Appendix D — Further information on documents and risk assessments used in the research
This Appendix provides further information on the documentary evidence reviewed in this research, with details from policy documents and other literature on the nature of the documents, their typical contents, and the purposes for which they are produced.
D.1 Risk assessment methodologies
Chapter 6 describes the sample in terms of their assessed risk, differentiating among assessments made using varied methodologies. This section sets out key contextual information about these. All page references are to HMPPS’s Public Protection Manual (2016) unless otherwise indicated.
The risk assessment methodologies used in the overall Offender Assessment System (OASys), which are described and noted in Chapter 6, fall into two main groups: those using actuarial methodologies, which produce a numerical prediction, and those using structured professional judgement, which produce a qualitative assessment of the level of risk involved. There are differences within these groups, which are set out more thoroughly in Table D.1.
The wider OASys contains a larger range of assessments, and those described in Table D.1 are not comprehensive: only those methodologies covered by the documents I reviewed, and which are reported on in Chapter 6, are covered. Risk assessments for sexual violence, though relevant for some of the men that I interviewed, and though included in the overall OASys, were not included, nor summarised in the documents made available to me. It was only noted whether participants had such an assessment.
| Methodology | Assessment | Further details |
|---|---|---|
| Actuarial1 | OGRS3 (Offender Group Reconviction Score, version 3) |
|
| OGP (OASys General reoffending Predictor) |
|
|
| OVP (OASys Violence Predictor) |
|
|
| Structured Professional Judgement | RoSH (Risk of Serious Harm) |
|
D.2 Summary of documents reviewed
The selection of documents reviewed in the OMUs at both prisons were not the same, and in fact overlapped only in part. This section offers brief further context on these documents, the purposes for which they are produced, and the information they contain. Not every document was available for every prisoner who consented to my viewing his records.
RC1 forms
The RC1 form summarises the decision-making process and basis for regular individual security categorisation reviews. Categorisation reviews for determinately sentenced prisoners “must” take place every 12 months, until three years remain to the earliest possible date of release, at which point they shift to being completed every six months (Ministry of Justice and HM Prison & Probation Service 2021 para. 8.5). For lifers, the timescales are not specified, but recategorisation is meant to take place alongside the sentence planning process, except if the prisoner “needs to be able access [sic] a progression opportunity […] at a prison of a lower category”, in which case the review should be expedited (Ministry of Justice and HM Prison & Probation Service 2021 para. 8.6).
The vast majority of those I reviewed at Swaleside had been completed during the last 12 months, with only a couple dating from as much as two years beforehand; I was told by an OMU staff member that they were normally done annually. This was something of a puzzle, given the requirement for them to be done alongside sentence planning, and the plainness with which so many men at Swaleside said they had not had a sentence plan or an OASys done for many years. I formed the impression that the RC1s were handled mainly as a formality and a paper exercise, and this accorded with what I was told by prisoners, as well as with the fact that only two examples contained any comments or contributions from the subject: one in which the prisoner had responded himself, and one from a highly litigious former category-A prisoner at Swaleside whose solicitor had contributed on his behalf. It is an explicit expectation that prisoners “must” be provided opportunities to participate in the sentence planning process (Ministry of Justice and HM Prison & Probation Service 2018 para. 4.38); whether or not they were provided such opportunities, only one man from thirty in the sample took it.
The forms themselves were recommended to me by OMU staff at Swaleside as the best single, recent, brief, but comprehensive overview of how the prison understood the individual under review. This was a fair description. Each form followed a standard format, containing:
- an overview of key sentence information
- categorisation history including the recommendation for the current review period
- risk data including most recent OASys assessments, RoSH levels, and any escape or abscond histories
- an offence narrative excerpted with varying degrees of abridgement from the most recent OASys, which usually also used extracts from the judge’s sentencing remarks and the MALRAP file summarising police records (Ministry of Justice 2019) to comment on and often contradict the subject’s account of the offence
- a summary of risk-reducing work prescribed by the sentence plan
- a summary of case notes from the review period, excerpting positive and negative entries by staff about the subject
- IEP levels during the review period
- Adjudications during the review period
- Contributions from the key worker and the prisoner (both almost always empty)
- Contributions from the Security department
- The outcome of the review
Enhanced Behaviour Monitoring assessments
Enhanced Behaviour Monitoring assessments (EBMs) are made, on reception to open prison, for all high-risk prisoners (and all lifers, regardless of their assessed risk). They are also completed for all prisoners in closed men’s prisons who are part of a Progression Regime (HM Prison & Probation Service 2019, paras. 1-2). As such, EBMs would have been produced for men on Swaleside’s PIPE unit, but not for the mains prisoners there; I was not offered access to them at Swaleside, and hence the only ones I accessed for the present sample were at Leyhill, where the OMU recommended them to me for similar reasons as the RC1 was recommended at Swaleside.
The EBM assessment is not a risk assessment, but “a means of collating all known risk related information about an individual” (HM Prison & Probation Service 2019 para. 4.3). Conducted initially from written prison records by a prison psychologist, and mandatorily completed to a tight timescale (4 weeks) from the prisoner’s arrival in open conditions, the EBM ‘formulates’ the individual’s risk, lists factors of concern, and defines behaviours prison staff should interpret as indicating elevated and reduced risk. Based on the EBM, some prisoners are subjected to additional surveillance and monitoring and further review on an at least monthly basis. For others, the initial assessment document simply remains on file. Regardless of how it is used, the EBM remains a comprehensive overview of past assessment documents, and is expected to play a part in (but not to stand in for) subsequent risk assessments, such as those relating to ROTL.
The content of these documents was in many ways like that of the RC1 forms, in that they gave a similar overview of sentence information, OASys data, risk-reduction work prescribed and completed, and overviews of current custodial behaviour. However, they differed in offering more commentary, and also making an enormously more nuanced, detailed, and—in terms of behaviours—descriptive and ‘net-widening’ formulation of risk. This was absent from the RC1s.
The general impression I formed of the EBM assessments was that they represented a sharp and striking ‘tightening’ of penal supervision, that they were probably best thought of as a reflection of the changing means employed to manage risk after the move from closed conditions, and that (from prisoners’ point of view) they might often seem to dash hopes and expectations for a ‘loosening’ of controls that might have built up earlier stretches of the sentence. Even so, some participants clearly found the EBM process, and the regular review meetings and feedback loops it resulted in, very helpful—saying it was the first time they had seen their risk so clearly and the conduct expected of them so clearly described. But lengthy lists of risk factors, including some which were very vague, left scope for broad categories of behaviour to be labelled ‘risky’.
ROTL risk board records
Life-sentenced and other indeterminately sentenced prisoners are eligible only for Restricted releases on temporary licence (ROTLs), for which the process of risk assessment is more specific and more demanding than for Standard ROTLs (Ministry of Justice and HM Prison & Probation Service 2022). The risk board documents shared with me by Leyhill’s OMU recorded the decision made by a ROTL board, consisting of the subject’s offender supervisor and a senior manager. Compiled through a lengthy process in which offender managers, probation victim liaison staff, victims, police, and immigration enforcement authorities were all required to be consulted, the board then considered risk assessments written by the Offender Supervisor, considering each of the types of ROTL for which permission was being requested, and made recommendations for a decision and any licence conditions required. The final decision, in all cases affecting lifers, would be made by the prison governor or his/her deputy.
The contents of the risk board documents were, again, similar in the underlying basics to the other assessments: sentence details, risk data, OASys records including an offence narrative, details of past offending, gists of security intelligence, and risk assessments. The latter were, like the EBM assessment, more detailed than anything in the RC1s, but not because they described in depth what might be the indicators of increased risk; in this case, they were more informative about the prisoner’s release plans and circumstances, family relationships, and similar. The Board’s recommendations, and the proposed licence conditions, were a good indicator of the trouble (or otherwise) that the prison expected to see if releases went wrong.
Police National Computer printouts
The Police National Computer is a database of criminal records managed by the Home Office, and containing entries for approximately thirteen million people as of 2021.12 Printouts supplied to me by the OMUs at both prisons listed past convictions and out-of-court measures for each participant, first summarised by conviction type, and then listing all past convictions, cautions, etc. by date, number, and the court or police force involved. The printouts also gave details of any aliases or other names by which the subject had been known.
Judge’s sentencing remarks
The judge’s sentencing remarks were the final documents consulted. They were also the hardest to obtain: at Swaleside, I was told by the OMU manager that they were not routinely held by the prison and that they had to be requested from the courts, but by one of the Psychology staff that they were usually available in the paper ‘warrant file’, which (as I discovered) contained all documents relating to the prison’s legal authority to hold prisoners in custody. The OMU allowed me to consult these files on the shelves; they were often hard to find and I discovered a number of cases where prisoners’ documents had been mixed up or placed in the wrong file, but I was able to find complete sentencing remarks in about three-quarters of cases, versions with pages missing in one or two, and an alternative document giving similar information.13 Sentencing remarks were often also quoted, usually in highly compressed and selective form, in the OASys assessments.
I found them fascinating documents to read, in part because they were so individual, so different from other documents, so expressive, and so clearly shaped by the writing style of the sentencing judge. In all these features, they stood out from the more administrative and consistent documents produced by prisons and probation services. They varied considerably in content and length, and it was evident that their nature changed very substantially over time. Those from before 2003 were highly condemnatory, used moralistic language, and were usually also very short: one transcript ran to exactly 129 words.
Those from after 2003 were enormously more detailed and much longer—sometimes ten or a dozen pages, and sometimes more. They varied greatly in style and content. Most narrated the offence in some detail and then assigned the sentence, but some began by setting the sentence and then referring to features of the case by turn, and without sequence. Some were matter-of-fact, others laid censure on thickly; some speculated into the lacunae left by known facts, filling in details such as motive probabilistically and from imagination, while others noted aspects of the case which might never be known. While it was usually clear what was being censured, some documents were confusing to read and no doubt more confusing to have listened to.
Appendix D — Further information on documents and risk assessments used in the research – Moral messages, ethical responses: punishment and self-governance among men serving life sentences for murder Appendix D — Further information on documents and risk assessments used in the research – Moral messages, ethical responses: punishment and self-governance among men serving life sentences for murder Appendix D — Further information on documents and risk assessments used in the research – Moral messages, ethical responses: punishment and self-governance among men serving life sentences for murder Moral messages, ethical responses: punishment and self-governance among men serving life sentences for murder
Not recorded in documents I reviewed and not represented in the research.↩︎
Either specific children, or children in general (p. 2).↩︎
Specific persons with an existing connection to the subject of the assessment (e.g. victims, former partners) (p. 2).↩︎
To be interpreted “either generally or a specific group such as the elderly, vulnerable adults […], women or an ethnic minority group” (p. 2).↩︎
Any person working with the subject in a professional capacity (p. 2).↩︎
Not recorded in documents I reviewed and not represented in the research.↩︎
Either specific children, or children in general (p. 2).↩︎
Specific persons with an existing connection to the subject of the assessment (e.g. victims, former partners) (p. 2).↩︎
That is, the public “either generally or a specific group such as the elderly, vulnerable adults […], women or an ethnic minority group” (p. 2). Unlike the ‘known adult’ assessment, the ‘public’ assessment represents the judgement that in any circumstances the assessor can imagine, the prisoner might in future harm either a person in a specific category, or simply any person. In ‘known adult’ assessments, the assessor must fit the assessment to a defined set of actual relationships. They can, in principle, inquire into the nature of these relationships, and hence are constrained by known facts. In the ‘public’ assessment, they are constrained only by what can plausibly be imagined.↩︎
Any person working with the subject in a professional capacity (p. 2).↩︎
HC Deb 18 January 2021, vol 687 col 621.↩︎
For example, the sentencing judge’s recommendations to the Lord Chief Justice, or in some other cases, a Court of Appeal resentencing judgement which summarised the original court’s censure before adding its own glosses.↩︎