I thought I was unshockable…..

In 1982, long before the internet as we now know it arrived in our midst, the US General Accounting Office advised  the Subcommittee on Juvenile Justice of the House Committee on Education and Labor that

As a result of the decline in commercial child pornography, the principal Federal agencies responsible for enforcing laws covering the distribution of child pornography – the U.S. Customs Service and the U.S. Postal Service – do not consider child pornography a high priority.

Things changed.

In 1995, arguably the modern Internet’s Year zero, in Greater Manchester, a major conurbation in England, the police seized 12 unique child sex abuse images. The UK police say at that time they knew of 7,000 in total. Inexplicably, in the same year Interpol claimed to know of only 4,000. In both sets “just” a few hundred children were depicted. That’s in the whole of recorded history up to that point and across the entire planet.

Things have changed again. And, once more, not for the better.

Thanks to some excellent work by the NSPCC, aided and abetted by brilliant journalism on the part of The Daily Mirror, it feels like several pieces of a complex jigsaw have started falling into place. The picture which is emerging is profoundly disturbing, verging on completely unbelievable.

FOI requests

The NSPCC sent out Freedom of Information requests to all 43 local police forces in England and Wales. The results form the basis of the first part of the Mirror’s story. It appeared yesterday (Monday). There is more today and a final article tomorrow.

Inter alia the NSPCC asked the local forces how many indecent photographs of children (child sex abuse images), as defined by the Protection of Children Act, 1978, each of them had seized between March, 2010 and April 2012. Five forces replied in the time available.

Police Force

Images Seized




















England & Wales


The population figures for each police force area and the total for all areas are based on mid-2010 estimates. We can see that between them the five forces had 4,006,600 inhabitants. This represents 7.25% of the entire population of England and Wales. A very big sample.


It’s hard to know why the 38 absentee forces did not reply. Maybe if they had had a little more time there would have been more responses, so for now I’m going to assume the sample provided by the five that did reply is not significantly skewed one way or another. It understates the number of people who live in major conurbations but I don’t think one’s place of abode is going to be a major factor in determining the likelihood of one being involved in downloading child abuse images.

As you would expect, police force size broadly corresponds with the number of people living in an area. However, different police forces give a higher or lower priority to online child protection work and it is not clear why. Not all forces have a dedicated online child protection unit or team but across the whole country I think the picture will even out. Hyper active police forces will make up for less energetic ones and vice versa.

Anyway what we are trying to glimpse here, in a sense, is not what the police are doing but what the people living in their operational areas might or are likely to be doing. What we know about the individuals who are arrested for child abuse image offences is that they come from all walks of life. Thus, assuming the demographics of the five forces’ areas combined are not massively different from the demographics of the rest of England and Wales and, once more, given the size of the sample, I’d say we are good to go.

Astonishing figures

Simply ramping up pro rata the number of seized images suggests something like 360 million might have been seized overall across England and Wales in the two years in question. If you exclude Cambridgeshire’s numbers as being somehow freakish you end up with 8,837,511 images from 5.8% of the population. That still produces a potential national total in excess of 150,000,000.

I can think of no good reason why we should exclude Cambridgeshire’s numbers but in a statistical exercise like this eliminating the extremes at both ends is not an uncommon practice. If you did that here, and excluded both Cambridgeshire and Nottinghamshire, the population of the remaining three forces would be 2,130,300. That is 3.8% of the total population. They yielded 8,600,655 images to seizures by the police. This would imply a total of 226, 333,026 images being seized by all forces.

The Met, the police force for London, are known to do an enormous amount of work in this area but we don’t have a reply from them. If we assume police in London only seized the same number of images as Cambridgeshire, but from a much larger population base, you also come out with a number in the 200 millions so maybe that is nearer the mark. As an aside, assuming people living in England and Wales are no more or less likely to engage in downloading child sex abuse materials than inhabitants of other countries with similar levels of fast broadband connections, it ought not to be too difficult to make an educated guess about the number of images in circulation worldwide.

In the quote I gave to the Mirror I took an ultra-conservative line. I said I thought the number of images likely to have been seized, or which could have been seized, across the whole of England and Wales in the period in question was very definitely going to be north of 100 million. Somehow that didn’t make it into the paper but given we were starting with a definite base of 26 million that had actually been seized I don’t think I was being too daring or sensationalist. Even 26 million is a troubling number so anything above that will only be more so.

What do the numbers tell us?

Some elements within law enforcement are very dismissive when it comes to discussing the number of images being seized or being circulated.  They point, for example, to a documented a case where one UK man was found with 2.5 million images. I’m told that in Cambridgeshire one of the people they arrested had five million images on his machines. If that stands up it could be a world record for an individual as up to now I had always thought that was held by Canadian Arthur Sayler with four million.

The implication is obvious. What matters much more is knowing how many individuals are involved in downloading images and the number of child victims depicted in those images. I get that and will come back to both these points.

Nevertheless, despite the caveats, there is a kind of macabre poetry in the mind-boggling figures shown. Accepting that knowing about the numbers of images being seized or in circulation will never tell us the full story, they definitely tell story. In this case the enormous numbers send out a very clear message that something has changed within society, something big and bad has happened and I’m not sure we have properly understood or come to terms with it. 

The link between possession offences and contact offending

Before going on we need to register one important recurring issue in this debate. “Simply” downloading or possessing a child sex abuse image is a criminal and abusive act in and of itself without more.  It violates the rights of the child in the image and because it encourages the production and distribution of more of the same or worse it places children all over the world at risk.

Nevertheless we  also know that a high proportion of people who download or possess child sex abuse material will never, themselves, go on to engage in contact offending of any kind against children.  It may raise other issues about them as parents, teachers or whatever but nothing which instantly suggests a child in their social, professional or family circle is in immediate danger of being sexually abused.

However,  countless studies have shown there is a raised probability that a person involved in knowingly downloading or possessing child sex abuse material will engage in contact sexual offending against children, their own or other people’s, or they will have offended against children in the past. There can therefore be no question, from a child protection perspective,  we would prefer law enforcement had the resources to address both categories of offenders. Since we know, in these times of austerity, that is simply a pipe dream, finding offenders in the second group needs to be  prioritised

Images and degrees of danger

The problem is how to do that. We need a way of grading “dangerousness”. The volumes in a person’s possession are no guide. At the time of the police’s initial encounter  with someone found with but a single image, it could easily be the case he is every bit as dangerous or be more dangerous than someone found with millions. The  latter may just be an obsessive collector.

The “horribleness”  or violence of an image, equally, does not necessarily tell you very much about a person’s dangerousness.  This may also sound counter-intuitive but it is nevertheless a fact. Men found with legal pictures of children wearing  bathing costumes can be every bit as worrisome as someone found with tens of thousands of pictures of children enduring the worst forms of sexual torture.

From a public policy  and child protection perspective this is bad news.   Absent a reliable way of assessing and grading dengerousness the only 100%  safe working assumption is anyone and everyone found in possession of an image has to be viewed as being a potential contact child sex abuser. A safeguarding investigation ought to be launched in respect of every child with whom he is or has been in contact. The enormous number of cases now arriving at law enforcement’s door means there is no way, in any reasonably foreseeable future, that is going to happen but we have to keep such an objective in our sights.

Is there a way, merely from looking at an individual’s stockpile of images, that the police can quickly and reliably identify  and separate out the more dangerous from the less dangerous? This is the Holy Grail. It would help the police, the courts, the prisons and the probation services to target their scarce resources more effectively. Kent police are leading an EU-wide and EU-funded study which looks extremely promising. Fingers crossed.

How many images and how many children?

Whether the number of images circulating in England and Wales is 26 million, 200 million or 360 million that does not mean there are 26, 200 or 360 million unique images going around. Whatever the true number there will be tens upon tens of millions of duplicates of the same images.

We noted earlier that in 1995 there were thought to be up to 7,000 unique images involving a few hundred children. That’s as close as we will get to any kind of historic data. Nothing more reliable is available and the police are still rather coy about firm numbers. However, not very long ago,  it was being suggested  in law enforcement circles there were now probably about 3,000,000 unique images known to the police around the world, which in practice means in INTERPOL’s database or the databases of a comparatively small number of countries and organizations that currently have one.

Equally, there is still no certainty or precision about the number of individual children in the known unique images. One suggestion was there is about 150,000 or thereabouts.  Sexting images are now starting to make their way into the body of illegal material being exchanged by paedophiles and collectors. On top of that there is a growing volume of child abuse videos. The police have thousands of hours of material of this type.

The above notwithstanding, no one I know suggests the number of children who have been identified and rescued from stills or videos, worldwide, will be much above 10,000.

The UK is starting to publish some hard numbers. In 2012-13 in their Annual Report CEOP indicated

790 children were safeguarded and protected from sexual abuse as a result of CEOP intelligence and operations

although it is unclear how many of the 790 were connected to or arose from a case that started with a report of an image.

How many people are downloading child sex abuse material?

Writing in 1997 Sir William Utting, in People Like Us” described the production and distribution of “child pornography” as a “cottage industry”, meaning it was small scale and highly localised. That year in England and Wales 184 people were convicted for offences relating to the possession, making or distribution of child abuse images. As I have already observed, today we are in an entirely different place. Convictions are now approaching 2,000 per annum and last year 2,312 people were arrested.

Returning to the number of images shown above, it is entirely possible that even if the number of images seized in England and Wales is as large as 360 million it could still represent the aggregated activity of only a comparatively small number of people.  We just have no sure way of knowing. Or do we? There are straws in the wind. Read on.

A network of abuse

I now want to put another piece of the jigsaw in place. Having thanked the Mirror for their part in drawing this matter to the attention of the wider public and, we hope,policy makers who are in a position to do something about it, here I must acknowledge the part played by another newspaper, The Sun. Although the police reconfirmed the same information to the Mirror, it first appeared in an article in The Sun on 1st July, 2012. I do not know how I missed it. Actually I do. It appeared on page 21 and was not promoted by them or picked up by other parts of the media. It’s amazing the story didn’t spark a riot.

I said earlier we don’t know how many people in total might be involved in collecting child abuse images and I also referred to the current level of arrests and convictions, both hovering around the 2,000 mark.

Now sit tight. The Sun’s article features Peter Davies, Head of CEOP, the UK’s lead police agency for online child protection. Peter acknowledges for the first time in public something insiders have known for a while. The police have the technical capability to monitor the number of people exchanging child abuse images over Peer2Peer networks. And they have been using it. If you want to know more about how they do it watch Prime Time Investigates on Irish TV. We use the same techniques.

On 1st July the British police revealed to The Sun that their estimate of the number of people involved in exchanging child abuse images over Peer2Peer networks in the UK is

50,000 to 60,000

I do not know if the 50-60,000 number derives from only one Peer2Peer network or if it is culled from several. For the moment let’s assume it is several, or even all of them though that is rather unlikely.

Down from 80,000

Originally I had heard unconfirmed reports the number of people in the UK using these networks to exchange child abuse images was around 80,000. I now understand that on further analysis, to eliminate possible duplicates and other potential errors CEOP shaved off 20 to 30 thousand. I’m guessing some of these calculations are a bit rough. We could be plus or minus several percentage points but not an order of magnitude.

Seven times larger than Operation Ore

Think about that for a moment. Let’s take the lower number and make 50,000 the baseline. As we have seen, we have yet to break the 2,000 barrier in terms of convictions in any year and the number of arrests is not much higher. And 50,000 is roughly 7 times larger than the 7,200 names that formed the basis of Operation Ore. Operation Ore nearly brought parts of our criminal justice system to a full stop. 

The arrests and prosecutions that were made under Ore were spread over the best part of three years and not everyone on the list of 7,200 received a knock on the door never mind got arrested. Not by a long chalk.

The unvisited

I have heard various estimates of how many Ore suspects were “unvisited”. According to Ore’s Wikipedia entry about 3,000 of the 7,200 Ore suspects never heard a dickie bird from the police. That number is disputed in some quarters. The suggestion is “only” 2,000 suspects remained “unvisited”. Nobody says the number was small. In other words between 60% and 70% of the suspects received a visit from the police. 30% to 40% did not.

Returning to the 50,000, Peter Davies tells us

We can zoom in on the street they live in. They have no idea we are watching them.

Peter goes on to say these guys should know they

….have nowhere to hide. We will hunt you down.

That’s fighting talk

I’m afraid there is nothing in the article to give us any reason to believe the bulk of the 50,000 will be “hunted down”, at least not any time soon.

Peter Davies reassures us the police are engaging in a range of operations to apprehend perpetrators, and I know that is true, yet I picked up no sense whatsoever that our police service as a whole is equipped to tackle the problem on the scale suggested by CEOP’s revelations. CEOP’s staff resources are low and appear to be getting lower. It’s the same with many local forces. I think a lot of people would take an enormous amount of convincing before they believed CEOP’s transition into the new National Crime Agency is in any way going to compensate.

If it took the police the best part of three years to knock on the doors of the Ore suspects they decided deserved a visit, and upwards of 30% were never seen at all, what do we imagine is going to happen facing a number like 50,000?

None of this is the fault of the police. I believe the police have never been given the right level of resources to do the job that needs to be done. What I find slightly disappointing, though, is that the police service does not make more of a fuss about it. If a Cabinet Minister swears at one of the boys in blue it is headline news for days with the Police Federation calling for a resignation. Chronic under resourcing of child protection which leaves children at risk struggles to get air time. It’s a funny old world.

It gets worse

If you read the article in The Sun you will see Peter Davies also quotes a figure which was originally published at paragraph 20 of a CEOP Report that came out earlier in the year.

Around 55% of the people who view (child abuse images) go on to commit child abuse.

Of course viewing the images is itself a form of child abuse. What is being said here is that 55% of people who view child abuse images will commit a contact offence involving a child.

Some very knowledgeable people in the child protection space do not accept that 55% is a reasonable proportion. They say it is way too high. Indeed, moving from the quote in The Sun to the CEOP Report where it came from CEOP makes clear there are other possible outcomes. These indicate proportions that could be much higher or much lower and they make clear all they are establishing is a correlation, not causation. That’s as maybe. For now let’s just accept that 55% reflects the police’s own assessment of the situation. However, we should also note the following statement which appears in paragraph 24

The link between viewing (child abuse images) and contact offending, identified but not yet quantified by research, highlights the need to consider each possession offender as a potential contact offender to some extent.

What if….?

I have often wondered what the public outcry would be like if any of the Ore “unvisited” had gone on to commit offences which resulted in a child being harmed and it emerged that the perpetrator had previously come to the police’s attention because he was on the Ore list but the police had done nothing.

By the way, as I understand matters and for the avoidance of doubt, it would almost certainly be inaccurate to say that 55% of the “Ore unvisited”, which would be between 1,100 and 1,650 people, were likely to have gone on to commit hands on offences against children.

The reason for that is the police first carried out a triage exercise. They analysed all the data they could collect about the 7,200 and prioritised visiting and arresting people who were in a position of trust or had access to children through their work (priests, teachers, youth club leaders and the like), who had previous convictions for child sex offences or where it was known children were living in the same house. The 2 to 3,000 Ore unvisited were therefore not a random cross section of Ore suspects. They were the ones judged least likely to be any kind of risk to children.

Matching Ore will not be good enough

In the un-sifted, non-triaged 50,000 Peer2Peer users, if CEOP’s meta-analysis is to be believed, there could be up to 27,500 child abusers in amongst them. It beggars belief.

Even if the police were only to repeat the same proportion of arrests they managed in Ore, we are still talking about them visiting between 30,000 and 35,000 people in order to find the 27,500 future or current child abusers. I wonder if anyone thinks that will happen either at all or within any sort of reasonable time scale?

And just to depress you thoroughly and put this in some sort of context, here’s an interesting factoid. On 12th October, 2012, the operational capacity of the prison system in England and Wales was 91,000, and it was pretty much fully taken up. Stick that in your pipe and smoke it. What would we do with 27,500 newly convicted child abusers? Could that daunting number be in any way influencing the police’s or the CPS’s operational policies or priorities?

As a society we are going to have to look again at the role of preventative measures.  As a police officer in The Mirror article mournfully accepts

We cannot arrest our way out of this

If it is beyond us to arrest these guys when they commit offences, what are we doing to divert them or help stop them getting involved with child abuse images in the first place? How much more do we need to do to reach out to children to help them resist abuse or report it as soon as it happens? How do we get major institutions to put in place better safeguarding measures? Shades of Jimmy Savile and his 40 year reign of abuse at once spring to mind.

Organizations like the NSPPC and the Lucy Faithfull Foundation have been speaking about this for years and have had woefully insufficient recognition and support from the authorities, local and national.

Improved technical measures are essential

Thanks to the IWF and the approach of the internet industry as a whole in relation to blocking access to Newsgroups and web sites containing child abuse images, that part of the challenge has been substantially dealt with. But as we have just seen we are nowhere near solving the problem of Peer2Peer networks. The IWF has no remit to tackle Peer2Peer. It is an altogether different kettle of fish. Improved technical measures will have to be part of the answer. To return to those haunting words

We cannot arrest our way out of this

Part of a new approach?

Now we go back to The Daily Mirror and the final bit of the jigsaw. As I understand things in tomorrow’s edition The Mirror will bring into the public domain information which, hitherto, had again only been shared within a limited circle. 

The Metropolitan Police confirmed to the Mirror that they had run a small experiment. During it, instead of going to arrest people whom they suspected of some level of involvement with child abuse images they sent them letters. Seemingly the letters did not renounce the possibility of police officers going to arrest the addressees at some later date. They were meant as a warning shot.

The Met did a risk assessment before they sent the letters to the individuals. The circumstances in which they were dispatched appear to be quite limited. Only 11 letters went out and the experiment is now over.  I believe at least one other local police force is doing or has done the same thing, perhaps on a larger scale or on a more sustained basis.

At first sight this practice will strike many as bizarre or dangerous or both. I don’t think anyone should try to pretend or dignify it by saying it is the best possible outcome in the best of all possible worlds. This is a response by part of the police service to the overwhelming mismatch which exists between the level of resources available to them and the number of cases piling up. They are managing a crisis.

These crimes will only be solved, children rescued, perpetrators apprehended, through a combination of technology and people sitting at a desk looking at computer screens. This is not “visible policing” or “Bobbies on the Beat”. Some of the victims or perpetrators that might have to be investigated or located may not even come from your police area! In some forces that alone can make a case fall to the bottom of the list.

Chief Constables do not get Brownie points with the Home Secretary for delivering on online child protection. Online child protection, indeed child protection generally, is not a national policing priority against which a Chief Constable’s performance is measured. It should be.

However, with the letter strategy, the Met seem to have concluded it is better to do something which might result in a person ceasing to engage in criminal activity than it is to do nothing in order to maintain the manifest pretence that they are on top of things, completely in control.

I have spoken about this issue to a number of professionals who work with child sex offenders with convictions for downloading child abuse images. They claim to be 100% certain that if some of the people they are now dealing with had at any stage received a letter of the kind I have just mentioned it would have given them such an almighty scare it would have instantly jolted them out of their illegal conduct. As a result fewer children would end up being abused to create images for them to consume, it would have pushed several potential offenders away from a course of conduct which would otherwise lead them to engage in the hands on abuse of their own or other people’s children.

I attended a meeting at Interpol’s global HQ in Lyon not that long ago. I raised the question of these types of letters. The idea went down like a lead balloon. Some overseas police forces made clear that in their countries it would, in fact, be a crime to send an epistle of that sort. It would be tantamount to warning a possible criminal of a potential arrest, giving them an opportunity to abscond and destroy evidence.

I get that. I sympathise. Sending warning letters to potential suspects is used in other parts of policing so it is not a wholly unheard of notion, although I can see that when it comes to protecting children people will worry it is a step too far.

However, I do not sympathise with or get, in effect, deceiving the public into thinking all is well when it plainly isn’t. Policing has to be about more than finding new and ingenious ways of managing an Inbox without alarming anyone. We’re talking about children, not public relations.

Returning to the UK I refuse to believe, for example, that if our Ministers were fully seized of the current situation they would simply say “Carry On”. Somehow, somewhere, this problem needs to be debated openly and honestly with all the interested parties. Even if austerity means our options are limited for a while the ad hoc nature of policy making up to now is far from satisfactory.

The internet is not to blame for this state of affairs. People do bad things, not otherwise inanimate machines. But what undoubtedly is the case is that whatever we have all been trying up to now to deal with this problem it ain’t working. Government, police, the internet industry, NGOs, must think again. Either we need to redefine the problem or we have to come up with new and convincing strategies for dealing with it as it is.

About John Carr

John Carr is one of the world's leading authorities on children's and young people's use of digital technologies. He is Senior Technical Adviser to Bangkok-based global NGO ECPAT International and is Secretary of the UK's Children's Charities' Coalition on Internet Safety. John is now or has formerly been an Adviser to the Council of Europe, the UN (ITU), the EU and UNICEF. John has advised many of the world's largest technology companies on online child safety. John's skill as a writer has also been widely recognised. http://johncarrcv.blogspot.com
This entry was posted in CEOP, Child abuse images, Internet governance, Regulation, Self-regulation. Bookmark the permalink.