nextgenforensic

Long-term reconviction rates for individuals convicted of indecent image offences appear to be low

Ian A. Elliott

Although they’re a relatively small proportion of individuals convicted of sexual offences there is increasing concern about the behaviours and management of individuals with offenses relating to indecent images of children (IIOC) online. The consensus in the literature appears to be that, contrary to popular notions, sex offenders don’t reoffend at high rates and that the rate for IIOC offenders is lower than those who commit contact sexual offences. This post summarises the findings of our new study into (relatively) long-term reconviction rates for IIOC users.

In their meta-analysis, with a combined sample of 2,630 IIOC offenders, Seto et al. reported a reoffending rate of just 4.6% for new sexual reconvictions after approximately six years, with 3.4% committing subsequent IIOC offenses and 2% committing subsequent contact offenses. Since that meta-analysis, more recent studies (such as Eke et al., Faust et al., and Krone et al.) have also reported sexual reconviction rates (IIOC and contact) ranging between 5% and 11% over approximately 4-5 years.

These findings generally support Seto et al.’s assertions that although many IIOC offenders are likely to be sexually interested in children, those pedophilic interests do not necessarily result in contact sexual offenses, and that there appears to be a distinct group whose sexual crimes appeared to be related to illegal pornography or online solicitation alone.

“Two of the key limitations of the existing reconviction studies have been the small sample sizes and the short durations over which sample participants are tracked for future reoffending.”

Nonetheless, two of the key limitations of the existing individual reconviction studies have been the small sample sizes and the short durations over which sample participants are tracked for future reoffending in the community. To address these concerns, we conducted a study to provide some longer-term data using an IIOC sample where follow-up rates were approximately 13 years on average and to compare rates for individuals both with and without current or historical contact offenses.

That sample consisted of 584 IIOC-only and 106 mixed IIOC/contact probationers who attended a U.K. National Probation Service community intervention. Approximately 1-in-7 of this IIOC sample had a previous contact offense, which chimes well with the previous meta-analyses that suggest around 1-in-8, but worth noting that these were proven priors and there is evidence to suggest that self-report of undetected contact offences may be higher. As Seto et al. noted in their meta, it’s not controversial to suggest that some online offenders will have committed undetected contact offenses since for most crimes there will always be a proportion whose offences go undetected.

“We found a proven reoffending rate of 24.8% for any reconviction and 12.6% for sexual reconvictions after an average of 13 years in the community.”

Two separate reconviction analyses were planned. The first was intended to establish overall reconviction rates for the IIOC and the mixed group for the longest follow-ups available. The second was a fixed 5-year analysis to draw comparisons between this sample and the findings of other previous primary and meta-analysis data. We also analysed risk assessment and psychological test data to see if any variables predicted new reconvictions. “Reconviction” was defined as any post-index offence sexual, violent, or non-sexual non-violent conviction or caution, excluding summary motoring offences and “pseudo-reconvictions” (convictions for historical undetected crimes).

Across the whole sample, we found a proven reoffending rate of 24.8% for any reconviction and 12.6% for sexual reconvictions after an average of 13 years in the community. Proven reoffending rates differed significantly between those who had known IIOC offenses only (22.4% for any reconviction and 10.1% for sexual reconvictions) and those who also had known prior or concurrent contact sexual offenses (37.7% and 26.4%). Only 2.7% of the IIOC-only group was convicted of a subsequent contact offense indicating low rates of “escalation” from IIOC to contact offences. The remainder were convicted of new IIOC offences. The fixed 5-year rates for IIOC-only offenders were also consistent with previous research (around 5%).

“Prior general offenses were found to be significantly predictive of general reconviction and prior sexual convictions were significantly predictive of sexual convictions, albeit weakly.”

Risk assessment findings concur with previous research that the Risk Matrix 2000/S is a weak-to-moderate predictor of reconviction for the IIOC population and that much of the predictive power is driven by historical offenses. Like many other sex offender risk assessment tools, it was better at measuring relative risk (i.e., who is riskier than who) but not absolute risk (i.e., even high risk offenders were still only reconvicted at low rates). Prior general offenses were found to be significantly predictive of general reconviction and prior sexual convictions were significantly predictive of sexual convictions, albeit weakly. Sadly, we couldn’t retrofit the new CPORT risk assessment, developed specifically for the IIOC population, as we didn’t have item-level sexual interest data: a requisite for scoring the tool.

Having an “untreated” profile, as determined by improved scores on a variety of relevant psychological self-report measures after participating in a treatment program, was also not found to predict future reconvictions. This was probably due to the fact that prior to the intervention, only about one-third of the sample seemed to have any problems with socio-affective functioning (self-esteem, loneliness, etc.) and only a small minority self-reported pro-offending attitudes as they related to sexual contact between adults and children.

“Having a “treated” profile, as determined by improved scores on a variety of relevant psychological self-report measures after participating in a treatment program, was not found to be predict future reconvictions.”

This mirrors previous research findings that IIOC offenders don’t inevitably hold those dysfunctional attitudes and beliefs related to sexual contact between adults and children that have been found to be criminogenic. Instead, IIOC samples seem to rationalize and excuse their use of IIOC, and post-hoc rationalizations and excuses have yet to be found to be statistically associated with reoffending and are falling out of vogue in the new world of positive psychology-informed offender treatment program.

This calls into question the value of the ongoing quest to uncover “implicit theories” that IIOC offenders hold at the level of beliefs and that provide an etiological explanation of their behavior, arguably at the expense of the quest for solid evidence of the underlying “implicit theories” concept. At least, I think it’s fair to say that the role of pro-offending thinking in IIOC offenses is due a conceptual deep-dive, since I personally don’t know of good evidence that IIOC offenders actually endorse many of the field’s traditional cognitive distortions, as opposed to being less inclined to reject them than individuals without sexual convictions.

Despite the groups appearing well-adjusted in terms of social competency and offence-supportive attitudes, approximately 80% were found to be dysfunctional on impulse control measures and appeared to remain so at the end of the program. This also mirrors other findings suggesting that IIOC offenders often have measurable problems with impulse control and that mixed internet offenders can be distinguished from IIOC-only offenders by poor self-management.

“Despite the groups appearing well-adjusted in terms of social competency and offence-supportive attitudes, approximately 80% were found to be dysfunctional on impulse control measures and appeared to remain so.”

The benefit of these blogs as compared to the slightly more constrained world of peer-reviewed publications, is that you can speculate a little more wildly. It was too much of a stretch for our reviewers, but, personally, I think these findings add weight to the notion that IIOC use is often a poor self-management strategy – a form of malign self-medication – for dealing with underlying paedophilic interests. IIOC users might typically represent a subset of an otherwise well-adjusted paedophilic population that for one of several possible reasons – a lack of self-control being one – is willing to consider IIOC use as an acceptable alternative to sexual contact.

As with all studies, limitations should be borne in mind when evaluating these findings, including (but not necessarily limited to) the following. Despite being a relatively large sample for the IIOC population, it’s still a small sample for a reconviction study. The low base-rate of reconviction also made predictive analyses difficult, notably making some group contrasts impossible. This study should not be viewed as an evaluation of the effectiveness of those community programs for the IIOC population as there is no counterfactual or comparison group. Only proven reconvictions were examined in a criminal justice sample which is likely to lead to underestimations of both prior offending behaviours and post-release offending behaviours, since offenses go undetected.

“We drew two key conclusions: (1) there is a need for comprehensive assessment of offending history – known and self-reported; (2) there needs to be a nuanced discussion on the nature and efficiency of treatment for this group, given their low reconviction rates and normal functioning in [treatment] domains.”

We echo the sentiments of Seto and colleagues that the lower reconviction rates should not be used to minimize the seriousness of these crimes. We drew two key conclusions from the findings. The first was that there is a clear need for comprehensive assessment of offending history, both known and self-reported, for this population so that sound decisions can be made for offender management. The second was that there needs to be a more nuanced discussion on the nature and the efficiency of treatment for this group, given their low reconviction rates and their normal functioning in domains that are otherwise problematic for those with contact offenses.

In closing, it’s worth remembering that to generate a longer follow-up, this sample was drawn from participants on programs between 1998-2006. As the sands of time seem to slip by faster as we get older, it’s easy to forget that back in 2006 – only a smidge over a decade ago – the iPhone was just a rumour, less than half of Americans had a high-speed internet connection, and Facebook was basically a fancy Harvard campus people-finder. The sheer speed of technological advancement online makes it incredibly difficult to be confident in what we know and don’t know about IIOC users as a population.

But, at least to some extent, the known-knowns are clear: for whatever reason, the IIOC population does not seem to be one that once convicted of an IIOC crime, is typically convicted of a new sexual crime – especially a new contact offence.

 

Suggested citation:

Elliott, I. A. (2019, April 28). Long-term reconviction rates for individuals convicted of indecent image offences appear to be low [Weblog post]. Retrieved from https://wp.me/p2RS15-i5