Last week Keith Bristow, Director General of the National Crime Agency, confirmed what had previously been disclosed on national TV in 2013 by Assistant Chief Constable Peter Davies, the then CEO of the Child Exploitation Online Protection agency (CEOP). Bristow was quoted in the following terms
I don’t think I can be more candid than say if there are 50,000 people involved in this particularly horrible type of criminality (downloading child abuse images), I don’t believe that all 50,000 will end up in the criminal justice system being brought to justice.
So far so regrettable but, as we have seen, it was not in itself a new disclosure. What made Mr Bristow’s words unusually interesting was the fact that he had chosen to raise the issue again. Taken together with the Government’s recent acknowledgement of the wider position (think Savile, Rotherham, Rochdale and the rest) that rather elevated the point, giving it new political legs and salience. Entirely predictably Bristow led the early evening news and was all over the papers the next day.
Am I reading this right?
But what Bristow apparently said next was definitely new, and at the same time far from reassuring. In effect he established or gave official blessing to the notion of a hierarchy of child sex abusers. Since it appeared in my newspaper in reported speech, not quotation marks, and it differs from the official press release I cannot be certain if the journalist got him absolutely right. Nevertheless the sentiment is clear enough. Indeed to a degree it is hinted at in the official release. Here is The Guardian‘s account
Keith Bristow…..said the police were simply unable to pursue all of those engaged in downloading child abuse images. He said law-enforcement agencies must instead focus on the highest-risk offenders, who at the top end were criminals who sexually abused children. This meant that lower-risk suspects could fall through the net. (italics added for emphasis)
Now at one level Bristow is expressing what looks like a common-sense idea. Let’s go after the baddest baddies first and to qualify for that soubriquet you have to be suspected of being a contact offender, actual or potential. I’m not sure I would define dangerousness quite so narrowly but it sort of sounds ok because it suggests we might be able to prevent more children from being abused or stop any on-going abuse that is already underway.
But in this newly declared hierarchy it seems the unpalatable truth is some of the guys at the bottom end (who presumably are judged not to be potential or actual contact offenders?) likely will never be troubled by a knock on the door. The starkness of this latter declaration shocked many pundits and members of the public alike although this has in reality pretty much been the position for a good many years. Only now are senior police officers being open about it. Not before time.
How do you grapple with such gigantic quantities of data?
The problem, though, is starting with the large volumes of data that are now flooding into police operations rooms about online accounts that are being used to engage with collecting or exchanging child abuse images, we simply do not know how, rapidly, reliably or easily, to get from an undifferentiated mass to a list of the baddest or the most likely baddest, about whom further enquiries ought to be made as quickly as possible.
The broader debate about child abuse in Britain
Rightly there is a much broader debate going on about how we try to stop child abuse happening in the first place (and see Simon Bailey’s comments) , how we help children and young people to develop the knowledge, confidence and skills to avoid being groomed, report abuse and get help should abuse nevertheless happen. That forms an important part of the backdrop to a discussion about child abuse images but there are quite distinct and particular aspects to the online images issue which need separate consideration.
Online child abuse images
In the remainder of this blog I want to focus on some of the specifics which arise in connection with the existence on the internet of extremely large volumes of images and from the fact that they are regularly being accessed by so many people who live within our borders. It is important also not to forget that because a large proportion of these images are produced and stored outside of the UK there are limits to what we can achieve here solely as a result of policies we might adopt in our green and pleasant land.
Finding the baddest from among image collectors
Faced with a new list of persons suspected of involvement with online child abuse images it is probably obvious that people with prior convictions for child sex offences ought to be high on the list for further investigation and probable arrest but please note that, historically, a goodly proportion of those who have been identified as being involved in downloading images have no prior convictions of any kind for anything.
Persons in positions of trust or authority, or individuals who work with children should also come near the top but it by no means follows that a person who does not fall into one of these first two categories is going to be any less dangerous than anyone who does.
There appears to be no common factors or themes
In fact if we know anything about child sex abusers in general and in particular guys who get involved in collecting child abuse images it is that we have not yet been able to discover any common, defining or overarching characteristics which would allow us easily to separate them out from the general population. There is no demographic that yields a larger number of baddies than any other at least not if we set the standard as being the potential for a given person to engage in contact sex offending with his own or other people’s children.
About all we do know, and we know it with a high degree of certainty, is that everyone found in possession of child abuse images, at first sight carries with them the raised probability that they are already or will eventually become a contact child sex offender. Thus all image collectors are abusers by proxy and an unknown number will be current, past or future hands on abusers as well.
Some (barely believable) studies have suggested that the proportion of individuals arrested solely for possessing child abuse images who have already abused children in real life might be as high as 80%. A CEOP study of likely contact offenders at one point put the figure at over 50%, other reports have come out around 15% or thereabouts. All of the numbers I am aware of imply that the proportion is too high to ignore if safeguarding children from future or current contact offending is at the core of your concerns.
It follows from this argument that, yes of course, if the police can assemble additional information about individuals which points to the presence of extra risk factors then that should move someone up the list for investigation and likely arrest but emphatically the absence of such additional data at the time of the initial enquiry should never be taken to imply that everyone else is probably “just a harmless saddo who is only into pictures so we needn’t fret so much”.
The images themselves tell you little
Our difficulty is that we simply have no way of knowing, solely on the basis of the types of images a person collects, the volumes, the demographic they belong to or the length of time they have been collecting, how to separate out actual or potential contact offenders from those who undoubtedly will never lay hands on a child with evil intent.
More research is urgently needed
It is widely assumed, and this is an assumption I share, that if we did enough, large scale research we might be able to get a better handle on how to spot the potential or actual contact offenders (past, present and future) or at least prove beyond all reasonable doubt that there is nothing to spot because there is no pattern to find.
If we could find a pattern then, no argument: those who comply with it are the ones to go for first but since at the moment we do not have one I am bound to feel our current selection methods are going to be random to some extent or are being tailored more to the available resources. Nothing wrong with that at one level but we should say that is the case and not try to dress it up with a spurious justification which in fact amounts to deception.
What we would all love is, based on observed behaviours, for someone to develop a magic formula that algorithmically could look at a mountain of digital data about a large number of suspects and from that in a few seconds produce a list of the ones that ought to be looked at urgently. That might or might not include everyone the police currently classify as being among the most dangerous, but I have a hunch it would draw in others.
It’s a pipe dream
Anyway, this is a pipe dream because the research has just not been done and as far as I know no one is currently doing it on a large enough scale to allow such an algorithm to be produced even though there are many academics and police officers who would give their eye teeth to get such a project going. It’s a question of money and nobody is putting it up.
In the meantime what we must absolutely never, ever do is simply look at the type of pictures a person has been collecting, or the numbers they have amassed, or the period of time over which they have been collecting them, and deduce from those facts alone whether or not the individual concerned is a greater or lesser risk in relation to current, past or future contact offending.
I have often heard it said that because so and so was found in possession of what we used to call Level 5 images – featuring scenes of sexual torture or rape of children, that the individual concerned must plainly be a greater threat to children in terms of contact offending in the past, present or future. Counter intuitive it may be, but there is absolutely no basis for such a belief.
The Bradbury lesson
Take the case of Myles Bradbury, a paediatric hematologist at Addenbrooke’s Hospital in Cambridge. In an operation led by the Canadian police information concerning Bradbury was passed to CEOP but officers graded the images as Level 1 – described as ‘depicting erotic posing with no sexual activity’ – or lower and did not pursue the matter. Bradbury was eventually convicted on 12 counts of engaging in sexual activity with a child, mostly at his hospital. Now this error was finally picked up internally by the police but the sloppy thinking which allowed it to happen in the first place is in many ways more worrying.
We must never lose sight of the importance of dealing with the images
There is yet another aspect of this discourse that is potentially worrying. Implicit in a lot of people’s thinking is that the images matter only insofar as they point to victims who can be identified, located and rescued, or they enable the police to identify persons who are currently involved in child abuse, have been in the past or are likely to be in the future.
Of course nobody would argue with any of those objectives but the point is this relegates the images, as such, to a level of instrumentality. The images matter only in so far as they are useful because they serve a law enforcement end. If they do not then maybe it’s not so important to get them deleted or to ensure that access to them is denied. I have even heard it suggested that in some ways it is a great advantage to the police to have known images out there because it allows them to identify potential offenders whom they might otherwise never detect. Assuming for the moment there could be something in that argument (and I don’t think there is) we now know the police and the criminal justice system as a whole cannot cope with the existing volumes of offenders so an argument that rests only on the premise that such an approach will help them find more offenders or suspects really falls at the first hurdle.
Images are a cause and an effect
For a good many men the images are both a cause and an effect of paedophilic behaviour (and that is true whether the images are real or pseudo). So even if the police are not able to identify, locate and rescue every child in every image, or even if it is impossible to identify, locate and arrest every perpetrator or downloader, that does not mean there are two classes of image: those with some investigative potential and those with none.
All child abuse images matter for at least two reasons: as previously noted, real or pseudo they are both a cause and an effect of paedophilic behaviour which by definition is likely to put children in danger both now and in the future. Then in the case of real images there is the question of the rights of the children depicted in them. They have a legal right to human dignity, to say nothing of the egregious breach of their right to privacy. We cannot allow or be indifferent towards the continued circulation of these images or keep them in circulation as bait not least because they put the child at risk of further abuse or harm both in the here and now, and also potentially into the distant future ( see Amy’s case).
This is not a problem the police can solve
For these reasons I do not see the problem of images on the internet as being principally a problem for the police. Yes we need agreed mechanisms for determining whether or not particular images would be illegal in each participating jurisdiction but beyond that and after that the challenge lies with the internet industry to stop or reduce the traffic. The internet industry is central to this project and happily most of the large and responsible players are up for it.
We need to employ more police officers but that alone will never do it
In some quarters there is undoubtedly a degree of ambivalence towards the idea of looking to technology to solve the problems presented by the extremely large quantities of child abuse images being circulated on the internet. Undoubtedly some would rather we simply increased the number of police officers employed on this work so that anyone who engaged with the images might reasonably fear that sooner rather than later there would be a knock on their door following which they would be arrested then convicted. I do not doubt there is an extremely strong case for more resources to be given to the police to fight this type of crime and I hope that Theresa May’s different reviews and enquiries reach the same conclusion. But let us not kid ourselves. Even if we were living in times of super abundance, as opposed to times of austerity, we still need to suppress the trade in images and that is first and foremost a challenge to the industry not the police.
And it involves the whole criminal justice system
Moreover it is important to remember it is not just the numbers of trained police officers that is important here. With an increased number of arrests for image related offences would also come an increase in the demand for forensic examinations of seized equipment, an increase in the demand for staff capable of carrying out psychological or other assessments of those arrested, an increase in probation and prison staff, possibly also an expansion in the number of prison places, not to mention an increase in the number of court rooms, judges, lawyers and associated staff to make all this work. As I have said, it is doubtful, in the midst of a global recession, whether or not it is realistic to expect countries in the richer parts of the world to be able to contemplate the sorts of expansion in public expenditure such an approach suggests but it is surely completely unrealistic to expect less prosperous parts of the globe to be able to do likewise.
I am not sure when, exactly, it became clear that the police had been beaten by the volume of offenders and images but I sincerely doubt that it is a recent phenomenon. Look at Operation Ore. That started back in the 20th Century when the UK police were handed over 7,500 names and credit card numbers of guys who appeared to have been buying child abuse images over the internet. In other words this should have been easy peasy. No question of time consuming tracking down of names and addresses of users where the only information available was an IP address. The credit card number takes you straight to an individual with a specific home address. Yet it is understood that between 2,000 and 3,000 of the 7,500 never had any contact with the boys and girls in blue.
I said earlier that I thought the timing of last week’s announcement was interesting in the context of the current debate in the UK about child abuse in general. I suspect over the coming months and beyond we are going to learn a lot about several shortcomings and failures on the part of law enforcement and others in relation to how they have handled activity in this space but it would be both wrong and unjust to lay all the blame at the door of the police, which was how things were shaping up. If the police have erred at all it is because they did not speak out publicly sooner.
Time for radical experimentation?
However, rather than try to allocate blame and start name calling we should all focus on working out what we do next. And in that regard I think we need to get ready for perhaps some radical experimentation. If we cannot arrest everyone, surely there are other things we can do which might persuade people to cease their involvement with collecting or distributing child abuse images online? Maybe this is a second best outcome, but it is definitely better than no outcome at all.
Desiderata 