Table 5 shows, for all prisoners in the sample for whom such an assessment could be made, the numbers of 'high-risk' and 'not high-risk' prisoners reconvicted for a sexual crime or a sexual or serious violent crime in the four-year and six-year follow-up periods.

Bearing in mind that these findings on the 'accuracy' of Board members' assessments of risk relate only to reconvictions, not to reoffending, they can be summarized as follows:

'True positives'

At least one panel member successfully identified as 'high risk' all of the small number who subsequently were convicted of a sexual offence at the four-year follow-up stage and all but one at the six-year follow-up stage: recall that nearly all these offenders committed offences which were punished by severe sentences. Board members also identified as 'high risk' two-thirds of the small number reconvicted of a serious violent offence. These offenders were the 'true positives'. Good examples are:

Case A 

was aged 26 and serving four years six months for five offences of actual bodily harm and two of indecent assault, all against the same adult female victim, plus possession of a controlled drug. This was his last review.

Chairman (a psychiatrist) : 
It is another awful sex offence ... A excused himself on the grounds that he had been sexually abused as a child. He also said he had a head injury, he was on drugs for epilepsy and he used the injury as an explanation for not going to work ... His prison behaviour was quite good and he was described as remorseful. He has a considerable criminal record and he is not a very nice man ... I am inclined to release on supervision rather than release cold ... One issue is whether he has properly admitted his offences.

Independent member: 
I think he should stay in until his EDR. In one report it said that two previous girlfriends had complained about his behaviour but had been too frightened to take their complaints further. The assaults were degrading acts and whatever women come into contact with him will be in danger.

Judicial member: 
He is a thoroughly unpleasant character from a thoroughly unpleasant family. However, I do not want to let him out cold and for this reason want to release. The choice is between a 'short launch' or now and I come down in favour of a launch ... It's better that he is supervised so that he knows he will go back to prison if he steps out of line.

Chairman: 
My initial inclination was to release, although I suggest release on a short launch in September.

Independent member: 
He says he doesn't need treatment for his sexual problems ... This fellow will be back before EDR, he won't respond to supervision and he's a tearaway. He is a very, very bad risk ... There is a risk of bringing the parole system into disrepute and in my view this outweighs everything else.

Chairman: 
I think he is unlikely to offend before his EDR [two-thirds date] although I agree he will reoffend eventually.

Judicial member (concluding the discussion): 
It's worse to release him with no supervision ... Nothing appals me more than the chap coming out of prison with no supervision.

[Page 383]

He was granted a 'launch' of four months' parole. Twenty-two months after his release he was sent to prison for five months for assaulting a police officer. About five months after being released from that sentence he was again reconvicted, this time of multiple offences including false imprisonment, discharging a firearm, grievous bodily harm, rape of a female aged over 16 and indecent assault on an adult woman. He was given a Hospital Order without restriction of time.

Case B, 

aged 48, was being reviewed for parole for the last time. He was serving 11 years for an indecent assault on the eight-year-old son and ten-year-old daughter of his part-time employers. He had nine previous convictions for sexual offences on girls and boys, the first when he was 16 years old, and he had subsequently served five prison sentences, the longest being four years.

Chairman (criminologist) : 
... He has a long history of offences against children ... The main points which struck me were that he had deteriorated a lot in prison ... The Prison Probation officer says that he can go to a [probation hostel] and can get work. He has been doing anger management and he is recommended for parole by his Home Probation officer. Therefore I said 'yes'. He can go now with a forward date.

Judicial member: 
I only questioned a forward date in the light of the fact that he is about to begin the core sex offender programme.

Psychiatrist: 
I suggest that he be released when he has completed his course, as is recommended in the dossier.

Independent member: 
He is a high, high risk this man, but there is now a Centre he can attend to continue work on his sex offending in the community.

The panel therefore recommended that he should be paroled when his course had been completed. Four years and ten months after his release from prison he received a sentence of nine-and-a-half years' imprisonment for indecent assaults on boys.

'False positives'

In contrast, a high proportion (92 per cent) of those identified as 'high risk' by a member of the Parole Board panel turned out not to have been convicted of a sexual offence by the end of the four-year follow-up period. In this sense they were 'false positives'. 

Even if one included reconviction for a serious violent offence, the 'false positive' rate at four years was as high as 87 per cent. When one considers only those followed-up for six years, the 'false positive' rate fell, but was still 78 per cent for a sexual reconviction and 72 per cent when sexual and serious violent reconvictions were combined. 

The following examples illustrate the type of cases where great concern was expressed but no subsequent reconvictions were recorded.

Case C 

was at his first review and serving his sentence in a Regional Secure Unit.

Independent member: 
He is serving four years for rape, a section 53 case [aged 16 when convicted] ... He committed the offences against children he babysat with: a five-year-old girl, her brother aged four and a six-year-old girl. He refuses to accept all the blame. The social work report outlines the difficulties. He will need a lot of help and support on release. There is no sign at present that he shows or feels remorse.

[Page 384]

Psychiatrist: 
His sincerity is unconvincing. In my view he must be an extreme risk to young children at present.

Independent member: 
The paper does not help very much. I say no. He has no understanding of sexual offending and no remorse. He is a very high risk of reoffending.

There was no information on whether he was paroled at a later date, but assuming he was not, in the six years since the two-thirds point of his sentence had been reached (when he would have been released automatically) no reconvictions of any kind had been recorded against him.

Case D 

was aged 42 and serving his sentence for seven offences of serious indecent assault against female stepchildren and the female children of neighbours. It was his second, although not his last, review.

Psychiatrist: 
He is serving eight years for indecent assaults and he will be on licence for 21 months if he were granted parole. The police report gives a dreadful catalogue of sexual abuse over 13 years. However, at least he pleaded guilty ... 
In prison he behaves well but it has been an off and on business whether he goes to his offending behaviour course. In the dossier, the medical officer says that he refused to go to [a therapeutic prison]. On another page it says that he recently withdrew from his offending behaviour group. 
His Home Probation officer is keen on parole, but his report was written before the prisoner withdrew from the offending behaviour group. 
The Local Review Committee [LRC] says no and I say no. He is clearly an extremely grave risk to children and the longer period of parole is not justified or he can be refused on Criterion B [criminal and other history].

Both the other members also said he should be refused parole. He was also refused parole at his next and last review, being released after serving two-thirds of his sentence without statutory supervision by the probation service. No reconvictions of any kind were recorded in the five years following his release from prison.

Case E,

aged 45, was serving four-and-a-half years for buggery of his wife and indecent assault on her two daughters (aged 12 and 15) and son (aged 10) from a previous marriage and his own six-year-old daughter. The offences had continued for several years. He had previously served a two-year prison sentence for attempted buggery of the boy, but on release from prison had returned to the family. This was his second and last parole review.

.Judicial member: 
... He is in for just about everything. He has a very worrying previous conviction in 1987 for buggery of his stepchild. There is a long history of sexual assault in the family and he has been cautioned for abuse of his stepdaughters, and while he was on bail [for the 1987 offence] his stepdaughter alleged assault. In 1988 he was released and returned to his family and then his stepdaughter reported more assaults and gross indecency. It does not sound a very hopeful background ... 
At his first review he was minimizing his actions and he was deeply defensive in relation to his entrenched paedophilic inclinations ... 
At his second review the probation officer says he is continuing to deny the offences and fails to recognize the damage he has done. [The dossier] notes he withdrew from the treatment programme after one session. He says he finds it difficult to discuss offences which did not take place. 
A psychiatric report talks of little insight and says he is manipulative and there is a family problem. 
It's amazing how little insight he has at this stage. It's a complex case. He has an immature attitude but predicting outcome is very difficult. Releasing him

[Page 385]

with no support will not help him ... The Local Review Committee thought him suitable because he needed a structured release plan but I cannot recommend him. I consider he is a risk.

Independent member: 
A very, very big risk.

Psychiatrist: 
I agree. The Social Services believe he has assaulted all the children who have been in his care.

Independent member: 
A bully, a really dangerous man.

Parole was not recommended. The reasons given were: 'Too high a risk for release on licence. Parole will not affect the risk of reoffending.' 
Over the following six years and three months no convictions of any kind were recorded against him.

'True negatives ' and 'false negatives'

Where the prisoner was not identified as a 'high risk' by any member of the Parole Board panel, the prediction turned out nearly always to be correct. By this criterion they were 'true negatives'. Indeed, amongst the six-year follow-up sample only one such prisoner was reconvicted of a sexual offence, contrary to the Board's assessment. In other words, he was a 'false negative'.

Case F 

was aged 49 and was being considered at his last review.

Independent member: 
He has a Reconviction Prediction Score of 13 per cent and is serving eight years for rape on young girls aged six, plus indecent assault on others. They were local children. On the second LB2 [prison report] it says he does not appear to have attended but it seems that no sex offender courses were being offered where he is... His home leave is OK which would indicate he would be reliable on licence... It is almost a year to his EDR [two-thirds date], the LRC are saying no and that he should have counselling. 
There are two choices. One to parole him with a long licence. I do not like that. There are children on the outside. So, secondly, there could be every effort in the next six months for him to attend a sex and alcohol course.

Chairman (Psychiatrist) : 
Well, he has no previous sex convictions.

Judicial member: 
It says constantly through the documents that he is full of remorse. The LRC are absurd to say he has not confronted it, although he had not done any courses. The psychiatric report does not end up by assessing risk. Obviously there is a risk, there always is, but I do not think on 12 months at the end of a long time at [X prison] he would get and complete a course. I would be inclined to give him parole with a condition that he would get it outside.

Independent member: 
What about a psychiatric report?

Chairman: 
Will it tell us anything we don't know already?

Judicial member: 
Shall we postpone until after Christmas with a shorter risk period?

Chairman: 
It would give probation a chance to arrange a programme. We ought to have the usual conditions.

Parole was recommended with a forward date. He was not reconvicted during his parole licence but five years and three months after his release from prison he was reconvicted of an indecent assault on a female under the age of 16 and sentenced to three years' imprisonment. In this sense he was a 'false negative'.

[Page 386

could be made, and not all the prisoners coming forward for review would have begun such a programme of treatment even if it had been available. One of the reasons for this is that prisoners who deny their guilt are not thought to be suitable candidates for a sex offender treatment programme. For if they insist that they 'have not done it' how can they 'address it'? This is why such offenders are generally regarded as especially 'risky'.

We therefore analysed the judgements of risk made about those sex offenders who were said, during the panels' discussions, to be 'deniers'. In this sample, they accounted for a third of all those for whom an assessment of risk was made and they were distributed almost evenly among offenders against adults and intra-familial and extra-familial offenders against children. 

As many as two-thirds of the 'deniers' were identified by at least one panel member as being 'high risk " compared with half of the 'non-deniers'. However, only one 'high-risk' 'denier' was subsequently reconvicted of a sexual offence. All but one of the other prisoners who were subsequently reconvicted of a sexual or serious violent crime had not been 'deniers'. The difference was statistically significant. [*5]

  [*5] Among the 144 prisoners who could be followed-up for at least four years and for whom the Board had made an an assessment of risk only of the 47 'deniers" was reconvicted of a sexual or serious violent offence, compared with 17 of the 97 'non-deniers'. 
This is statistically significant: Χ² = 6.86, 1 df, p > 0.009. 
Standardizing for a four-year follow-up for 'deniers' and 'non-deniers', the figures were 1/47 and 12/97 respectively. 
This remained statistically significant, but at a lower level. Χ² = 4.05. 1 df, p > 0.044. 
At the six-year follow-up the difference was" significant at the same level: Χ² = 3.96. 1 df, p > 0,047.

Thus, it appears that 'being in denial' was given much more weight as a risk factor by Parole Board panel members than turned out to be justified by the criterion of subsequent reconviction for a sexual or serious violent crime. 

Amongst those followed-up for four years, 'deniers' made up 44 per cent of the 71 prisoners who had been identified as 'high risk' by at least one panel member and yet had not subsequently been reconvicted of a sexual or serious violent offence. 

This raises the question whether, as appears to be commonly assumed, sex offenders who have not completed a sex offender treatment programme because of denial are particularly at risk of being reconvicted of a sexual or serious violent offence on their release from prison.

There are several possible explanations. 

The difference in reconviction rates between 'deniers' and 'non-deniers' could, for example, be due to a higher rate of hidden reoffending amongst the former than the latter, although this does not seem particularly plausible. 

It is also possible that a few of the 'deniers' may have been truly innocent. But even if not innocent, some 'deniers', when faced with the stigma of conviction and punishment, may not accept their deviant sexual acts as a reflection of their 'real self'. Nor may they wish to associate with those they regard, unlike themselves, as 'real' sex offenders. 

It is possible that such persons may be less likely to become 'secondary deviants', that is, persons who accept and seek to justify their sexual deviance. Rather, their denial may signify that they are trying to distance themselves from admitting the possibilities of any such behavioural aberration in the future. 

This hypothesis would be consistent with a 'labelling' perspective of deviant conduct. We are not experts on the dynamics of sexual offending as such but, in our view, these findings do suggest that the link between denial and reoffending is a subject that warrants more thought and research.