Serious further offences: An exploration of risk and typologies
| DOI | 10.1177/0264550508099712 |
| Date | 01 March 2009 |
| Published date | 01 March 2009 |
| Author | Jackie Craissati,Oliver Sindall |
| Subject Matter | Articles |
Article
The Journal of Community and Criminal Justice
Copyright © 2009 NAPO Vol 56(1): 9–27
DOI: 10.1177/0264550508099712
www.napo.org.uk
http://prb.sagepub.com
Serious further offences: An exploration of risk
and typologies
Jackie Craissati, Oxleas Foundation NHS Trust
Oliver Sindall, Canterbury Christ Church University College
Abstract This study examined 94 serious further offences (SFOs) committed by
offenders managed by the London Probation Area over a 14-month period
between 2004 and 2006. Data was collated on background and criminogenic
variables for the sample, as well as retrospectively administering a range of vali-
dated actuarial measures of risk prediction. The predictive validity of key variables
embedded within the probation OASys tool were also examined, specifically the
utility of the DSPD variables and the OGRS and OGRS-2. The SFO offender charac-
teristics – in keeping with previous studies – were wide ranging, as were the risk
profiles of the offenders, the majority of whom fell into the medium risk range. The
risk assessment and personality-based assessment tools routinely available to the
probation service – that is, DSPD and OGRS scores – were found to be at least as
accurate as other actuarially based methods of risk prediction. There were no clear
identifying features of this sample of offenders which seemed to differentiate them
from a much wider sample of the probation caseload, although the research was
not designed to provide a comparison study. However, examination of the nature
of the SFOs suggested there were some key situational contexts which were associ-
ated with serious harm to others (rather than simply likelihood of re-offending).
These contexts included pervasive weapon use, unexpected victim responses during
acquisitive crimes, and disputes occurring within an anti-social male subculture.
Recommendations included a greater emphasis on harm reduction approaches
to probation interventions in terms of risky lifestyles and weapon use.
Keywords offending, risk, typologies
Introduction
Serious further offences (SFOs) are defined as occurring when an offender under
probation supervision – either on a community sentence or on licence following
9
10 Probation Journal 56(1)
custody – is charged for murder or other very serious offences. Such incidents must
be reported immediately to the Probation Unit at the Home Office and are subject
to management review. The impact of SFOs reaches beyond the general conse-
quences of re-offending for the victim and their family, society and indeed the
offender himself; there are significant implications for SFOs – complicated by
the media attention which tends to be associated with such incidents – for the
probation service and its employees, not least of which is the risk of a loss of public
confidence and reduced morale within the service.
SFOs have been the subject of study, and reports include Hedderman and
Vennard (1997) who gathered data on 107 incidents drawn from 128 notifications
and management reviews reported over a six month period in 1997. A subsequent
analysis of 90 management reviews for the London Probation Area was reported
by Gavin (2003); and HM Inspectorate of Probation (2005) reported on a sample
of 292 management reviews drawn from 2002 to 2004. All the sampling methods
varied, and different variables were assessed; with the absence of any comparison
groups, it is difficult to make wider generalizations from the data. However there
is reasonable stability of findings across the samples: for example the offenders
tend to be in their mid-20s at the time of the SFO, approximately 50 per cent have
prior violent convictions and around half are considered to pose a high risk to
others prior to the SFO. In terms of the SFO incidents themselves, approximately a
third to a half involve new charges of murder or attempted murder and just under
a third comprise new charges for rape; substance misuse is involved in the SFO in
a quarter of cases and mental illness is present in less than 10 per cent of instances.
The one study that did take a case-control approach was that of Soothill et al.
(2002): the researchers used data from Home Office datasets, and analysed the
previous criminal histories of 569 males convicted of murder, aged under 45 years,
focusing on those with previous criminal convictions (386 who were matched to
two sets of control groups, a group of offenders with a general criminal history and
a group of offenders with a previous violent conviction). A further sample was of
678 offenders convicted of a serious sexual assault on adult women, with previous
convictions, comparing them to a control group of offenders with a general criminal
history. The study found that any history of convictions for wounding, robbery and
arson indicated a substantially increased risk for a future conviction for murder
(as did the rare crimes of manslaughter, blackmail and kidnapping). These findings
were fairly similar for future risk of a serious sexual assault, although cruelty to
children was also a risk indicator here, as well as a history of minor sexual assaults.
Furthermore, familial homicide offenders were found to have very different risk
factors to the other offenders, and specifically a threat to kill was associated with
a 12 times greater likelihood of family murder, but also of further general violence.
Robbery greatly increased the risk of a subsequent conviction for murder of a male
stranger or acquaintance. However the authors do point out that there were very
few significant factors in the offenders’ prior criminal histories which had any
impact on the likelihood of future homicide or sexual assault, and this included
the number and range of prior convictions.
A comparable system of incident reporting is available for suicides and homi-
cides committed by individuals with a history of mental illness (Dept of Health,
Craissati ● Serious further offences: Risk and typologies 11
2001). The National Confidential Inquiry into Suicide and Homicide was notified
of 1579 homicides in England and Wales from 1996 to 1999. They found that
the majority of perpetrators and victims were young men, around a third of whom
killed a relative or former partner, and a further quarter killed a stranger. The
commonest method of killing was stabbing. Fifteen per cent of all the homicides
were committed by individuals considered to have an abnormal mental state at
the time – although 34 per cent had a lifetime history of mental disorder – and
the rate of reported homicides by the severely mentally ill has remained steady at
around 50 nationally per year. The most common diagnoses were alcohol and
drug dependence and personality disorder, with 5 per cent of all perpetrators
having a diagnosis of schizophrenia. Mentally ill perpetrators were less likely to
kill a stranger than those without mental illness, and alcohol and drugs were more
likely to contribute to the offence in people convicted of homicide who were not
mentally ill. Nine per cent of all perpetrators had been in contact with mental
health services in the year before the offence, and at final service contact, both
the immediate and long-term risk of violence were estimated to be low or absent
in over 75 per cent of cases.
A subsequent study looking at homicide inquiry reports in the severely mentally
ill where there was a prior history of violence (Maden, 2006), examined 25 cases,
and retrospectively applied the HCR-20 (Webster et al., 1997), an actuarially based
structured risk assessment approach designed for application with mentally dis-
ordered offenders. Maden found that there was a high rate of scores on the histori-
cal items – static items emphasizing criminogenic variables more than specific
mental illness variables – which tended to be under-emphasized by general mental
health teams. He also identified the problems of substance misuse, personality
disorder and criminogenic variables in the histories of the homicides committed
by perpetrators with a diagnosis of severe mental illness.
Earlier research by Maden et al. (2005) found that OGRS-2 scores (Copas and
Marshall, 1998) – now embedded within the OASys – were associated with violent,
sexual and any other convictions at two and four year follow-up with a sample of
mentally disordered offenders, suggesting again that historical criminal factors
are the greatest risk factors associated with reconviction in mentally disordered
offenders.
The aim of the current study was to draw on the findings from the above studies,
and to:
● Provide a descriptive profile of serious further offences (SFOs) and SFO
offenders, drawing on key background and offence-related variables
available in the probation information system and files;
● Provide a comprehensive retrospective review of the risk status of the SFO
offenders, comparing widely used independent actuarial scales with the
risk measures intrinsic to the OASys;
● Identify possible targets, arising from the research findings, which might
enable the probation service to identify risky situations in relation to SFOs.
12 Probation Journal 56(1)
Method
All the subjects were offenders under the supervision of the London area of the
National Probation Service. All the subjects were selected for committing a serious
further offence (SFO) during a 14-month period between 2004 and 2006.
The data was derived from the Probation Service computer databases (Delius
and OASys) and, where possible, the hard case files. Information was collected
from CPS documents, pre-sentence...
To continue reading
Request your trial