

12. The performance of the seven PFI prisons against contract has been mixed. All but one have incurred financial deductions for poor performance, although the level of financial deductions in themselves do not provide a full picture of performance in a prison. In most cases, the financial deductions have tended to be highest in the first year of operation and generally reduced in the following years. The Home Office intends that in future there will be a much greater link between financial deductions and the performance of the prison.[12]

13. Comparing the performance of PFI prisons against publicly-managed prisons is difficult because of variations in their age, design and function and because of the different ways in which their performance is measured and targets are set. PFI prisons tend to perform better than public prisons in relation to the Prison Service's decency agenda, such as respect shown to prisoners, and the availability of purposeful activities for prisoners. They tend to perform less well in areas such as safety and security. Indeed, it is unusual for any prison, whether privately or publicly managed, to perform equally well on both counts, indicating that there is a difficult balance to be struck between the two areas of work.[13]

14. One of the seven PFI prisons, Ashfield Young Offenders Institution, which was opened in 1999, has performed significantly worse than the other six. The Prison Service became so concerned about safety at Ashfield that it put in its own management team for five months in 2002 and moved the young offender population to other establishments. In October 2002, control of Ashfield was returned to the contractor, Premier Custodial Group Ltd, after conditions improved, although an improvement was to be expected as the institution was half empty of prisoners but operating with a full complement of staff.[14]

15. Premier was in breach of its contract for Ashfield on several occasions. Rather than terminate the contract, the Prison Service had opted to put in public service management to improve the prison's performance. Under the contract, Premier had incurred financial deductions for poor performance and non-availability when young offenders were moved to other institutions. Premier's income was halved but it still had to maintain a full complement of staff, the cost to Premier amounting to £4.2 million. However, it was not only the contractor who suffered losses. Some of the risk of poor performance was borne by the public sector as young offenders had to be redirected to other institutions which were already overcrowded. The Prison Service has made clear to Premier that the prospect of contract termination remains if the improvement in performance is not sustained.[15]

The use of performance measures and targets

16. Public prisons are measured against 48 Key Performance Targets and have to comply with 61 Prison Service Standards (Figure 4). Her Majesty's Chief Inspector of Prisons has also set out 4 key constituents of a healthy prison. PFI and privately-managed prisons typically have to meet 30 to 40 performance measures set out in their contracts. They also have to comply with the 61 Prison Service Standards and their contracts are being amended to incorporate the relevant Key Performance Targets. Many of the targets and measures overlap and there are a number of other areas where the details of the measures are different or where different measures are used to assess performance against the same standard .[16]

17. There are inconsistencies between the targets set for PFI and public prisons. For example, the average target for purposeful activity in a public prison is 20.6 hours a week per prisoner, as opposed to 29.5 hours a week in PFI prisons. The Prison Service said that it had sought higher levels of performance in the early PFI contracts to get better value for money and to create a benchmark which would help to push up performance in the public sector. The PFI prisons had largely succeeded in meeting these targets, and most public prisons were also improving. But the Home Office recognised that prisons should be concentrating much more on those aspects of purposeful activity known or believed to reduce re-offending, such as education and offending behaviour programmes.[17]

18. The Home Office and contractors agreed that the large number of performance measures overburdened prisons, making it difficult for managers to prioritise between targets, and to monitor performance against every indicator accurately. Targets needed to be consistent if meaningful comparisons were to be drawn. Contractors were also concerned that some of the targets were input-based, for example, measuring compliance with the Prison Service Standards to prevent self-harm by prisoners rather than monitoring actual incidences of self harm. This practice stifled the potential for innovation, which was meant to be a key benefit of the PFI. In the most recent contracts for PFI prisons under construction at Ashford and Peterborough, two-thirds of the performance targets were to be measured yearly, rather than monthly, in order to reduce the reporting burden. PFI contracts were also being amended to bring them into line with current Prison Service priorities to increase educational activities in prisons, such as offending behaviour programmes, and to assist with the resettlement of prisoners prior to their release.[18]

Figure 4: Performance measures in a PFI prison

Source: C&AG's Report

19. The Prison Service has Service Level Agreements with five public prisons, which set out the resources to be provided and specify the required level of performance in terms similar to a contract with a PFI prison. The Home Office said that the intention was to move public prisons on to a series of such agreements so that there was a similar contractual arrangement with public prisons as with privately-managed prisons. Two of the existing Service Level Agreements were the result of a performance testing regime to identify failing prisons. A failing prison has to produce an action plan to improve, which is then implemented through a Service Level Agreement. If the plan is not acceptable, or is not fulfilled, the prison is contracted out to the private sector without an in-house bid. Two further prisons—Liverpool and Dartmoor—have now been selected to go through this process. They will face the sanction of being contracted out unless they significantly improve their performance within six months.[19]

20. The Prison Service is also developing a quarterly system of ranking prisons, known as the Weighted Scorecard. It scores an individual prison's performance against its targets, its previous performance and the performance of other prisons in the same category. Individual targets are weighted according to the type of prison.[20]

Monitoring prison performance

21. Unlike public prisons, the performance of PFI prisons is monitored by an on-site Controller from the Prison Service. In the past, there have been different approaches to the role, with some Controllers becoming too close to the contractor, whilst others were adversarial. The Home Office agreed that there had been inconsistencies. The Controller's role would be simplified to focus purely on contract monitoring. All Controllers would be line managed by one person, a senior member of the Home Office, who would be developing training to ensure that there was a consistent approach to assessing a prison's performance.[21]

12   C&AG's Report, para 5; Q 67 Back

13   C&AG's Report, paras 8, 2.16; Q 87 Back

14   C&AG's Report, para 1.15, Appendix 2 Back

15   Qq 9-10, 126, 129, 145-157; C&AG's Report, para 5 Back

16   C&AG's Report, Fig 1, paras 2.2-2.3  Back

17   Qq 3-6, 83-88  Back

18   Qq 1-2, 24, 111-116  Back

19   Qq 8, 16, 81, 106 Back

20   C&AG's Report, para 2.4 Back

21   ibid, para 1.23; Qq 79, 93-94, 130-131 Back

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2003
Prepared 2 December 2003