§ Mr. George HowarthTo ask the Secretary of State for the Home Department what is the estimated cost of implementing the findings of the review of the open prison estate. [35865]
§ Miss WiddecombeResponsibility for this matter has been delegated to the Director General of the Prison Service, who has been asked to arrange for a reply to be given.
145WLetter from Richard Tilt to Mr. George Howarth, dated 9 July 1996:
The Home Secretary has asked me to reply to your recent Question about the cost of implementing the findings of the review of the open prison estate.The cost of implementing the recommendations arising from the review is being examined as part of the preparation of the Prison Service's implementation plan.
§ Mr. HowarthTo ask the Secretary of State for the Home Department if he will publish the statistical analysis of the cost drivers of publicly-operated prisons undertaken by the Home Office research and statistics directorate as referred to in paragraph 11 of the review of comparative costs and performance of privately and publicly operated prisons produced by Coopers and Lybrand. [35134]
§ Miss WiddecombeResponsibility for this matter has been delegated to the Director General of the Prison Service, who has been asked to arrange for a reply to be given.
Letter from Richard Tilt to Mr. George Howarth, dated 9 July 1996:
The Home Secretary has asked me to reply to your recent Question asking if he will publish the statistical analysis of the cost drivers of publicly operated prisons undertaken by the Home Office Research and Statistics Directorate as referred to in paragraph 11 of the Review of Comparative Costs and Performance of Privately and Publicly Operated Prisons recently produced by Coopers and Lybrand.The Research and Statistics Directorate's analysis is undertaken every year to compare publicly managed prisons and is used for a range of management purposes by the Prison Service.The analysis uses regression analysis, a statistical technique, to identify the relationship between the cost of a prison, and a group of independent variables (for example, size of prison, prisoner mix and so on). In other words, to identify which variables appear to explain existing variations in costs. From this is derived an equation for calculating a "predicted" cost for each prison, in the light of the known variables, which can then be compared with its actual cost.Using this methodology, the variables with the greatest statistical links to costs are:
- baseline certified normal accommodation
- the number of category A prisoners
- whether of not sentenced category B prisoners are held
- the number of unsentenced adult prisoners
- the number of unsentenced prisoners under 21
- the degree to which the prison is in open conditions.
These variables, and their relative weightings, were used to help select the public sector prisons most comparable to the privately managed prisons in Coopers and Lybrand's study.