some default text...
Select Committee on Public Accounts Minutes of Evidence


Supplementary memorandum submitted by HM Prison Service

Questions 27 and 28: How many prisoners completing accredited offending behaviour programmes still have two or more years to serve?

We have been able to make some estimates for the cognitive skills and CALM programmes. The issue is complicated by the fact that release dates are variable, because prisoners will get varying periods of parole, so in general we have used parole eligibility date as the point of reference. The data do not include life-sentenced prisoners, because these have no fixed parole eligibility date and there is no comparable reference point that we can use. But subject to those caveats, evidenced estimates are that, in 2000-01, the numbers are:

    —  588 (12 per cent) of 4,974 prisoners that took a cognitive skills course;

    —  52 (28 per cent) of the 184 of the CALM graduates for whom a parole eligibility date is available in the data we have.

We do not have equivalent data on the number of prisoners who completed the Sex Offender Treatment Programme or the Cognitive Self Change Programme in these circumstances. The Sex Offender Treatment Programme is much smaller, and there are many more variables that come into play when making such judgements. For example, the programme is in fact a number of component programmes; and the index offence of those who undergo it may not be an overtly sexual one - it could be murder, say, or manslaughter.

The Cognitive Self Change Programme is also a small programme and the nature of it means that, at present collecting such data would not benefit its effective delivery or development. We will be monitoring this issue and considering whether we should start collecting it in respect of those programmes.

There are in any event a number of reasons why we do not see early completion of a programme as a bad thing:

    —  the prisoner may be doing a cognitive skills programme as a foundation for other work: for example, it is a distinct advantage if a sex offender does a cognitive skills programme first, because this is likely to make him more receptive to the very difficult and demanding nature of the Sex Offender Treatment Programme; and prisoners who undergo the core Sex Offender Treatment Programme may need to go on to the extended programme;

    —  doing a programme early gives the prisoner the opportunity to practice the skills he has learned in the stable conditions that prison provides;

    —  the Sex Offender Treatment Programme has now, and the cognitive skills programmes should have soon, a follow-up booster programme;

    —  undergoing a programme generally has a beneficial effect on the prisoner's general self-awareness and behaviour in prison, so doing a programme early can contribute importantly to the prisoner's general development and to good order and discipline in the prison.

Question 50: Prisons with newly introduced drug treatment programmes?

In recognition of the need to expand the number of places on rehabilitation programmes in the juvenile estate, funding was reallocated to three Young Offender Institutions (YOIs). Delays in recruiting staff (due to a shortage in the market) and the transfer of one prison from one Area to another has meant that only one of the three YOIs has the proposed drug rehabilitation programme running.

The details are as follows:

    Glen Parva—the programme based on the Phoenix House model began on 11 February. The programme provides places for ten prisoners and lasts for eight weeks. The programme will run six times a year and the annual number of entrants will be 60.

    Ashfield—responsibility for the prison has been transferred from the South West Area to the Juvenile Operational Management Group and the proposed new drug rehabilitation programme is currently on hold.

    Bullwood Hall—the programme will not be operational this financial year. The recruitment process for drug workers is underway. The annual number of entrants to the programme is expected to be 50.

Question 59: Cost of renegotiating contracts at contracted out prisons to effect regime changes?

There have been two occasions when a major change of regime at a private prison has meant additional cost to the Prison Service:

At HM Prison and Young Offender Institution Ashfield, where the juvenile population was increased in 1999 from 80 to 300 and the regime had to be changed so that it was compliant with Prison Service requirements for juveniles. The cost is £1,059,532 per year at current prices; and

At HM Prison and Young Offender Institution Parc, where new separate provision for 28 juvenile offenders has been agreed with effect from 4 March 2002 at an annual cost of £1,051,111 and start-up costs of £363,975. This is in response to the growing number of juvenile offenders in Wales.

Other, smaller changes to regime have been generally cost neutral. For example, at HM Prison Rye Hill, where in exchange for a reduction in the unattainable target of purposeful activity of 35 hours of purposeful activity per prisoner per week to 32 hours, the prison has agreed to increase literacy and numeracy accreditations and key work skills.

I mentioned HM Prison Lowdham Grange in my evidence to the Committee. Although there have been changes to the regime, it has not been necessary to change the contract to reflect this.

Questions 60 and 61: Purposeful activity figures for HM Prisons Dovegate, Forest Bank and Rye Hill?

Purposeful activity performance for Rye Hill for January to March 2001 was 26 hours per prisoner per week. For the year to date (April 2001 to January 2002) it currently stands at 29.3 hours per prisoner per week. This is provisional and will be finalised at the end of the year.

Dovegate opened in July 2001 and will be excluded from this year's (2000-01) Annual Report. Year to date performance (July 2001 to January 2002) is 28.7 hours per prisoner per week.

Appendix 4 of the NAO report gives the 2000-01 figure for Forest Bank—27.5 hours per prisoner per week.

Questions 94-97: Extent of the use of volunteers in helping prisoners with literacy and numeracy?

At present there is very little involvement by voluntary groups in support for numeracy, but there are a number of literacy and family learning projects operating across the prison estate. These are helping prisoners to improve their own literacy or reading skills as well as to support their children's learning development. These projects depend for their success on the support of voluntary organisations such as the National Literacy Trust, Safe Ground, the Federation of Prisoners' Families Support Groups, Parents in Prison and the Writers in Prison Network Ltd. The Prisoners' Learning and Skills Unit conducted a survey last year which identified over 40 prisons that were running family literacy or family learning projects.

The Prison Reform Trust has recently carried out a survey of volunteering by prisoners which show that at least 3,000 prisoners volunteer. Some of these are involved as "study buddies" or peer partners and offer reading support to other prisoners. Peer partnership schemes operating currently have been very successful and we would like to see this kind of good practice replicated more widely.

A survey carried out in September 2000 for the Home Office's Active Community Unit suggested that there are at least 4,000 volunteers working in prisons. We suspect the actual number is substantially higher. This figure includes 130 volunteers involved in vocational education (mainly as mentors) and 49 in health education. It does not include 1,400 prison visitors and 1,700 Board of Visitors members or volunteers based in visitor centres outside prisons. A recent survey carried out by the Prison Service Chaplaincy showed that there are 6,000 chaplaincy volunteers.

The Prison Service has been successful in securing matched funding from the Active Community Settlement for projects which will increase the numbers of volunteers by about 1,000 over three years. Total funding for the period 2001-04 stands at £892,000.

Question 163: Prisoners unable to complete programmes or courses due to transfers to other prisons?

Information about the number of prisoners forced to drop out in the middle of an offending behaviour programme, due to a transfer to another establishment, is being collated and I will write again with the data. (Subsequent information is at Annex A).

A prison would, however, be penalised at the annual audit of programme implementation for transferring programme participants for non-clinical or non-security considerations so I am confident that, in practice, the number of prisoners forced to drop out as a result of transfer to another establishment is small.

There is no centrally held information on prisoners forced to drop out of drug treatment programmes due to transfers. However, Area Drug Co-ordinators would be aware of any significant problems in their establishments and no such problems have been reported to the Drug Strategy Unit. The decision to transfer a prisoner to another establishment is an operational one but we would expect every effort to be made for those prisoners on a drug treatment programme to complete the programme before being transferred. The transfer of prisoners mid-way through a programme is inadvisable given the importance of completing the sequential content of the programme and the therapeutic benefits of staying within the same group. In addition, a transfer may cause problems because of differences in content from one programme to the next. So although we recognise the importance of maintaining continuity of treatment, regrettably, there will be occasions when the operational need for transfers will be of over-riding importance.

The Prison Service does not collect information centrally on the participation and achievement of individuals in learning, only at establishment level. Nevertheless, we recognise that the issue of prisoners being unable to complete their course because of transfer to another establishment is one that we need to address. Many establishments try hard to retain those prisoners who are working towards qualifications until they have been able to complete their course. For example, HMP Birmingham aims to retain as many prisoners as possible who are following basic and key skills courses, and HMP Leeds aims to retain prisoners undertaking hairdressing qualifications. We also know that the timing of tests at the end of courses are often not flexible enough to accommodate prisoners, and are working with examination awarding bodies to address this. I am aware of isolated examples of establishments arranging transfers to ensure that prisoners will be able to continue their studies, and am keen to encourage all prisons to follow this example where practicable.

Question 187: Reconviction rates—prisoners not reconvicted?

I am attaching below some key statistics on released prisoners not reconvicted within two years. It should be noted that these are essentially the reverse side of the coin from the reconviction statistics.

As I indicated at the hearing, it is important to recognise that reconviction rates are only one measure of reoffending. While some ex-prisoners reoffend and are not caught, others get caught and convicted for what may be a very trivial offence relative to the one they were originally imprisoned for. So offending may be reduced in that the seriousness of offence is reduced.

Moreover, reconviction rates give no indication of rates of offending, which may also be significantly reduced. So, for instance, an offender who may have been committing several offences per week before custody, may be released and commit one offence in the first year and be reconvicted. Reoffending is reduced, but not reconviction results.

Offender Behaviour Programmes generate reductions in reconviction which can be as high as 15 per cent in some cases. This shows that where appropriate funding and resources are available, the Prison Service can and does achieve significant reductions in offending with some of the highest risk offenders.


In 1997 42 per cent of persons discharged from custody were not reconvicted within 2 years. The "success" rate varies by type of offence, length of sentence, age, gender and whether a person has any previous convictions.

For males, 41 per cent are not reconvicted; for females it is 49 per cent; for young offenders it is 25 per cent; for adults 46 per cent.


Offence categor within 2 yearsy
Percentage not reconvicted
Violence against the person
53 per cent
Sexual offences
82 per cent
22 per cent
43 per cent
27 per cent
70 per cent
Drug offences
64 per cent


Sentence length
Males—percentage not reconvicted
Females—percentage not reconvicted
Upto and including 12 months
39 per cent
44 per cent
Over 12 months upto and including 4 years
44 per cent
65 per cent
Over 4 years upto and including 10 years
69 per cent
79 per cent
Over 10 years not including life
77 per cent
100 per cent

  By examining the number of previous convictions, non-reconviction rates were higher for those whom a custodial sentence was the encounter with the Criminal justice system.

Prisoners released and not reconvicted within 2 years
Number of previous convictions
1 or 2
11 or more
Males aged under 17
Males aged 18-20
Males aged 21-24
Males aged 25-34
Males aged 35 and over
All males
All females
All prisoners


Reconviction is more likely to occur closer to discharge than later on.

When considering all prisoners discharged from custody, 89 per cent will not have been reconvicted within 3 months of release, falling to 57 per cent one year after release, and 42 per cent two years after release. Seven years after release 27 per cent will not have been reconvicted. It is not unreasonable to assume that the effects of any sentence of imprisonment diminish over time.

Question 187: Prisoners with evidence of mental ill health?

A survey of mental ill health in the prison population of England and Wales, undertaken in 1997 by the Office for National Statistics, showed that some 10 per cent of remanded men, 7 per cent of sentenced men and 14 per cent of all women prisoners had suffered from a functional psychosis in the past year. Some 59 per cent of remanded men, 76 per cent remanded women, 40 per cent of sentenced men and 63 per cent of sentenced women had a neurotic disorder. Applying these rates to the current population would indicate that, at any one time, around 5,300 prisoners would be suffering from a functional psychosis and around 30,000 from a neurotic disorder. Some prisoners will have both conditions, which taken together constitute the mental illness element of the survey.

The ONS survey also reported that only one in ten prisoners or fewer showed no evidence of any of the five disorders considered in the survey (personality disorder, psychosis, neurosis, alcohol misuse, and drug dependence). That would indicate that, at any one time, upwards of 60,000 prisoners would have a mental health problem of some kind.

Statistical returns from Prison Service establishments indicate that at the end of June 2001 128 prisoners had been accepted for admission to hospital for in-patient treatment for mental disorder and were awaiting a place. A further 175 were either awaiting assessment or, having been assessed, were awaiting a decision concerning acceptance. At the end of December, the respective figures were 121 and 132.

Question 196: Details on whether education spending has kept pace with the increase in the prison population?

Prior to 2001-02, levels of expenditure at individual prisons reflected differing levels of commitment to education by prison governors who were, in the absence of an allocation methodology, able to cut education budgets to create efficiency savings. In April 2001 prison education was ringfenced and is now administered jointly by the Home Office and Department for Education and Skills. Overall funding is set to increase from £58 million this year to nearly £70 million by 2003-04—a real terms increase of 15 per cent. Education funding continues to be ringfenced once transferred to establishments and cannot now be spent on other services.

The following table sets out education expenditure and the prison population over the last ten financial years.

Education spending
Prison population
£29.372 million
£31.227 million
£31.089 million
£36.756 million
£39.965 million
£34.507 million
£36.174 million
£39.049 million
£47.450 million
£50.890 million

There is some difficulty in comparing like for like each year due to the accounting procedures then relevant for VAT. We believe 96-97, 97-98 and 98-99 years are net of VAT.

Question 196: Cost benefit analysis?

During the 2000 Spending Review the first proper cost benefit analysis of programmes to reduce re-offending were made. Though to an extent crude and limited by the available data, these represent a significant step forward in enabling judgements to be made about the effectiveness of investment.

The analysis was limited to those programmes for which the Prison Service was seeking additional funding and for which reasonable estimates of the impact on re-offending were possible. The table below sets out the details of the calculations in respect of Drug Treatment Programmes, for the teaching of basic literacy and numeracy, and for medium-intensity Offending Behaviour Programmes.

It is important to note that the calculations are based on the best evidence available at the time. As programme evaluations are completed, more accurate, programme-specific calculations will be made. A forthcoming study of one resettlement programme, the High Intensity Treatment Programme at Thorn Cross, will give details of the cost effectiveness of that programme, based on the methodology set out below but applied to a more detailed breakdown of performance.



The costs of programmes were calculated based on the extra resources identified by Prison Service during the Spending Review as required to deliver the outputs proposed.


The benefits were calculated based solely on benefits attributable to reductions in re-offending (including social and psychological costs). Benefits arising from other outcomes, such as for instance greater prosperity due to the increased employability of ex-offenders, were not calculated.

The calculations were based on:

    —  the output targets proposed;

    —  an estimate of the reduction in reconvictions associated with those outputs (based on Home Office researchers' advice on the best available international evidence of average programme effectiveness for each type of intervention);

    —  an estimate of underlying offending rates, and hence of offences avoided for each reconviction "saved" (based on estimates of general offending levels, and on data on types of offending and frequency of reconviction for different prisoner populations)

    —  estimates of the costs of each type of offence (based on RDS work which led to the later publication Economic Cost of Crime, Home Office Research Study 217 (Brand et al., 2000), taking into account financial and non-financial costs to the community, as well as costs in the Criminal Justice System).

previous page contents next page

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

© Parliamentary copyright 2002
Prepared 5 September 2002