Annex I
RECONVICTION RESULTS FROM HMP GRENDON THERAPEUTIC
COMMUNITY
1. Grendon prison opened in 1962 and until
the opening of Dovegate in 2001 was the only prison in England
run solely on Therapeutic Community (TC) lines. It comprises an
Assessment Unit and five wings each run as a TC for around 40
prisoners. Following early studies that showed no effect on reconviction
rates, research evidence of a positive treatment effect on reconviction
began to emerge in the early 90s in studies by Cullen (1993) and
Newton and Thornton (1994). Better quality evidence has been published
in the last six years.
2. Marshall (1997) undertook a four-year
follow-up study of 702 prisoners admitted to Grendon during the
period 1984 to 1989, with two control groups. One was a group
of 142 prisoners on the Grendon waiting list who never went there,
the second was a group of 1,425 prisoners drawn from the general
prison population with similar offence characteristics. Marshall
suggests that the most likely reasons for the waiting list prisoners
not being admitted was either because no place became available
soon enough or because they were released on parole earlier than
expected. As neither reason would be expected to substantially
affect re-conviction rates this makes them a suitable comparison
group, but rather small, so the larger general prison group was
also needed.
3. The following is a summary of the key
findings.
4. Prisoners selected for Grendon tended
to be higher risk offenders when compared with other prisoners
of similar age and serving similar sentences for similar offences,
which means that previous studies comparing Grendon prisoners
with the general prison population were not comparing like with
like.
5. Prisoners on the waiting list were significantly
more likely to be reconvicted, to receive a prison sentence, and
to be convicted of a violent offence than those in the general
prison population.
6. Although both the treated and waiting
list groups had higher reconviction rates than the general prison
population, prisoners who went to Grendon had lower rates of reconviction,
fewer custodial sentences and fewer reconvictions for violent
offences than prisoners on the waiting list who never went there.
7. Re-conviction rates were lower for prisoners
who stayed longer at Grendon. Divided into roughly equal groups
of those who stayed 0-6, 6-12, 12-18 and over 18 months, the biggest
reduction occurred for those staying over 12 months. Those who
stayed over 18 months showed reductions in reconviction of a fifth
to a quarter.
8. Although numbers were small there was
a reduction in re-conviction for sexual or violent offences among
repeat sex offenders in the admitted group compared with the waiting
list group, and a reduction in reconvictions for violent offences
among repeat violent offenders in the admitted group compared
with the waiting list.
9. Marshall also found that reconviction
rates were higher if offenders were transferred to another prison
rather than released into the community following their time at
Grendon.
10. Taylor (2000) followed up Marshall's
admitted and waiting list groups at seven years. He found substantially
the same results, with the difference in reconviction rates between
the admitted and the waiting list group at the 0.1 level. Overall
reconviction rates were 73% for the waiting list group and 66%
for the admitted group. Reconviction rates for a violent offence
were 37% and 30% respectively. Taylor examined those Grendon prisoners
serving life sentences (n=104) who were released on life licence.
At four years 8% of those who could be followed up had been re-convicted
for a standard list offence compared with an expected rate of
24% for a comparable group matched for criminal and demographic
characteristics. At seven years these figures were 11% and 28%.
Taylor summarises his conclusions as follows:
"Grendon appears to select offenders who
have a high risk of reconviction. This may be due to the fact
that the prison selects offenders with personality disorders.
There is also some evidence of a treatment effect, particularly
for those who stayed at least 18 months, life sentence prisoners
and repeat sexual offenders."
RECONVICTION, READMISSION
AND COST
OFFSET AT
HENDERSON HOSPITAL
11. Established in 1947 as the Social Rehabilitation
Unit at Belmont Hospital and subsequently renamed in 1960, Henderson
Hospital was where Maxwell Jones developed what became the blueprint
for Democratic Therapeutic Communities for people with psychopathic
or personality disorders. Still going strong, it has survived
various closure threats, produced a stream of research publications
including Rapoport's classic Community as Doctor (1960),
and been "cloned" in two new "Hendersons"
in Birmingham (Main House) and Crewe (Webb House). Outcome research
has generally included both hospital readmission and reconviction
as a combined measure, and has tended to cite success rates rather
than reconviction rates, so that comparison of Grendon and Henderson
figures needs to be done carefully. Follow up periods also tend
to be shorter.
12. Copas et al (1984) followed up
245 referrals to the Henderson between 1969 and 1971. One hundred
and ninety-four were admitted, 51 either failed to attend for
interview or were rejected as unsuitable. Thus the latter group
is not an ideal control group as it is likely to differ at least
in terms of motivation, although the two groups had similar rates
of previous convictions: 53% of admissions and 57% of those not
admitted had at least one previous conviction. Success was measured
in terms of no further convictions or hospital admissions over
periods of three and five years after discharge or from the date
of assessment interview. Success rates at three and five years
were 41% and 36% for the admitted group and 23% and 19% for the
non-admitted group.
13. Personality disorder was assessed using
the Hostility and Direction of Hostility Questionnaire. Both the
admitted and not admitted groups obtained mean total hostility
scores above 30, higher than the scores obtained by men on reception
at Grendon. In terms of reconviction, 63% of admissions who had
one or more previous convictions and who stayed at least 18 days
were reconvicted in the three years after discharge, in comparison
with 90% of the non-admitted group—which as mentioned above
is likely to have been a less motivated group. The following table
demonstrates a clear relationship between successful outcome (no
readmission or reconviction at three years follow up) and length
of stay:
|
| Not admitted
| Admitted group: length of stay in months
|
|
| | 0-1
| 1-2 | 2-3
| 3-4 | 4-5
| 5-6 | 6-9
| 9+ | Total
|
| Number of cases | 51
| 51 | 29
| 25 | 19
| 18 | 14
| 21 | 17
| 194 |
| % success at three years | 23
| 29 | 34
| 36 | 32
| 44 | 43
| 62 | 71
| 41 |
|
14. In an interesting development Dolan et al
(1996) extended the criteria of reconviction and re-admission
rates to include the economic cost of the service. In a small
study they calculated the cost of psychiatric and penal services
in the year before admission and the year after discharge for
a series of 29 consecutive admissions. Full data was available
for 24 patients, who achieved an average of six PD criteria each
on a self-report diagnostic questionnaire. Costs were identified
for in-patient beds, secure beds, outpatient therapy, day-hospital
attendance and prison. The average cost per patient in the year
before admission was estimated to be approx, £14,000. In
the year after discharge this cost fell to £1,300—less
than one tenth. The authors concluded that if this saving were
maintained for two years it would recoup the full cost of an average
stay of 7.5 months in the Henderson.
| 
