A serious problem

Given that the last month in Australian media has been full of appalling stories of pedophilia (apart from the Morecombe case itself, there are the terrible allegations from several women about - alleged - comic actor Robert Hughes, and the extraordinary case of the Cairns based gay couple who acquired a baby with the explicit purpose in mind,)  I thought I would look around for recent commentary on the whole issue of pedophilia and recidivism.

This one from Harvard Medical School in 2010 is pretty good.  The problem is, as the title says, there is pretty widespread pessimism about how it can be treated with any sufficient degree of certainty that you can prevent recidivism.   This is a real nightmare for those in the justice system, and the community more generally. 

Look at Cowan himself.  After serving 4 years for a second rape of a boy, and being sent by his parole office in Darwin to Queensland so that could do a sexual offenders program, he apparently got into an evangelical Church and met his wife there, with whom he had two children (one before he killed Daniel, one after.)  He certainly looks clean cut and reformed in this street interview in Brisbane in which he proclaimed he was a new man.  For any psychologist dealing with him, this must have sounded like the best possible type of support group to help ensure he did not re-offend.  Yet the urge to sexually assault boy strangers was strong enough that he acted on it.

The thing is, what confidence should the public have in sexual offender programs delivered through the justice system?   The Harvard article above notes this about recidivism:

Estimates of recidivism vary because studies define this term in different ways. One review found recidivism rates of 10% to 50% among pedophiles previously convicted of sexual abuse, although this could include anything from an arrest for any offense to reconviction on a crime against a child. One long-term study of previously convicted pedophiles (with an average follow-up of 25 years) found that one-fourth of heterosexual pedophiles and one-half of homosexual or bisexual pedophiles went on to commit another sexual offense against children.

This would indicate the recidivism rate for someone like Cowan is extraordinarily high, and his case is far from atypical.

Are judges kept fully aware of the research on the topic?

And what about the style of treatment and its efficacy that is attempted in the sexual offender programs in Australia?   I see there is a 2010 Queensland study on the sexual offender program, but it is only for about 409 released from 2005 to 2008 - a pretty small sample size over a very short period.  (The report acknowledges these shortcomings;  it also notes that for offences against children, there is often a substantial delay before the offence is reported.)

The report finds that the program seems to have a positive effect (in that those who did the program have a somewhat lower recidivism rate for sexual offences), but obviously the study is looking at a very short period.    The parts that discuss sexual offence recidivism rates generally are of interest. On the "down" side:

Despite a great deal of research effort devoted to the question, knowledge concerning the dynamic risk factors associated with sexual offender recidivism (otherwise referred to as criminogenic needs) remains limited. Some of the psychological characteristics commonly targeted for improvement in sexual offender programs (e.g. sexual preoccupations; impulsivity; intimacy problems) tend to be only weakly (even if significantly in a statistical sense) associated with sexual offender recidivism. Other common treatment targets (e.g. denial; victim empathy) have recently been shown, on average, to be unrelated to sexual offender recidivism (Hanson & Morton-Bourgon, 2005).

 On the (slightly) more positive side:

A meta-analytic review of sexual offender recidivism studies, involving more than 80 separate studies and almost 30,000 sexual offenders, found an average sexual recidivism rate of 13.7% over an average time at risk of five to six years (Hanson & Morton-Bourgon, 2005). In that meta-analysis, rates for nonsexual violent recidivism were 14.3%, and for ‘any’ recidivism 36.2%. For the present evaluation, the average time at risk of two years and five months was a little less than half the average observation period for the studies included in Hanson and Morton-Bourgon’s meta-analysis. Notwithstanding jurisdictional variations in crime reporting.  and recording, it would be reasonable to expect that over the next two or three years observed recidivism base-rates for the Queensland offenders would be more similar to international averages.

But what about for longer periods?   This passage seems to me to be written with unwarranted optimism, given the number quoted:

Notwithstanding the limitations of official recidivism data, it seems clear that a significant proportion (perhaps the majority) of convicted sexual offenders do not go on to commit further sexual offences, even without treatment. Studies that have followed sexual offenders for 20 years or more still tend to find sexual recidivism rates well below 50% (Hanson, 2000; Janus & Meehl, 1997). This raises further questions about the universal need for specialised sexual offender treatment.    

What's missing, of course, is more specific figures for the rate of pedophiliac recidivism.

From reading the Harvard commentary, and this Queensland report with its hesitation about how effective these programs really are, it would seem reasonable to conclude  that the rate of recidivism for pedophiles is significantly higher than for sexual offenders against adults, and is very likely to stay that way even if they undergo a sexual offender course.

More pessimism can be taken from the psychologist who actually treated Cowan.   While he considers sexual offender programs do help (and while ever they have some effectiveness, I wouldn't argue against them - except to the extent that they may give unwarranted confidence to judges when sentencing,) he freely admits you can never really know if someone will re-offend:

He says evidence suggests such programs are effective when they are "well designed and well implemented".

"But there are never any guarantees for individuals, and when you're dealing with high-risk offenders in the first place there's bound to be some recidivism among that group," he said.

"The difficulty is predicting who among that group is and isn't likely to go on to commit further offences."

He says sex offenders often display no observable behavioural features apart from the offences they commit.

"In fact for many sex offenders, the most unusual thing about them is that they have committed sex offences," he said.

"We know that in retrospect, but it's obviously very difficult to know in prospect."

It seems to me that if the long term recidivism rate for guys like Cowan is around 50%, and if sexual offender programs have only ever been shown to have modest effects (say, reducing re-offense by pedophiles to 40%?) this is a really significant matter which must be taken into account in terms of how both sentencing and the terms on which they should ever be released back into the community.

(If any psychologist thinks my guesstimate figures in that paragraph are  wrong - let them come out and be specific about what figures they know about.)

So how is the justice system supposed to deal with these offenders?    The judge in the Cowan case has been criticised for making the observation that an unintended consequence of tougher sentencing for sex offenders is that it can give them a greater incentive to kill their victim.   This makes some intuitive sense, but really, it has got to be balanced with the actual evidence about the rate of recidivism.   What I find particularly hard to credit is that second offence sentences for sexual assaults on children should take this effect into account. 

The Courier Mail is running hard on the previous judges being too lenient in sentencing.   I generally am wary of such media campaigns for various reasons, but it is interesting to note that there was an appeal by the Prosecution which failed against the 7 year (with non parole of 3 1/2 years) sentence for Cowan's second offence in Darwin.  The point is that, reading around on the topic, there do appear to be grounds to be concerned that judges may have an inadequate appreciation of the recidivism rates for this particular type of offender.  If the media can play a role in highlighting this, I don't have a problem with that.

But even then, longer sentences for early offences is not the only issue.   The community hates the idea of released repeat offenders of any age being anywhere near them, and who can blame them, really?  Chemical or physical castration if issued as a punishment by the State feels either too medieval, or "Clockwork Orange," to most liberals, I assume, but it is interesting to see it is still a live issue as a voluntary measure in the United States, and chemical castration is a recently introduced punishment in Korea.  

(One argument noted against physical castration is that men can seek to counteract its affects by testosterone patches - which are presumably not so hard to get your hands on these days.)

The article about the Korean situation notes the 50% recidivism rate for pedophiles:

Surgical castration reportedly produces definitive results, even in repeat pedophilic offenders, by reducing recidivism rates to 2% to 5% compared with expected rates of 50%. Chemical castration using LHRH agonists reduces circulating testosterone to very low levels, and also results in very low levels of recidivism despite the strong psychological factors that contribute to sexual offending (10). Chemical castration has some advantages over surgical castration. First, although chemical castration is potentially life-long for some offenders, it might allow sexual offenders to have normal sexual activity in context with psychotherapy. Second, some sexual offenders may voluntarily receive chemical castration. Third, chemical castration may be a more realistic restriction than electronic ankle bracelets or surgical castration. Fourth, unlike surgical castration, the effects of anti-libido medication are reversible after discontinuation. Finally, the general public may feel relieved knowing that sexual offenders are undergoing chemical castration.

 I'm not at all sure that the public thinks the reversibility of chemical castration is an advantage, but at least chemical castration (I assume) avoids the issue of physical castration being countered by testosterone.

I do not have a conclusive view on this, but I would not reject a serious discussion about the voluntary use of chemical castration as a precondition for release of offenders such as Cowan, regardless of their age.

Update:   I see that a New South Wales Parliamentary committee is said to be already considering chemical castration as a sentencing option.   This report notes that civil libertarians were unhappy with the suggestion, but interestingly, the opinion of one important group of professionals sounds quite sanguine and practical about it:

But the Royal Australian and New Zealand College of Psychiatrists president-elect, Louise Newman, believes chemical castration, which works as long as the patient stays on the medication, should be considered for the more "hard core" sexual offenders.

"It's certainly not a cure or a way of reducing all risk, but it might be seen as a useful component of treatment or management for some of these very difficult cases, where we're unlikely to see response to other methods," Professor Newman told ABC News Online.

"They're not commonly used at all and they are not needed other than in the fairly difficult group of very severe offenders."

The drugs impact on liver function, so those undergoing the treatment usually have their health checked regularly.

Professor Newman says that although some sex offenders have a good chance of recovery, there will always be a small group of people who will continue to pose a risk to the community.

"There's always some risk of reoffending and obviously someone can offend in various ways," she said.

"Clearly there is a group of people for whom risk is significant and they might have ongoing and really severe difficulties in managing their impulses.

"Containment and monitoring of those sorts of people might actually be necessary to ensure community safety."

 I could not think of a more ideal candidate for having his release on parole subject to a lengthy mandated period of chemical castration than Cowan on his second sentence.  Of course, this would presumably have stopped a relatively young man from marrying and having kids as he did, but look at what it could have stopped.

I also see from Wikipedia that, surprisingly, in the United States, California was the first state to introduce it as a punishment for child molestation.  It is a mandatory if going on parole after a second offence.  The article notes about 7 other states have "experimented" with it; I am a little surprised that it has not been used more widely in some of the more conservative states. 

The Wikipedia story also lists quite a few European nations that have implemented it since about 2010.  There would seem to be a bit of an international movement towards accepting it as useful particularly for repeat offenders against children.

It seems to me that it is time for it to be discussed in more detail in this country, too.