Statistics Laundering: false and fantastic figures

Graham, I.; Jan 03 2010

Type of Work

Essay, research report

Notes:
- This is the second version of this article. In the first version, sent to me by e-mail, none of the many links worked. In the html version I saw ".../.../.../...". In this version all links are removed - Ipce webmaster.
- The References and Notes are given here in a separate PDF document:
  graham_refs_and_endnotes.pdf

I. Graham, 25 Nov 2008, later reviews until 2010.

I think we have what I would call a case of information laundering: You state a figure on something, somebody else quotes it, and then you and others [quote] it back, and thus it becomes clean and true. ... Perhaps this discussion will help instill more rigor in the future.

Alexander Seger, Head of Technical Cooperation, Dept. of Crime Problems, Council of Europe, April 2006

This research paper contains information about various alarming and sensational, but out-of-date, false and/or misleading 'statistics' concerning the prevalence of 'child pornography' material on Internet Web sites, etc., which appeared in Australian media reports/articles, government agency reports, etc., in 2008 and 2009.

While sometimes statistical exaggerations are not important, those referred to herein are being used to directly exaggerate the prevalence and hence risk level of certain threats, and to indirectly weaken the position of those attempting to critically assess the nature of the threats, and whether proposed public policy solutions are effective and proportionate.

Often these 'statistics' are promulgated without source citation.

Section 2 of this paper identifies the origin/source of such 'statistics' and documents the history of same where 'information laundering' has occurred (i.e. the source or detail has been changed in the course of repetition).
Section 3 provides information about estimates/statistics originating more recently, and which appear more likely to have some connection with reality.
Section 4 documents some examples of Australian media misreporting about web sites and police operations which may contribute to false perceptions about the number of web sites containing child sexual abuse material.

This paper does not contend that there is not a serious problem in relation to the use of Internet technologies to distribute/obtain child sexual abuse material. The core point made herein is that the use of web sites for such purposes has long been, and still is being, vastly exaggerated in the media, by advocacy organisations, etc.

Meanwhile little if any attention has been given to credible evidence that there is a vastly larger problem involving the use of non-Web Internet technologies which will not be affected in any way by the Federal Government's plan to spend AUD$ 44.5 million on 'blocking' of accidental/unintentional access to web sites. The problem can only be reduced by better funded and resourced specialist units of law enforcement agencies.

Commerce and Number of Web Sites/Images

"it is estimated that 100,000 commercial websites offer child pornography"
"more than 20,000 images of child pornography are posted on the internet every week"
"child pornography is one of the fastest growing online businesses generating approximately $US 3 billion ($ 3.43 billion) each year"
"child pornography is a $ 20 billion industry worldwide"
"approximately 20% of all Internet pornography involves children"
[Added 30 Nov 2008]

Predators, etc.

"40 per cent of arrested child pornography possessors sexually abused children"
"50,000 predators prowling for children online at any given time" [Added 30 Nov 2008]
"over half of 11-to-15 year olds surveyed who chat online are contacted by strangers"
"40 per cent of children who chat online said they had been contacted by someone they didn't know" [Additions: 4 Jan 2010]

Access "attempts" (hits):

"British Telecom (BT) blocks over 35,000 attempts per day to access child pornography websites"
"During 2006, the Norwegian system blocked 1.7 million attempts to access child pornography. The Swedish system blocked 15,000 attempts during its first few weeks"

Recent (2007-2008) 'Statistics' - possibly more realistic

Web sites: US Customs' officers' estimates 2007-2008
Web sites: U.K. Internet Watch Foundation ("IWF") 2008 & 2007 statistics [Updated 30 Apr 2009]
Peer to peer (P2P) networks: Special Agent Flint Waters (Lead Agent for the Wyoming Internet Crimes Against Children Task Force) statistics 2008

Media Misreporting about Web Sites and Police Operations

March 2008: Operation Achilles - Australia, USA, UK, Germany, etc.
June 2007: Operation Lobate-AU, Operation Chandler-UK, Project Wickerman Two-CA
June 2008: Operation Centurion-AU [Added 29 Nov 2008]
December 2008: Operation Resistance-AU; September 2008: Operation Carousel-Brazil, Spain [Added 23 Apr 2009]

1. Introduction & Background

On 8 January 2008, The Australian national newspaper published a highly controversial opinion article by Bernadette McMenamin, CEO of Child Wise/ECPAT in Australia, (Filters needed to battle child porn [*1]).

(Ms. McMenamin is also a member of the Federal Labor Government's 'Consultative Working Group to improve Cyber-Safety' announced in May 2008.)

Ms. McMenamin opined that persons opposed to the Australian Government's mandatory ISP-level filtering plan [*2] must not be a "decent human"; misrepresented the purpose of the government's plan; and promulgated sensational statistics without citing sources. Among other things, Ms. McMenamin said:

"Let's put this argument into perspective. Child pornography is one of the fastest growing online businesses generating approximately $US 3 billion ($3.43 billion) each year. It is estimated that 100,000 commercial websites offer child pornography and more than 20,000 images of child pornography are posted on the internet every week."

In 2005 the United States National Center for Missing and Exploited Children revealed that 40 per cent of arrested child pornography possessors sexually abused children.

Source: "Filters needed to battle child porn", Bernadette McMenamin, CEO of ChildWise (ECPAT in Australia), Opinion, The Australian, 8 January 2008

The alleged 'statistics' in the first paragraph above also appear, without source, on a Child Wise website page encouraging readers to sign a Child Wise petition to the Federal Government recommending government mandated ISP-level filtering (as at mid November 2008).

Ms. McMenamin's opinion article resulted in, among other things, numerous long term Internet users questioning the source of such 'statistics' in online discussion/commentary fora because they had never seen such material online. Such questions were still occurring in November 2008. The short answer to such questions is that the 'statistics'/'estimates' in the first paragraph quoted above have no factual basis, and were invented at least 5 years ago.

Perspectives

As this writer is opposed to the Federal government's AUD$ 44.5 million plan to 'block' access to 'child pornography', 'X-rated material', 'violence', 'prohibited' material, 'inappropriate' material and 'unwanted' material on a secret blacklist compiled by a government agency, which at best can only prevent accidental/unintentional access to content on web sites, presumably the writer is one of the people regarded by Ms. McMenamin as an [in]decent human.

Nevertheless, this writer agrees with Ms McMenamin's apparent opinion that arguments should be put into perspective. In that regard, for example, in 2005 the United States National Center for Missing and Exploited Children ("NCMEC") revealed that in a one year period there were an estimated 1,713 arrests for Internet-related crimes involving the possession of child pornography (of which 55% also involved sexual victimization of children or attempts to do so), and an estimated 65,000 arrests for all types of sexual assaults committed against minors [*3]. The NCMEC did not, however, widely advertise the latter and failure to give equal prominence to the much larger contextual statistic undermines a reader's chance of putting the Internet-related crimes statistic into perspective.

Exaggeration is not a new phenomenon

The writer is of the view that public policy should be evidence-based, not based on myth, fiction, fantasy, exaggeration, or misrepresentation of academic research findings, promulgated by advocacy organisations (whether overseas-based or Australian-based) no matter how well intentioned any such organisations may be.

In April 1992, Kenneth Lanning (then Supervisory Special Agent, Behavioral Science Unit, Federal Bureau of Investigation Academy) remarked in Child Sex Rings: A Behavioral Analysis For Criminal Justice:

"Some professionals [dealing with child sexual abuse], however, in their zeal to make American society more aware of this victimization, tend to exaggerate the problem. Presentations and literature with poorly documented or misleading claims about one in three children being sexually molested, the $ 5 billion child pornography industry, child slavery rings, and 50,000 stranger-abducted children are not uncommon. The problem is bad enough; it is not necessary to exaggerate it. Professionals should cite reputable and scientific studies and note the sources of information. If they do not, when the exaggerations and distortions are discovered, their credibility and the credibility of the issue are lost.
Source: Child Sex Rings: A Behavioral Analysis For Criminal Justice, Kenneth Lanning, Behavioral Science Unit, Federal Bureau of Investigation Academy, April 1992 [*4]

Unfortunately, little if anything has changed in that regard since 1992. Also unfortunately, since then the Internet has become awash with false, exaggerated and misleading 'statistics' about child exploitation/abuse, some of which falsely purport to originate from a credible source - a tempting resource for people seeking a number to prop up an article, media release, proposal or argument.

While the promotion of exaggerated, or even false, 'statistics' on such issues is not a new phenomenon, there appears to be a growing trend in Australia of giving new life to out-of-date 'statistics' that have previously been found to be false or misleading. Many of these 'statistics' originate from child protection campaigners/organisations in the USA and UK which need to keep the public scared in order to maintain/increase their funding by government and/or public donations - members of the "Fear Industrial Complex":

"[T]he fear industrial complex is composed of politicians, activist groups and corporations that all sell us on the idea that they can provide safety from the very dangers they are scaring us about.

Whenever somebody's trying to scare us, the question to ask is: Are they benefiting from it, and in what way?" said [Barry] Glassner [author of The Culture of Fear: Why Americans Are Afraid of the Wrong Things]. "If they're selling us a product, if they're selling us their political campaign or their cause or whatever it is, we should ask how big is the danger, really? Is it big, is it small, or is it just that they stand to benefit by making us scared?"
Source: "The 'Fear Industrial Complex' How the Media, Government and Corporate America Bank on the Business of Fear", John Stossel & Natalie D. Jaquez, ABC News 20/20, 23 Feb 2007 [*5]

Some statistics invented/promulgated by such organisations have since been the subject of information laundering with the result that they are sometimes incorrectly attributed to law enforcement, or other government, agencies. Moreover, some politicians and some law enforcement agency officers, including within the Australian Federal Police, promulgate false and misleading 'statistics' in media releases, speeches, etc. Perhaps that occurs under an assumption that some overseas law enforcement agency did in fact 'estimate' the number, perhaps it does not. Whichever is the case:

"Unfortunately, the mere fact that a number has proliferated, even at the highest levels of officialdom, does not demonstrate the number is true."
Source: "Don't believe the hype" (An edited extract from Risk: The Science and Politics of Fear by Dan Gardner), Sydney Morning Herald, 26 April 2008 [*6]

Fear of Crime & Damaging Consequences of Internet Fear Mongering

Some argue that sensationalism, exaggeration and even lies are justified for the purpose of attracting attention and raising public awareness about child sexual exploitation/abuse. However as Associate Professor Stephen Smallbone (Program Leader: Preventing Sexual Violence and Abuse, Griffith University) has observed:

"Publicity surrounding Operation Auxin [October 2004], as with much crime reporting, is something of a two-edged sword.
- On one hand, the media attention given to this event is likely to have a significant general deterrence effect.
- On the other hand, since the emphasis is on reporting but not understanding the phenomenon, the community can be left with a feeling of increased threat.

Because the threat is largely unspecified and unexplained, the community can also be left feeling that it has no means to protect itself. Thus the gap between the actual and perceived risk of crime -- what many refer to as the fear of crime phenomenon -- is effectively widened.

Media attention to events like Operation Auxin -- and more broadly to sexual crime in general -- could provide a better public service by helping to specify and explain the threat.

But the problem does not rest with journalists alone. Many journalists rely on the opinions of academics, professionals and interest groups, who themselves bear responsibility for their contributions to the public debate. Unfortunately, there are many agendas that are served by selective information, and in some cases by plain exaggeration.

While the media can play an important role in preventing crime, it also can play an important role in preventing the fear of crime. While one sexual offence is one too many, we need to keep a sense of proportion."

Source: "Fear and loathing", Stephen Smallbone (School of Criminology and Criminal Justice, Griffith University), The Courier Mail, 6 October 2004 [*7]

Moreover, the saturation of fearmonging 'educational' or awareness raising media reports and advocacy organisation campaigns may result in not only adults, but also children, 'switching off'. As Nancy Willard (Director, Center for Safe and Responsible Internet Use and author of Cyber-Safe Kids, Cyber-Savvy Teens: Helping Young People Learn to Use the Internet Safely and Responsibly) warned in December 2007:

"Internet fear-mongering is undermining our ability to effectively address the risks to young people online. In their zeal to address these concerns, some individuals and organizations are exaggerating the problem. They present poorly documented statements and misleading claims. Here is one example: One in seven young people has been sexually solicited online. ...

Young digital natives are dismissing fear-based Internet safety guidance because they perceive that adults fear what they do not understand. As a result of this fear-mongering, they may be less inclined to report serious concerns to adults, even when they should. They have reason to fear that adults will overreact. Increasing their disinclination to report is a serious and damaging consequence of the irresponsible fear-mongering.
Source: "Delete Internet Fear-Mongering", Nancy Willard, Center for Safe and Responsible Internet Use, 7 December 2007 [*8]

Further, as reported in the Queensland Police Service Safety Bulletin:

"Detective Superintendent [Ross] Barnett [head of the QPS Sexual Crimes Investigation Unit] encouraged parents to be vigilant as they were the first and best line of protection for their children.
However, parents shouldn't be unduly alarmed about incidents of paedophilia in the community.
There hasn't necessarily been an upsurge in incidents, more an increase in reporting offences and a willingness for public debate on the topic.
[... tips and suggestions...]
These suggestions are not intended to be an exhaustive list but are relevant to all situations. Parents and caregivers should use common sense to ensure the safety and welfare of their family.
It is important for parents and caregivers to keep the fear of violence in perspective as this could result in unnecessary limitations being placed on their lifestyle.
Source: "QPS leads the way in child safety on the internet", Simon Kelly, Queensland Police Service Safety Bulletin, September 2007 [*9]

Sensational false and/or misleading statistics unnecessarily increase fear in the community, including among politicians, often resulting in 'knee-jerk' reactions and legislation which does nothing to protect children from sexual and/or other violence and abuse, but interferes in the lifestyle of law-abiding citizens and wastes millions of dollars of taxpayer funding.

National debate requires facts and informed opinion, not hysteria

As noted by Professors Donna L. Hoffman & Thomas P. Novak, Vanderbilt University, in 1995 (in relation to a Time Magazine report about the subsequently discredited Martin Rimm 'cyberporn' study):

"The critically important national debate over ... rights and restrictions on the Internet and other emerging media requires facts and informed opinion, not hysteria. Misformation, when propagated, begets even worse misinformation."
Source: "The Cyberporn Debate", Professors Donna L. Hoffman & Thomas P. Novak, Vanderbilt University, 1995[10]

Promulgators of unsubstantiated sensational statistics about child pornography often accompany same with calls to "do something"/"give it a go" and the "something" is usually government censorship of the Internet by 'blocking' access to web sites. Others, who understand the technological nature of the world wide Internet (which comprises much more than "web sites") and associated impossibility of effectively 'blocking' Internet content and communications, have a different view about what the "something" should be.

While the mainstream media constantly talk about web sites as if "the Web" = "the Internet", rarely is public and political attention drawn to the problem of child sexual abuse material trafficking via non-Web Internet protocols such as peer to peer (P2P) networks, Usenet newsgroups, IRC (Internet Relay Chat), FTP, Instant Messaging and email.

Meanwhile all indications are that the vast majority of such trafficking does not occur via Web sites, and therefore so-called 'blocking' of web pages will have minimal if any impact on such trafficking, let alone on reducing child sexual abuse. According to the ACMA filtering trial report issued in July 2008, none of the ISP-level filter products were capable of filtering and blocking illegal material distributed/accessed via non-Web protocols. The writer considers the government's web site 'blocking' plan is the equivalent of placing road blocks at street intersections which currently have traffic lights for the purpose of preventing transportation of illegal drugs.

A national debate should be occurring about the merit (if any) of spending AUD$ 44.5 million on 'blocking' web pages rather than increasing law enforcement agencies' funding to enable them to rescue children when the map to their abuse is [or would be with adequate funding] sitting right in front of [them].

Rarely is it mentioned in the media that the Labor government's AUD$ 44.5 million 'blocking' budget has been funded in part by shaving AUD$ 2.8 million off funding previously budgeted by the Coalition Government for increasing the size of the Australian Federal Police's Online Child Sexual Exploitation Team ("OCSET"). AUD$ 44.5 million is 6 times OCSET's annual $ 7.5m budget as at 2007, which then had 35 specialist officers.

The writer believes that the AUD$ 44.5 million the Australian government plans to spend on 'blocking' access, which at best can only prevent accidental/unintentional access to material on web sites, would be far better spent on increased funding/resourcing of the online child exploitation units of Australian law enforcement agencies - most notably Queensland's Task Force Argos which, unlike some other Australian police services, has a demonstrated record of high competence in using a combination of technological know-how; old-fashioned detective work; and international police liaison to find, infiltrate groups and capture people who are sexually abusing children and distributing pictures of their crimes online, and most importantly, finding and rescuing their child victims [*11].

Funding Task Force Argos, not only to increase their own work, but to further enable them to train other Australian police service officers, would achieve far more in reducing child sexual abuse/exploitation and finding children currently being abused than so-called 'blocking' of web sites. (As at September 2007, Argos had a team of 30 detectives [*12]).

2. False and/or Misleading Statistics

"It is estimated that 100,000 commercial websites offer child pornography"

This 100,000 number originated in 2000 and has often been attributed to "Canadian Police". However, in April 2005 the officer-in-charge of the Royal Canadian Mounted Police's National Child Exploitation Coordination Center said it is not their estimate and that they obtained it from a January 2002 magazine article which attributed the 100,000 number to the U.S. Bureau of Customs.

Further, the alleged U.S. Custom's estimate did not include the words "commercial" or "offer". Considerably more recent estimates/statistics (2007-2008) place the number of commercial websites between 150 and 2,204, most of which remain 'live' for less than 50 days.

"More than 20,000 images of child pornography are posted on the internet every week"

This 20,000 'statistic' has nothing to do with numbers of postings on Web sites. It was invented in late 2003 by, or for, the U.K. NSPCC (known in the U.K. for its scare campaigns).

It is a misrepresentation of COPINE Project research about the number of image files, not all of which were child pornography, posted into a sub-set of Usenet newsgroups in August/September 2002. The 'statistic' was invented by inserting the word 'pornography' into a phrase originating in COPINE Project research findings: 140.917 child images files and dividing the number by six (weeks). According to the ACMA filtering trial report issued in July 2008, none of the filters were capable of filtering Usenet newsgroup postings.

"Child pornography is one of the fastest growing online businesses generating approximately $US 3 billion ($ 3.43 billion) each year"

This '$US 3 billion' figure has no credibility and even if it had credibility in 2008, then it could be regarded as 'good news' because it would mean (based on previously promulgated 'statistics') that there had been no increase at all in the five years to 2008, therefore 'child pornography' could not be one of the fastest growing online businesses.

The '$US3 billion' figure has been promulgated far and wide since at least mid 2003, when Jerry Ropelato, a Utah-based anti-pornography crusader and content filtering software promoter, commenced publishing it on his web site InternetFilterReview.com, without citing a source for that or many of the other scary numbers he promulgates.

"Child pornography is a $ 20 billion industry worldwide"

This out-of-date/discredited $ 20 billion 'statistic' was given new life in March 2008 when it appeared in Australian media reports as a result of a joint media release between the Australian Federal Police and Microsoft. The statistic was disowned in April 2006 by the organisations to which it had been, and still is being, attributed (i.e. the FBI and Unicef).

"Approximately 20% of all Internet pornography involves children"

This 20% 'statistic' originated in a 1994 study which found that approximately 20% of of 150 images posted to Usenet newsgroups (i.e. not Web sites) depicted naked persons who were or were claimed to be (by the poster) under 18 years. The images were mostly of the type found in nudist magazines and the researchers said they never came across an image depicting a sexual act between an adult and a child/adolescent, or acts between children. Obviously any 1994 statistic is well overdue for retirement.

"40 per cent of arrested child pornography possessors sexually abused children"

This 40% number was in a report distributed by the NCMEC in 2005. However, insofar as the phrasing of the assertion quoted above appears to imply that 40% of persons arrested for possession of child pornography were found to have sexually abused children, it does not accurately reflect the research findings.

The U.S. case academic research, from which the NCMEC's 40% figure originates, found in a one year period beginning 1 July 2000, an estimated

1,713 arrests for Internet-related crimes involving the possession of child pornography (of which 55% also involved sexual victimization of children, or attempts to do so), and an estimated
65,000 arrests in the U.S. for all types of sexual assaults committed against minors.

Of the Internet-related cases, one out of six [i.e. 16% of] the cases originating with an allegation or investigation of child pornography discovered a dual offender who had also sexually victimized children or attempted to do so.

"50,000 predators prowling for children online at any given time"

The number is what Kenneth Lanning (formerly FBI) calls a Goldilocks number - not too hot, not too cold, i.e. a number that might sound credible although it has no factual basis. The invented number has since been promulgated far and wide and is often incorrectly attributed to "law enforcement officials" or Alberto Gonzales, the former US attorney-general.

"Over half of 11-to-15 year olds surveyed who chat online are contacted by strangers"
and
"40 per cent of children who chat online said they had been contacted by someone they didn't know"

The above and similar alarming claims were made by the Australian Government in a AU$ 22 million NetAlert advertising/information campaign during the 2007 Federal election campaign. Unsurprisingly, given the government's efforts to prevent, or at least delay, public release of the associated Government-commissioned survey/research report, it did not support the alarming statistics used in the government's pre-election advertising campaign.

"British Telecom (BT) blocks over 35,000 attempts per day to access child pornography websites" and
"During 2006, the Norwegian system blocked 1.7 million attempts to access child pornography. The Swedish system blocked 15,000 attempts during its first few weeks"

These numbers actually refer to 'hits', not 'attempts' to visit a web site or web page, and are therefore largely meaningless.

More detailed information about the above 'statistics' is below. Section 3 later herein provides information about estimates/statistics originating more recently, and which appear more likely to have some connection with reality.

"It is estimated that 100,000 commercial websites offer child pornography"

Even if the estimated 100,000 'estimate' had any factual basis when it originated in 2000 (improbable), the number is likely to have markedly decreased given, since 2000, law enforcement agencies around the world have become increasingly active, inter-cooperative and successful in closing down such sites, and numerous governments have enacted new criminal offences, or heavier penalties, to deter establishment of new sites.

Far more recently, in late 2007/early 2008, estimates of 200 - 1,700 commercial web sites have been attributed to U.S. Customs officers, and in April 2009 the U.K. Internet Watch Foundation ("IWF") announced that the number known to the IWF during 2008 was 1,546 (a decrease from the 2007 year when the IWF reported 2,755) of which 74% (1,144) were commercial domains, and the majority (60%) of the total 1,546 domains were live for 50 days or less, and 81% were live for 100 days or less.

Hence, considerably less than 1,546 would have been 'live' at any given time

(due to law enforcement action and/or ISPs becoming aware of illegal material having been uploaded to their servers and deleting/disconnecting same).

More detail about recent estimates/statistics is provided later herein.

The origin and information laundering history of the out-of-date 100,000 'estimate' is outlined below.

5 October 2000: The Christian Science Monitor reported [*13]:

"Kevin Delli-Colli, director of the [U.S. Customs' new CyberSmuggling] center, says there are estimates of 100,000 Web sites that are involved with child pornography."

April 2001: a U.S. Bureau of Customs' newsletter Customs and Border Protection Today stated, in a list of dot points at the end of an article, that [*14]:

"An estimated 100,000 Web sites are involved in some way with child pornography."

2 January 2002: Articles in Red Herring Magazine claimed [*15]:

"The U.S. Customs Service estimates that there are more than 100,000 Web sites offering child pornography."

and on the same day another article by "staff" claimed:

"An estimated 100,000 Web sites sell illegal sexual images of children."

29 November 2004: Australian Federal Police Commissioner Mick Keelty stated in a speech that [*16]:

"Canadian estimates place the number of child pornography websites operating globally at over 100,000."

April 2005: An Associated Press article claimed [*17]:

"Canadian police estimate that more than 100,000 Web sites contain images of child sexual abuse."

April 2005: Canadian police said it was not their estimate (see below).

18 August 2005: The U.S. National Center for Missing and Exploited Children ("NCMEC") issued a media release [*18] claiming, among numerous other things, that:

"A 2002 report by ECPAT International and the Bangkok Post, estimated that 100,000 child pornography web sites existed on the Internet in 2001."

The '2002 report' was not named. However, testimony by the NCMEC's CEO in September 2006 [*19] cites, as the source of the 100,000 'statistic', a slide presentation about a 2002 survey by ECPAT and the Bangkok Post: Children At Risk Online [*20].

The slide presentation claims 100,000 sites existed in 2001 on the Internet about child pornography, but does not cite a source for the 'statistic', nor does it explain what is meant by 'about child pornography'.

8 January 2008: An opinion article by Bernadette McMenamin, CEO of Child Wise/ECPAT in Australia, published in The Australian national newspaper, stated [*21]:

"It is estimated that 100,000 commercial websites offer child pornography."

On the same day a report in The Australian IT stated *[22]:

"According to Bernadette McMenamin ... more than 100,000 commercial websites offer child pornography."

August 2008: A NSW Parliamentary Library Research Service Briefing Paper observed[23] that:

"Referring to the availability of child pornography material on the Internet, AFP Commissioner, Mike Keelty, said in 2004 that Canadian estimates place the number of child pornographic websites operating globally at over 100,000, generating around US$ 3 billion per annum."

April 2009: A report issued by the Australian Communications and Media Authority made the same claim as above about "Canadian authorities" [*24].

In April 2005, The Wall Street Journal's "Numbers Guy", Carl Bialik, investigated the 'Canadian Police estimate' reported in the Associated Press article and subsequently reported in "Retirement Is Long Overdue For Some Aging Statistics" that:

"... Jennifer Strachan, officer-in-charge with the Royal Canadian Mounted Police's National Child Exploitation Coordination Center, told me her office had supplied the statistic. Her source: the Jan. 3, 2002, issue of Red Herring magazine, which reported [*25], The U.S. Customs Service estimates that there are more than 100,000 Web sites offering child pornography. Ms. Strachan also directed me to a Web site called Internet Filter Review that carries the statistic, with no source. My phone call to a number listed on the site's registration wasn't returned, and an e-mail to the site's contact address got bounced back to me."

It wasn't directly from our research, Ms. Strachan says. So why cite the statistic? It's just to show the proliferation of the ability to access this information, whereas prior to the Internet, [child pornography] would be underground and you would have to know somebody. Ms. Strachan says she has also used the stat in public education and in funding proposals.

"I found other articles besides the one published in Red Herring that attributed the 100,000 stat to Customs officials. So I contacted the Department of Homeland Security, which absorbed part of the Customs Service in a 2003 reorganization. But spokeswoman Jamie Zuieback said she hadn't heard the stat before. After some digging, she found a reference to it from an April 2001 Customs press conference, but wasn't able to track down the original source. Ms. Zuieback also said the department has no current stats available on the number of Web sites that carry child pornography."
Source: "Retirement Is Long Overdue For Some Aging Statistics", Carl Bialik, 22 April 2005 [*26]

Further, with regard to the year 2000/2001 'estimate' of 100,000 Web sites involved in some way with child pornography, this may have included sites that did not contain illegal images. For example, Australian Institute of Criminology research involving observing online paedophile activity during a one year period (1996-1997) found that:

"Paedophiles were very concerned to conceal their identity. This was not unexpected given society's attitude to paedophilia. Many of the Internet links provided on WWW pages described anonymity and privacy techniques. ...

WWW pages were used to deliver coming out presentations (although most presenters hid behind masked identities). ...

They also acted as a vehicle for soliciting communications from other paedophiles. ...

[T]here appeared to be a relationship between the potential to be identified and the publication of explicit child pornography. The stronger the perceived anonymity the more explicit the child pornography. WWW pages offered weak anonymity for owners. While it may be possible to disguise identity when obtaining ISP facilities, maintaining pages provides ISPs with an opportunity to trace owners. Therefore, paedophiles took care not to display offensive pictures on WWW pages."
Source: "Paedophile Internet Activity", Patrick Forde and Andrew Patterson, Trends & Issues in Crime and Criminal Justice No. 97, Australian Institute of Criminology, November 1998 [*27]

- - -

"More than 20,000 images of child pornography are posted on the internet every week"

This 20,000 'statistic' has nothing to do with numbers of postings on Web sites. It was invented in late 2003 by, or for, the U.K. NSPCC (known in the U.K. for its scare campaigns [*28]).

(Note: Usenet is not a Web site, nor is it 'a site on the Internet'. Usenet newsgroups are more like, but not the same as, email mailing lists. Usenet has existed since 1980, over a decade before the Web was invented [*29].
According to the ACMA filtering trial report issued in July 2008, none of the filters were capable of filtering Usenet newsgroup postings. A few were able to completely block the NNTP (news) protocol which would block a huge amount of innocuous content/discussion etc.)

The origin and information laundering history of this 'statistic' is outlined below:

July 2003: John Carr, NCH U.K. (formerly National Children's Home) issued a paper claiming [*30]:

"In 2002, in a six week period covering August to mid-September [COPINE found] a total of 140,000 child abuse images were posted to the Newsgroups."

October 2003: The U.K. National Society for the Prevention of Cruelty to Children ("NSPCC") issued an advocacy research report, titled: Images of Abuse: A Review of the Evidence on Child Pornography [*31], which they say [*32]:

"Estimates that 20,000 images of child pornography are posted on the internet every week."

8 October 2003: UK media reported:

"20,000 child porn images a week put on internet, says NSPCC", Maxine Frith, The Independent (London), 8 October 2003 [*33]

"Researchers who monitored the internet over six weeks .. .found that 140,000 child pornography images were posted."
"Child porn 'endlessly recycled'", John Carvel, The Guardian, 8 October 2003 [*34]

The NSPCC warned last night...

"An analysis of 140,000 images of child pornography posted on websites over six weeks."

15 October 2003: John G. Malcolm, Deputy Assistant Attorney General, Criminal Division, US Department of Justice, testified before the Senate Judiciary Committee and stated [*35]:

"A recent study by the [NSPCC] indicates that approximately 20,000 images of child pornography are posted on the Internet every week."

December 2003: The NCH U.K. (formerly National Children's Home) issued a report written by John Carr which claimed [*36]:

"In 2002, in a six-week period covering August to mid-September, ... a total of 140,000 child abuse images were posted to these newsgroups."

Sep 2006 - Mar 2007: At least two U.S. State Attorneys issued media releases stating [*37]:

"According to the U.S. Department of Justice ... In addition, more than 20,000 images of child pornography are posted on the Internet every week."

8 January 2008: An opinion article by Bernadette McMenamin, CEO of Child Wise/ECPAT in Australia, published in The Australian national newspaper, stated [*38]:

"More than 20,000 images of child pornography are posted on the internet every week."

The NSPCC's October 2003 report, Images of Abuse, is not publicly available unless one purchases it (thereby helping fund the NSPCC's scaremongering activities).

However, one of the editors of the NSPCC's report is John Carr, who is also the author of at least two publications which refer to '140,000 child abuse images'; one dated July 2003 [*39] (i.e. several months before the NSPCC's report) and one dated December 2003 [*40]. The two publications contain almost identical paragraphs about '140,000 child abuse images' and although the former does not a cite source for the number, the latter does. It claims:

"Usenet newsgroups are one of the lesser known parts of the Internet.
Since the mid-1990s COPINE has been monitoring newsgroups that are known to contain child abuse images on a regular basis. ...
In 2002, in a six-week period covering August to mid-September, ... a total of 140,000 child abuse images were posted to these newsgroups ...
Abusive Images of Children and the Internet: Research from the COPINE Project, M. Taylor and E. Quayle, in Medical and Legal Aspects of Child Sexual Exploitation. A Comprehensive Review of Child Pornography, Child Prostitution, and Internet Crimes Against Children, edited by Cooper et al. GW Medical Publishing, St Louis, USA.
Source: Child abuse, child pornography and the Internet, John Carr, NCH, December 2003, p.12 [41]

The Taylor & Quayle (COPINE Project) paper cited by Carr is Chapter 13 of a two volume publication which has a purchase price of US$ 275.00 and is available in a few Australian university libraries.

Taylor & Quayle's paper does not say 140,000 child abuse images, as Carr's paper claims it does. It contains one paragraph about statistical findings in relation to newsgroups, which states:

...In the 6-week period from August to mid-September 2002, the COPINE Project downloaded a total of 140.917 child images files from the Usenet newsgroups monitored by the project. ...
Source: "Abusive Images of Children and the Internet: Research from the COPINE Project", M. Taylor & E. Quayle, Chap.13 in Medical and Legal Aspects of Child Sexual Exploitation, p.269 [*42]

The NSPCC's claim that 20,000 images of child pornography are posted on the internet every week has been invented by adding the word 'pornography' into the phrase '140.917 child images files' and dividing the number by six (weeks). Hence the NSPCC's claim has no factual basis.

(The Taylor & Quayle paper does not provide any statistics about the number of 'child abuse images' downloaded during the six week period in 2002, nor during any other period.)

The COPINE Project researchers were not downloading only 'child pornography/abuse image' files, but all image files posted into a monitored sub-set of Usenet newsgroups known to be used by paedophiles, because the project was focussed on material related to adult sexual interest in children, which is not limited to child pornography. Hence, the COPINE Project's database of downloaded images included pictures that do not fall within any legal definition of child pornography:

"Essentially innocent pictures can fall within the category of indicative (level 1 [on the COPINE categorizing scale]). Level 1 may include most common pictures of children, either commercially taken or from family albums.
Source: "Typology of Paedophile Picture Collections", Max. Taylor, Gemma Holland and Ethel Quayle, COPINE Project, 2001 [*43]

- - -

"Child pornography is one of the fastest growing online businesses generating approximately $US3 billion ($3.43 billion) each year"

This '$US 3 billion' figure has no credibility and even if it was factual as at January 2008, (when it appeared in an opinion article by Bernadette McMenamin, CEO of Child Wise/ ECPAT in Australia, with citing a source), then it could be regarded as 'good news' because it would mean (based on previously promulgated 'statistics') that there had been no increase at all in the five years to 2008, therefore 'child pornography' could not be one of the fastest growing online businesses.

The '$US 3 billion' figure has been promulgated far and wide since at least mid 2003, when Utah-based Jerry Ropelato commenced publishing it, without citing a source, on his web site InternetFilterReview.com, which has since become part of his TopTenReviews.com.

According to Texas-based Red Orbit News (5 Nov 2006) Ropelato was formerly chief operating officer of ContentWatch, a Salt Lake City-based developer of Internet filtering and virus protection software. He is also known locally as a speaker and presenter on Internet safety issues, and as a crusader against online pornography .*[44]

The fastest growing online businesses claim originated with the U.S. NCMEC, in August 2005, which based its claim on the then two-year old US$ 3 billion 'statistic' promulgated by Ropelato. (The U.S. NCMEC has a long history of promulgating exaggerated/false statistics [*45].)

The origins and history of '$US 3 billion' and 'fastest growing' claims is outlined below.

June 2003 - March 2007: Since at least 21 June 2003 [*46], Ropelato had been claiming on his web site that Child pornography generates $3 billion annually (i.e. not necessarily via the Internet) without stating a source for that particular figure, or any of the many other 'pornography statistics' he promotes.

Ropelato issued a press release making that and other uncredited statistical claims on 6 February 2004 [*47]. The claim remained on his 'Pornography Statistics' page until at least 6 March 2007 [*48], but had been deleted from the page by 15 March 2007 [*49] (according to the Internet Archive's WayBackMachine).

His press release of 12 March 2007 [*50] which claimed to 'update' his previous 2003 uncredited 'statistics' about the 'worldwide pornography industry' did not mention child pornography. At least two journalists have attempted, without success, to ascertain sources of 'statistics' from Ropelato (see below).

29 November 2004: Australian Federal Police ("AFP") Commissioner Mick Keelty stated in a speech: [*51]

"Canadian estimates place the number of child pornography websites operating globally at over 100,000, generating around US$3 billion per annum."

April 2005: The Wall Street Journal's "Numbers Guy", Carl Bialik, investigated the origin of an estimate attributed to 'Canadian Police [*52] and subsequently reported that he was directed by the officer-in-charge of the Royal Canadian Mounted Police's National Child Exploitation Coordination Center to a 2002 magazine article and also to a Web site called Internet Filter Review... . My phone call to a number listed on the sites registration wasn't returned, and an e-mail to the sites contact address got bounced back to me.

18 August 2005: The U.S. National Center for Missing & Exploited Children ("NCMEC") issued a headline grabbing press release titled "Child Porn Among Fastest Growing Internet Businesses" [*53] claiming:

"Within only a few years, child pornography has become a multi-billion dollar commercial enterprise, and is among the fastest growing businesses on the Internet."
Source: Top Ten Reviews 'Internet Filter Review' an online resource that reviews Internet Safety. (Reported that CP generates $3 billion annually)

The NCMEC's media release provided no source or grounds for their 'fastest growing' business claim other than the footnote mention of Ropelato's then two-year old uncredited US$ 3 billion 'statistic' (which did not mention the Internet). Hence, there was no basis for the claim (in 2005, nor its repetition in 2008) that the so-called commercial enterprise is growing at all, let alone is among the fastest growing.

September 2007: A research report published by the Australian Institute of Criminology ("AIC"), a Commonwealth statutory authority, stated [*54]:

"Affordable technology has greatly facilitated the production and distribution of child pornography - a multi-billion dollar industry globally. [Introduction, page xx]
...
TopTenREVIEWS™ has estimated that child pornography generates approximately US$3 billion annually worldwide (Ropelato 2007). [page 62]
Ropelato J 2007. Internet pornography statistics. Internet Filter Review.
Ropelato is the only source cited in the above AIC report for its multi-billion dollar industry claim."

8 January 2008: An opinion article by Bernadette McMenamin, CEO of Child Wise/ECPAT in Australia, published in The Australian national newspaper, stated *[55]:

"Child pornography is one of the fastest growing online businesses generating approximately $US 3 billion ($3.43 billion) each year."

In addition to The Wall Street Journal's "Numbers Guy", Carl Bialik (as referenced earlier herein), at least one other journalist has tried, also without success, to find out the original source of particular 'statistics' Ropelato promotes, after tracing other peoples'/organisations' claims to Ropelato.

In November 2005, Seth Lubove reported:

"Sen. [Blanche] Lincoln lifted the factoid from a report issued in July by Third Way, a new Washington think tank that helps Democrats grab on to red-state issues.
...
Where did Third Way get that notion?
From a May 12 story in the New York Times-owned Boston Globe headlined The Secret Life of Boys, which cites an outfit called Family Safe Media. The small firm in Provo, Utah, is in the business of scaring parents into buying software to protect their kids from Internet smut.

Jared Martin, who owns Family Safe Media, says he got his porn statistics from Internet Filter Review, a Web site that recommends content-blocking software. It is run by tech entrepreneur Jerry Ropelato of Huntsville, Utah, who pens antiporn screeds, such as Tricks Pornographers Play, and publishes curious and uncredited stats (for example, 17% of all women struggle with pornography addiction).

Most of the statistics there have come from literally hundreds of sources, all reputable, Ropelato insists. He says he got the age-11 item from The Drug of the New Millennium, a book about the dangers of porn self-published in 2000 by Mark Kastleman, a self-professed former porn addict in Orem, Utah, who counsels other porn fiends. I don't remember where I got that from, Kastleman says breezily. That is a very common statistic. And there the trail goes cold."
Source: "Sex, Lies And Statistics", Seth Lubove, Forbes.com, 23 November 2005[56]

"Child pornography is a $20 billion industry worldwide"

The history of this number is outlined below.

23 December 2004: A Council of Europe report titled "Organised crime situation report 2004, Provisional" [*57] stated:

"Experts assume that the number of Web sites containing child pornography has grown enormously in recent years. According to estimations by UNICEF, this market has a business volume of about $ 20 billion annually."

September 2005: ECPAT International (based in Thailand) issued a report, Violence against Children in Cyberspace [*58], which claimed:

"The production and distribution of abuse images of children is big business, estimated to be worth billions of dollars a year. Estimates of annual business volume range widely from $US 3 billion to $US 20 billion (the latter, according to the US Federal Bureau of Investigation).
(The ECPAT report provides no source for the $US 3 billion figure - presumably Ropelato, as detailed earlier herein.)

5 April 2006: Texas Republican Joe Barton, (as Chairman, U.S. House of Representatives' Committee on Energy and Commerce), issued a press release [*59] which claimed:

"Child pornography is apparently a multibillion ... my staff analysis says $ 20 billion-a-year business. Twenty billion dollars."

5 April 2006: A New York Times article [*60] attributed the entire claimed amount to the Internet:

"The sexual exploitation of children on the Internet is a $ 20 billion industry.

April 2006: The FBI and Unicef disowned the US$ 20 billion number."

7 March 2008: A joint media release between the Australian Federal Police and Microsoft"[*61] and articles in The Australian IT [*62] and ComputerWorld (AU) [*63] claimed:

"The FBI estimates that the production and distribution of child abuse images is valued at US$ 20 billion ($ 21.6 billion) annually."

5 June 2008: During an interview by Radio 2GB's Phillip Clarke about Operation Centurion, James McCormack (head of the AFP's High Tech Crime Operations) *[64] claimed:

"The FBI did a study a couple of years ago and they estimated the commercial child pornography industry was probably valued at anywhere between about three to twenty billion dollars of commercial activity per year, so it's a pretty signficant industry."

As a result of the April 2006 publicity, two U.S. journalists investigated the source of the $ 20 billion figure and reported their findings in: "Measuring the Child-Porn Trade", Carl Bialik, The Numbers Guy, The Wall Street Journal Online, 18 April 2006 [*65].

"How big is the online kiddie porn industry?", Daniel Radosh, radosh.net, freelance journalist, 5 April 2006 [*66]

In short, the trail to the origin of the claimed $20 billion 'statistic' went from Joe Barton's press release/staff analysis, to the U.S. National Center for Missing and Exploited Children (NCMEC), to 'McKinsey Worldwide', to the ECPAT International 2005 publication (mentioned above) which claimed ...$US20 billion (the latter, according to the US Federal Bureau of Investigation) [*67].

WSJ Numbers Guy, Carl Bialik, reported his findings from following the above trail in April 2006:

"Mr. [Ernie] Allen [CEO, NCMEC] faxed me an NCMEC paper that cites the McKinsey study in placing the child-porn industry at $ 6 billion in 1999, and $ 20 billion in 2004.
But a McKinsey spokesman painted a different picture for me:
The number was not calculated or generated by McKinsey, he wrote in an email. Instead, for a pro bono analysis for Standard Chartered, he said, McKinsey used a number that appeared in a report last year by End Child Prostitution, Child Pornography and Trafficking of Children for Sexual Purposes, an international advocacy group. [i.e. ECPAT, which attributed the number to the FBI]

FBI spokesman Paul Bresson told me in an email, The FBI has not stated the $ 20 billion figure ... . I have asked many people who would know for sure if we have attached the $ 20 billion number to this problem. I have scoured our Web site, too. Nothing!

I went back to the NCMEC Monday and shared what I found. In an email response, spokeswoman Joann Donnellan said, If it is determined that this ends up not being a reliable statistic, NCMEC will stop citing McKinsey as the source and will also stop citing a specific number. Rather, NCMEC will revert to what it has said previously ... that commercial child pornography is a multi-billion dollar industry.

This isn't the first number from the NCMEC that struck me as questionable... As I wrote last year...
Source: "Measuring the Child-Porn Trade", Carl Bialik, The Wall Street Journal Online, 18 April 2006 [*68]

Fifteen months later, Ernie Allen of the NCMEC was still citing McKinsey as source. On 24 July 2007 he told the U.S. Senate Committee on Commerce, Science and Transportation that:

"A recent report by McKinsey Worldwide estimated that today commercial child pornography is a multi-billion-dollar industry worldwide, fueled by the Internet."[*69].

The $20 billion figure was also found by Carl Bialik in a Council of Europe 2004 report which attributed the number to Unicef. Bialik subsequently reported on 27 Apr 2006:

"But Allison Hickling, a spokeswoman for the United Nations child agency, told me in an email, 'The number is not attributable to Unicef -- we do not collect data on this issue.'

I told Alexander Seger, who worked on the Council of Europe reports, that the Federal Bureau of Investigation and Unicef, both cited in Council reports, said they weren't the source for the $ 20 billion figure. He said the Council won't use the number in the future, and added in an email, I think we have what I would call a case of information laundering: You state a figure on something, somebody else quotes it, and then you and others [quote] it back, and thus it becomes clean and true. ... Perhaps this discussion will help instill more rigor in the future.
Source: Carl Bialik, The Numbers Guy, The Wall Street Journal Online, 27 Apr 2006 [*70]

In summary, the US$20 billion figure has been invented by some unknown person/organisation and since then been commonly attributed to the FBI or Unicef, both of which said in April 2006 that the 'statistic' did not originate with them.

The writer considers the fixation among advocacy groups and the media with attaching a dollar number to the problem is curious given commercial/monetary 'estimates' (or even factual statistics if it were possible to obtain same) are of minimal relevance to understanding or determining the extent of the problem.

This is because a significant amount, possibly most, of the trade in child sexual abuse images takes place at no cost via Usenet newsgroups, IRC (Internet Relay Chat), Instant Messaging, P2P technologies, email, etc. The images themselves, not money, are the trading currency. Extensive information about the nature of the non-commercial trade is available in the book: Beyond tolerance : child pornography on the Internet by Philip Jenkins. New York University Press, 2001. ISBN: 0814742629.

Non-commercial criminal activity was referred to during an AFP media briefing about 'Operation Centurion' (June 2008) which concerned a legitimate web site that had been broken into for the purpose of uploading illegal images to it:

Journalist: "The people who put these images up on a site, are they getting paid, [...inaudible...], where's the economic benefit?

AFP Andrew Colvin: We're not talking about a crime that's driven by a financial motive, there's other motivations here. So, while there may be some sites that attract a financial return, that's not the motivation here. So the answer to your question is no really, that's not what's motivating people, people aren't necessarily making a lot of money."
Source: Australian Federal Police media briefing [Mp3 audio], LiveNews/Macquarie National News, 4 June 2008 [*71]

- - -

"It is estimated that approximately 20% of all Internet pornography involves children"

Obviously any 1994 statistic is well overdue for retirement. Nevertheless fourteen years later, the 20% statistic is promoted in a Child Wise/ECPAT in Australia "Fact Sheet" titled "Child Pornography and the Internet" which attributes the number to a U.S. NCMEC report (of "2003") [*72].

The number is also promoted on the web site of NSW Christian Democratic Party politician, Rev. Gordon Moyes (without source citation) [*73] and on numerous child protection campaigners' web sites some of which cite the NCMEC.

The U.S. NCMEC's "Child Pornography Fact Sheet" states:

"A current study has estimated that as much as 20 percent of all pornographic activity on the Internet may involve children; however, accurate estimates are difficult to produce since a reliable methodology to measure the actual extent of these images online has yet to be devised."
Source: "Child Pornography Fact Sheet", NCMEC, (as at November 2008)[74]

The "current study" however is a study conducted 14 years ago in 1994 which analysed 150 postings into Usenet newsgroups.

The NCMEC's Fact Sheet reference numbers [12] and [13] cite page 3 of a March 2001 report commissioned by the NCMEC titled "Child Pornography: The Criminal-Justice-System Response" which states:

"One study has estimated that as much as 20 percent of all pornographic activity on the Internet may involve children;11 however, even the authors of this study admit to a high margin of error.
...
Excerpt from Canadian Parliament: House Debates, Government Order: Allotted Day on Child Pornography, 36th Parliament, 1st Session, Edited Hansand 1, No. 172, Feb. 2, 1999 (citing Mehta study, infra note 12).
Michael D. Mehta & Dwaine E. Plaza, Content Analysis of Pornographic Images Available on the Internet,
THE INFORMATION SOCIETY 153-62 (1997) (original study presented October 1994)"
Source: "Child Pornography: The Criminal-Justice-System Response", NCMEC commissioned report, March 2001[75]

Mehta & Plaza's report on their 1994 analysis of postings into Usenet newsgroups states:

"We examined the nature and content of 150 randomly selected pornographic images available through newsgroups ... Using content analysis, we identified themes that appear ... [and found] ... images of children/adolescents (phi=.21, p<.01) ...

If an image was accompanied by text suggesting that subjects were under the age of 18, we coded this image as a hit. In other words, models in such coded images may be technically adults by age, but props or descriptors give the illusion of being an adolescent.
As a qualifier, the vast majority of the small number of such images depicting children and adolescents probably come from nudist magazines or are taken by those who have visited such communities. We never came across an image depicting a sexual act between an adult and a child/adolescent, or acts between children.
Source: "Content analysis of pornographic images available on the Internet", Michael D. Mehta & Dwaine E. Plaza [*76]

In summary, claims in 2008 that approximately 20% of all Internet pornography involves children originate from a 1994 study which found that approximately 20% of of 150 images posted to Usenet newsgroups depicted naked persons who were or were claimed to be (by the poster) under 18 years.

"40 per cent of arrested child pornography possessors sexually abused children"

According to an opinion article by Bernadette McMenamin, CEO of Child Wise (ECPAT in Australia), published in the The Australian on 8 January 2008: In 2005 the United States National Center for Missing and Exploited Children revealed that 40 per cent of arrested child pornography possessors sexually abused children. [*77]

The 40% number was in a report distributed by the NCMEC in 2005 and the percentage concerned research findings in relation to a total of 429 cases during the 12 months beginning 1 July 2000. However, insofar as the phrasing of the assertion quoted above appears to imply that 40% of persons arrested for possession of child pornography were found to have sexually abused children, it does not accurately reflect the research findings.

The research found that one out of six, i.e. 16% of cases originating with an allegation or investigation of child pornography discovered a dual offender who had also sexually victimized children or attempted to do so.

Findings of the N-JOV Study

The source of the 40% figure is the second report on the National Juvenile Online Victimization Study ("N-JOV Study") [*78] conducted by researchers (Janis Wolak, David Finkelhor, and Kimberly J. Mitchell) at the Crimes Against Children Research Center, University of New Hampshire (in north-western U.S.A.). The research report was funded by the U.S. Congress Through a Grant to the National Center for Missing & Exploited Children.

(Note:
Although NCMEC's media release of 18 August 2005 cites the above research report as the source of numerous NCMEC claims about 'growing', 'increasing', etc, the research report did not find, or claim, that anything is increasing, growing, etc. The research concerned a one year period beginning 1 July 2000 and did not compare findings from that period with any other period.)

According to the research report:

"The goals of the National Juvenile Online Victimization (N-JOV) Study were to survey law-enforcement agencies within the United States (U.S.) to count arrests for Internet-related sex crimes committed against minors and describe the characteristics of the offenders, the crimes they committed, and their victims."

The above report was the second report on the N-JOV Study and it was focussed on a 'representative national sample' of persons arrested for Internet-related sex crimes who possessed child pornography, i.e. a sub-set of the cases identified in the N-JOV survey.

The researchers found that [U.S.] Law-enforcement agencies nationally made an estimated 1,713 arrests for Internet-related crimes involving the possession of child pornography during the 12 months beginning July 1, 2000. The estimate of 1,713 was projected from 429 actual cases identified.

The researchers also stated to give some perspective on this estimate of 1,713 arrests for Internet-related CP possession, we estimate there were approximately 65,000 arrests in 2000 for all types of sexual assaults committed against minors.

In the sub-set comprising persons who possessed child pornography (429 actual cases), 47% of the cases arose in the criminal-justice system as cases of child sexual victimization or attempted child sexual victimization (solicitations to undercover investigators) and 53% of the cases arose as cases involving child pornography possession.

The 40% number is a further sub-set which comprises cases/persons whom the researchers termed 'dual offenders' because They sexually victimized children and possessed child pornography, with both crimes discovered in the course of the same investigation:

"We found 40% of the cases involving CP possession in the N-JOV Study involved dual offenses of CP possession and child sexual victimization detected in the course of the same investigation. All of these offenders had identified child victims. An additional 15% both possessed CP and attempted to sexually victimize children by soliciting undercover investigators posing online as minors. When these cases of attempted child sexual victimization are counted, 55% of the CP possessors were dual offenders (unweighted n = 241, weighted n = 936)."

(Note that the actual number of dual offender cases identified was 241).

84% of the dual offenders were discovered in cases starting as investigations of child sexual victimization which subsequently turned up child pornography possession (55% child sexual victimization plus 29% solicitations to undercover investigators). 16% were discovered in cases starting as investigations of child pornography which subsequently detected a sexually victimized child or an attempt to do so (solicitation to an undercover investigator).

The researchers stated:

"When we looked at all of the cases originating as allegations or investigations of CP possession and examined how many resulted in the arrests of dual offenders, we found

In 14% of cases investigators found dual offenders who both possessed child pornography and sexually victimized children.
In 2% of cases investigators found offenders who possessed child pornography and attempted to sexually victimize children by soliciting undercover investigators posing online as minors.
84% of cases involved CP possession but investigators did not detect concurrent child sexual victimization or attempts at child victimization.

This means one out of six cases [i.e. 16%] originating with an allegation or investigation of child pornography discovered a dual offender who had also sexually victimized children or attempted to do so.

The research report also states:

"Limitations
The N-JOV Study is the first research gathering information about a national sample of arrested CP possessors. Data from a national sample is a strength of the N-JOV Study, but like every scientific survey, the study also has limitations. Readers should keep some of these important things in mind when considering the findings and conclusions of this study.
First, ...
Second, ...

Third, there is an additional caution to our findings about dual offenders. Knowing a considerable number of dual offenders were discovered during investigations of Internet-related, child-sexual-victimization and CP possession cases does not explain how possessing child pornography is related to child sexual victimization or whether it causes or encourages such victimization.

We did not have the data to determine this. In particular we had no information about the sequencing of the crimes committed by dual offenders or about undetected crimes they may have committed and little information about their criminal histories and how they used the child pornography they possessed."

In summary, the U.S. case research from which the NCMEC's 40% figure originates, found in a one year period beginning 1 July 2000, an estimated 1,713 arrests for Internet-related crimes involving the possession of child pornography (of which 55% also involved sexual victimization of children, or attempts to do so), and an estimated 65,000 arrests in the U.S. for all types of sexual assaults committed against minors. Of the Internet-related cases, one out of six [i.e. 16% of] the cases originating with an allegation or investigation of child pornography discovered a dual offender who had also sexually victimized children or attempted to do so.

- - -

"50,000 predators prowling for children online at any given time"

This 50,000 'statistic' originated in late 2005 in a controversial TV series for NBC's Dateline called "To Catch a Predator". The number was made up by a Dateline reporter, Chris Hansen - it has no basis in fact/knowledge. The number is what Kenneth Lanning (formerly FBI) calls a Goldilocks number - not too hot, not too cold, i.e. a number that might sound credible although it has no factual basis.

The invented number has since been promulgated far and wide and is often incorrectly attributed to "law enforcement officials" or Alberto Gonzales, the former US attorney-general.

For example, a Child Wise/ECPAT in Australia "Fact Sheet" titled "Child Pornography and the Internet" [*79] claims:

"Law enforcement officials estimate that as many as 50,000 sexual predators are online at any given moment."

For information about the actual origin of the 50,000 number, see:

"Prime Number", On The Media, WNYC-New York Public Radio, 26 May 2006 [*80].
Interview with Chris Hansen (Dateline), Kenneth Lanning (former FBI Agent), Carl Bialik (WSJ 'Numbers Guy') and others about the above and other 50,000 'statistics':

"BROOKE GLADSTONE:
The children. We are all of us frightened for the children. And we have plenty of numbers to justify that fear, like a 20-billion-dollar child porn industry or 50,000 predators prowling for children online, numbers that resound endlessly through the media ether, origins unknown.
Take that last number, 50,000 sexual predators logged in at any given time. That appeared late last year in a series for NBC's Dateline, called To Catch a Predator. Last week, Attorney General Alberto Gonzales cited it. Legal Times noted that spokespersons for the FBI, the National Center for Missing and Exploited Children and the Crimes Against Children Research Center say it's not based on any research they're aware of. The A.G.'s Office said it came from "Dateline."
...
BROOKE GLADSTONE: So why is 50,000 such an unaccountably sticky figure?
KEN LANNING: Maybe the appeal of the number was that it wasn't a real small number - it wasn't like 100, 200 - and it wasn't a ridiculously large number, like 10 million. It was like a Goldilocks number - not too hot, not too cold. ...

All Predators, All the Time? Maybe Not, Steven Levy, Newsweek, 3 July 2006 [*81]:

"How big is the threat of predators online? In the first two Dateline shows, Hansen reported someone's estimate that at any given moment, 50,000 potential child molesters were prowling the Internet. No hard statistics backed this up, and when critics questioned the figure, Dateline stopped using it."

"Don't believe the hype", an edited extract from Risk: The Science and Politics of Fear by Dan Gardner, Sydney Morning Herald, 26 April 2008[82]:

"Despite being the safest and healthiest humans to have lived, we allow experts to scare us witless, says Dan Gardner.
Recent figures suggest some 50,000 pedophiles are prowling the internet at any one time, says the website of Innocence in Danger, a non-government organisation based in Switzerland. No source is cited for the claim, which appears under the headline Some terryfying [sic] statistics.

It is indeed a terrifying statistic. It is also well-travelled. It has been cited in Britain, Canada, the US, and points beyond. Like a new strain of the flu, it has spread from newspaper articles to TV reports to public speakers, websites, blogs, and countless conversations of frightened parents. It even infected Alberto Gonzales, the former US attorney-general.

Unfortunately, the mere fact that a number has proliferated, even at the highest levels of officialdom, does not demonstrate the number is true. ..."

"Over half of 11-to-15 year olds surveyed who chat online are contacted by strangers" and "40 per cent of children who chat online said they had been contacted by someone they didn't know"

The above and similar alarming claims were made by the Australian Government in a AU$22 million NetAlert advertising/information campaign during the 2007 Federal election campaign. The 'statistics' were promulgated on billboards at tram stops, in newspaper and television advertisements [*83]; in a NetAlert booklet Protecting Australian Families Online [*84a] that was postal mailed to every Australian household and in an Australian Government Netalert 4 page Attitudes and behaviour of young people online - Research Summary [*84b] (issued by the Minister on 7 Sep 2007), which cited "A snapshot of online behaviour and attitudes of children, Wallis Consulting Group, July 2007" [*84c].

Peter Mares (host of The National Interest program on ABC Radio National) sought to obtain the government commissioned (Wallis) survey/research report, which allegedly contained these 'statistics', by contacting the Minister's (Senator Helen Coonan's) office, interviewing the Minister on radio and also lodging an FOI request.

Before the election, he reported that numerous bureaucratic blocks on release of the information had been put in his way. Among other things, it was claimed by the Minister that the research report could not be made public because it contained personal information. However, it was eventually released under Freedom of Information laws (received the day before the election) and it contained not a skerrick of 'personal information' that might have invaded anyone's privacy [*85].

Unsurprisingly (given the government's efforts to prevent, or at least delay, release), the Wallis survey/research report did not support the alarming statistics used in the government's pre-election advertising campaign.

[Added 4 Jan 2010:
Also, the Labor Party cited the above mentioned Government/Netalert "Research Summary" in "Labor's Plan for Cyber-safety" 2007 election policy document as the source of a number of 'statistics' in that document. For example: "Recent Government research shows that: ... 38 per cent of Australian children under the age of 13 have purposefully visited websites they think their parents would disapprove of them visiting".
However, the Government/Netalert "Research Summary" did not use the term "purposefully" nor any similar term. The Wallis survey report states that "6%" of children surveyed who were under the age of 13 (i.e. were aged 8-12 years) said they had "deliberately visited" websites that "their parents would not want them to see".]

For more information see:

"Stalking the cyber-stats on web stranger-danger", Transcript, Peter Mares interviews Sen. Helen Coonan, The National Interest, ABC Radio National, 16 September 2007 [*86]

"Can we get back to you on that (after the election)?", Peter Mares, The Age, 10 November 2007 [*87]

"Coalition internet campaign 'inaccurate'", Peter Mares, The Age, 15 December 2007 [*88]:

"The claim in the campaign regarding stranger contact does not appear in the government-commissioned research. The question was not posed in this form. Participants were asked: When chatting online, have you ever been contacted by someone you haven't met in real life? More than half answered yes.
But when asked who they chatted to or messaged, they said communication was mostly with friends (96%), friends of friends (31%) or people met online who their parents had said it is all right to talk to (20%). Only 14% chatted or messaged with just a mixture of people including strangers. ...

"Cyber stat alert", Transcript, Peter Mares interviews Dr Jane Burns, a specialist in adolescent mental health and Director of Research and Policy at the Inspire Foundation (which runs the Reach Out! website for young people), The National Interest, ABC Radio National, 16 December 2007 [*89]

"I'm talking to Dr Jane Burns from the Inspire Foundation, a mental health professional with a particular interest in the way in which young people use the internet and the way the internet can be used to help young people. And we're discussing the survey that lay behind the previous federal government's NetAlert campaign about the dangers of lurking online for children and teenagers. Jane Burns, one other thing the survey shows is that most parents - three-quarters, in fact - are already talking to their children about this stuff."

"How NetAlert accentuated the negative", Peter Mares, Comment, Creative Economy, undated (late 2007) [*90]:

"Research obtained under freedom of information laws raises questions about the accuracy of the Howard government's pre-election advertising campaign on internet safety."

A Snapshot of the Online Behaviour and Attitudes of Children, Wallis Consulting Group, report commissioned by the Commonwealth Department of Communications, Information Technology and the Arts, July 2007 [*91].

- - -

"British Telecom (BT) blocks over 35,000 attempts per day to access child pornography websites"

This misleading claim, made by numerous UK media outlets, originates from British Telecom's (BT) self-promotional 'advertising', on Safer Internet Day 2006 (7 February 2006), that its 'CleanFeed' system had blocked 35,000 hits per day over the past four months [*92].

Three years later, in April 2009, it was reported that: Between 35,000 and 40,000 attempts to access child pornography sites via BT Retail's broadband network are blocked every day [*93].

The 35,000 / 40,000 'attempts' claim is misleading and deceptive for a number of reasons.

Firstly, BT's numbers are actually about hits. 35,000 hits is not the same as 35,000 attempts to visit a web site (nor does the number signify a 'rise in internet child porn searches' as BBC News claimed in 2006). BT was asked in 2004 to explain what it counts as a 'hit' when it first issued such statistical claims, but BT has refused to do so.
Accessing one web page almost always generates numerous hits depending on the number of images and other items of content on the page. For example, accessing the home page of the Sydney Morning Herald newspaper or News.com.au generates over 90 'hits' (at April 2009).
An example of the significant difference between hits and visits to a website was included in the February 2006 Newsletter of the Virtual Global (Police) Taskforce ("VGT") which reported statistics about the VGT website in the first year since launch: Total number of hits is over 17 million. Total number of visits is 250,000 [*94].
The difference would be even greater in relation to sites where page content is primarily images, not text, for example, pages containing many thumbnail images.

Secondly, as a BT spokesperson admitted in April 2009:

"A [BT] spokesman said botnets and use of UK proxies from overseas made it hard to judge how many of the attempts to access child pornography sites are made by a real person, whether deliberately or by accident. ... The number has stayed roughly the same [in recent years] even though use of the internet has increased.
Source: "BT blocks up to 40,000 child porn pages per day" Chris Williams, The Register (UK), 7 April 2009 [*95]

The reference to 'botnets' presumably, and should, include the probability that many hits are caused by bots/search engines operated by e.g. spammers which access web site content looking for email addresses to harvest and subsequently spam etc.

Thirdly, BT's 'CleanFeed' system does not 'block'/prevent attempts to access to websites, it blocks accidental/unintentional access, as stated by BT's Mike Galvin in 2004:

"[Cleanfeed] is designed to block casual access to child abuse material on the Net. It won't stop a hardened paedophile and we're not saying that. One of the things we want to cut down on is unintentional access to this kind of material.
[emphasis added]
Source: "BT's modest plan to clean up the Net", John Leyden, The Register (UK) 7 June 2004 [*96]

Similarly, the IWF's FAQ about its 'Child Sexual Abuse Content URL List' [*97] states:

"Blocking is designed to protect people from inadvertent access to potentially illegal images of child sexual abuse. No known technology is capable of effectively denying determined criminals who are actively seeking such material; only removal of the content at source can achieve that goal."

In relation to the increase in 'hits' since BT's announcement of 11,000 hits per day in 2004, Ian Livingstone, Chief Executive of BT Retail, told BBC News:

"That increase is at least in part because of greater use of the Internet, but what it also represents I think is that there is more things like spam and viruses that are causing accidental attempts to access these sites."
Source: "Rise in internet child porn searches", BBC News (video), 7 Feb 2006 [*98]

- - -

(Note: In the above BBC News 'report', the reporter claimed there was an estimated 6,000 illegal websites. That was not true (unless it was the reporter's own estimate). The 6,000 figure, actually 6,128, was the number of URLs (pages) found by the IWF during 2005 to contain 'potentially illegal' content under UK law. The number of websites/domains was 2,966 [*99].)

For further information and commentary about BT's 'statistics', see:

Twisting the facts to fit the story - child porn nonsense, Kieren McCarthy, 7 February 2006 [*100]:

"Some 35,000 attempts to access child pornography websites are blocked in Britain every day, figures have revealed, screeches the Daily Mail this morning. The Times says the same, that the number of attempts to access websites hosting child pornography have trebled in the past 18 months.

And the BBC. And the Independent. And the Today programme on Radio 4 this morning. Oh, and the Scotsman. The Sun offers: Shocking new figures show UK perverts are trying to access child porn websites 35,000 times a day. And so on and so forth.

The only problem with this scaremongering story is that it is total bollocks. In fact, it is worse than that. It is a carefully engineered twisting of figures and facts to provide a dramatic story. Maybe that's just par for the course, but when you are talking about child pornography and building up an image of the Internet as a dangerous, lawless place, you would think media outlets like the BBC, Independent and Times would be a little more careful.

The facts
...
BT released figures yesterday that showed Cleanfeed had picked up four million hits over the past four months. This has been reported as 35,000 attempts to access child porn site per day.
....
BT uses the word hits. Since BT is the biggest ISP in the country, we can expect that it knows the difference between hits, page impressions and unique users.
The fact that the media choses to misrepresent this word as attempts is the first example of a conscious effort to twist the facts. ...

In relation to the BT 'statistics' announced in 2004, see:

"Porn filters ineffective against Tribbles", Lucy Sherriff, The Register (UK), 23 July 2004 [*101]:

"Letters.
This Friday's post bag was bulging with comments about the number of child porn sites blocked by BT's new CleanFeed filter, ISPA's subsequent call for clarification, and BT's response.
For anyone just back from an extended holiday in Cuba, the only way you could have missed this one, this is the story that BT kicked off by reporting that in the last three weeks, it has blocked nearly a quarter of a million attempts to access child porn. As always with statistical stories, the devil is in the details"

ISPA seeks analysis of BT's 'Cleanfeed' stats: Web filtering figures 'could be misleading', Tim Richardson, The Register (UK), 21 July 2004 [*102]:

"ISPA - the trade group for the UK's ISPs - wants to carry out an analysis of BT's "Cleanfeed" Web filtering system because of concerns that the data might be giving a misleading picture of the scale of Internet child abuse.
...
ISPA's Statement In Full:
ISPA welcomes new developments in the fight against child abuse images appearing on the Internet. However ISPA feels caution is needed with the information and statistics so far available on Cleanfeed.
It is very difficult to comment on the statistics reported by BT regarding Cleanfeed as BT has not passed the data to ISPA.
At present there seems to be a significant disparity in the statistics that are being reported.
20,000 URL requests per day reported by [BT Retail chief exec] Pierre Danon on Tuesday morning on BBC Radio 4 does not equate to 230,000 URL requests per day between June 21st and July 13th, which would mean around 10,000 URL requests per day.
There is also a need to understand exactly what Cleanfeed is detecting.
At present we do not know if Cleanfeed is measuring the number of 'hits' (attempts to download individual files from illegal websites) or 'visits' (number of attempts to visit the website). ...

"During 2006, the Norwegian system blocked 1.7 million attempts to access child pornography. The Swedish system blocked 15,000 attempts during its first few weeks"

The above statement appeared in a January 2008 opinion article by Bernadette McMenamin, CEO of Child Wise/ECPAT in Australia, published in The Australian national newspaper *[103].

The "attempts" claim is misleading for the same reasons discussed above in relation to BT's system.

The Norwegian and Swedish blocking systems referred to above are operated by Telenor ISP which stated in October 2006: The filter receives about 15,000 hits per day [*104]. As discussed above, hits are not the same as attempts to visit a site.

In addition, the Telenor system uses DNS blacklisting which does not enable 'blocking' of individual pages, only entire sites/domains. Hence, if a site contains some pages with illegal content and others with legal content, hits on pages with legal content are also 'blocked' and therefore included in the number of so-called 'attempts' to access illegal content.

In February 2008, PCWorld reported:

"Norwegian operator Telenor is one of the pioneers of such filtering, having put measures in place on its Internet access and mobile networks in 2004, said Morten Karlsen Sørby, Executive Vice President of Telenor Nordic.
The company blocked 274 sites in 2004, and 5,300 at the start of this year. It blocks around 15,000 attempts to access sites on the block list each day, Sørby said.
However, there will always be ways for organized criminals to beat such filtering, he said. If you are really interested in getting through the filter, it's possible, he said.
Source: "Carriers Team to Fight Child Porn on Cell Phones", Peter Sayer, PCWorld, 11 Feb 2008 [*105]

Telenor's DNS blacklisting method of 'blocking' is trivially easy to bypass by anyone, not only 'organized criminals', a fact which has become increasingly well-known since October 2006, and again since January 2008, when a Danish ISP added a music file sharing site and later a file-sharing search engine site to its DNS based blocking system after being ordered by a Danish court to block its customers access.

3. Recent (2007-2008) 'Statistics'

After considerable research in search of more recent and credible 'statistics' than those referred to above, the writer has concluded that many statistics, and often other 'facts', published on the topic of child sexual abuse material on the Internet are of dubious reality and/or veracity at best. Nevertheless, statistical claims originating vastly more recently than those above, and which appear more likely to have some connection with reality, are provided below.

Web sites: US Customs' officers' estimates 2007-2008

As noted earlier herein, the estimated 100,000 Web sites involved in some way with child pornography claim has existed since at least 2000, when it was publicised by the U.S. Customs Service. More recent statistics attributed to U.S. Customs' officers are as follows.

On 17 January 2008, an article in Buffalo News (western New York State) claimed [*106]:

"Organized crime is finding more money in child exploitation than in drugs, said Claude Davenport, an agent with the U.S. Immigration and Customs Enforcement.

It's estimated there are some 1,700 commercial child pornography sites - containing thousands of Web pages - operating worldwide. It's likely though, that less than 300 businesses operate all the sites, authorities said.

Some of the guys who run these sites have multiple locations because the government keeps shutting them down, said Don Daufenbach, a federal agent who works with Davenport. As soon as one site goes down, they move the site to another computer in another country. They're like cockroaches scurrying around the floor when the lights come on."

It is not apparent from the above which 'authorities', if any, said '1,700' sites. While in the context one could assume either or both of Claude Davenport and Don Daufenbach of the U.S. Immigration and Customs Enforcement Service, one could also assume that if they had made such an assertion, it would be attributed to them, rather than to nameless 'authorities'. The latter sometimes means that the journalist/reporter made up the number [*107].

A different number had been recently reported, on 2 December 2007, in an article in the Houston Chronicle [*108]:

[Claude] Davenport estimates 150 to 200 commercial child exploitation Web sites exist globally.
[The article states that Claude Davenport is chief of the U.S. Immigration and Customs Enforcement's Child Exploitation Section.]

Explanations for the variance above include that some journalists mis-attribute estimates to officers, or that some officers pick 'estimated' numbers out of the air, or that some journalists pick numbers out of the air and imply the source to be someone other than the journalist. While it is impossible to determine the explanation with certainty, the variations do not engender confidence in the reliability of such apparently sloppy journalistic reporting of hopefully expert quantitative estimates.

Web sites: U.K. Internet Watch Foundation ("IWF") 2008 & 2007 statistics

On 29 April 2009, the U.K. Internet Watch Foundation ("IWF") (which runs a hotline) published its 2008 Annual Report which states that there were 1,546 web sites (domains) containing potentially illegal child sexual abuse material ('indecent photograph of a child' under U.K. law) known to the IWF during the 2008 calendar year and that this was a decrease of 9% since 2007 (and a 21% decrease since 2006).

However, the reported 9% decrease from 2007 is curiously low given the number of domains reported in the IWF 2007 Report was 2,755. If that number for 2007 was accurate , and if the 1,546 number reported for 2008 is also accurate, then the decrease would be 44.83%, not 9%.

Apparently, statistics reported by the IWF, either in its 2008 Report, or in its Reports for 2007 and prior years, are incorrect. The 2008 Report indicates the IWF has revisited the numbers it reported in previous years and significantly reduced same in its 2008 Report - either the numbers were significantly exaggerated in prior year reports, or the decrease in known domains from 2007 year to 2008 year was 45%, not 9%.

A summary of domain number statistics reported by the IWF in its 2008 and 2007 reports follows:

2008 IWF Annual Report: 1,546 domains:
of which

74% (1,144) were commercial domains and
26% (402) were non-commercial domains where images are stored or swapped;
the vast majority (81%) of the total 1,546 domains were live (and therefore on the IWF blocking list) for 100 days or less, and
60% were live for 50 days or less;
there has been a 9% decrease since 2007 (and a 21% decrease since 2006) in the number of domains confirmed to contain indecent images of children, known to the IWF, during 2008.

2007 IWF Annual Report: 2,755 domains (the number having remained relatively static during the last three years according to the IWF): of which

2,204 (80%) were commercial operations and 551 (20%) were non-commercial domains;
the vast majority (71%) [of the total 2,755] were live for less than 50 days during the year.
there was a 10% decrease in the number of domains from 2006.

However, as mentioned above, either statistics reported by the IWF in its 2008 Report, or in its Reports for 2007 and prior years, are incorrect. This is evident in graphs contained in the IWF's reports. [... the links here do not work - Ipce ...]

[...] The 2007 report claimed a total of 2,755 domains were known during 2007, but the 2008 Report claims only approx. 1,700 were known during 2007. A similar discrepancy exists in relation to 2006 year numbers.

Source: IWF Annual Report 2007 page 7, 17 April 2008 & IWF Annual Report 2008 page 7, 29 April 2009

Similarly, the historical charts for the years since 1997 published by the IWF in its 2008 report show significantly lower number of domains known each year than previously published in the 2007 report. For example, the 2007 Report showed a peak of around 3,000 in 2006 year (3,077 reported in prior IWF report), but the 2008 Report shows only 2,000 in the 2006 year.

Source: IWF Annual Report 2007 page 6, 17 April 2008 & IWF Annual Report 2008 page 7, 29 April 2009

It is also curious that 2007 Report shows an almost flat line from 2003 to 2004, i.e. almost no increase, while the 2008 Report shows a marked increase from 2003 to 2004, and a similar discrepancy in relation to the 2005 to 2006 years.

Web sites: U.K. Internet Watch Foundation ("IWF") 2007 statistics

Note:
The 2007 statistics [...] have been brought into question as a result of numbers published by the IWF in its 2008 Annual Report (which indicates the numbers in 2007 were lower than reported in its 2007 Report); see above re IWF 2008 Annual Report.

On 17 April 2008, the U.K. Internet Watch Foundation ("IWF") (which runs a hotline) published its 2007 Annual Report [*109] and announced that the number of web sites containing potentially illegal child sexual abuse material ('indecent photograph of a child' under U.K. law) known to the IWF during the last three years had remained relatively static.

According to the report, the number known to the IWF during 2007 was 2,755 (a 10% decrease in the number of domains from 2006), of which 2,204 were commercial operations. There had also been a 15% decrease in the number of individual web pages during 2007 (down from 10,656 URLs of individual web pages or websites stated in the 2006 Report. The 2007 Report does not state the actual number). The Report also stated that:

"During 2007 the vast majority (71%) [of the total 2,755] were live for less than 50 days during the year."

Hence, considerably less than 2,755 would have been 'live' at any given time (due to law enforcement action and/or ISPs becoming aware of illegal material having been uploaded to their servers and deleting/disconnecting same). The IWF states that its list [*110] is updated twice a day and typically contains between 800 and 1200 live child abuse URLs at any one time and that the URLs are precise websites or web pages.

Reuters reported on 16 April 2008 that:

"This is the first time any organization has revealed the true scale of this issue and been clear that the problem is something that can be solved, the [The Internet Watch Foundation] said in a statement.

Chief Executive Peter Robbins said the new figure would help build the case for a global drive to eradicate the sites. He said: A coordinated global attack on these Web sites could get these horrific images removed from the Web.

Speculative figures can create a distorted picture of the scale of the problem of child sexual abuse Web sites, he added. The number of child abuse sites has remained static over the last few years, despite the growth of the Internet, he added.

The watchdog's annual report called for a worldwide campaign by governments, police, and the Internet industry to investigate and disrupt abusive sites.
...
Computer networks in Russia and the United States host the most child abuse images, although many other countries are involved, a watchdog spokeswoman said.

It can be hard to shut illegal sites because operators constantly switch countries, temporarily close them, or hop between different Internet hosting companies."
Source: "Watchdog wants global drive against online child abuse", Reuters, 16 April 2008 [*111]

The IWF 2007 Report also contains a graph [...] showing the number of web sites known to the IWF during the years 1997 to 2007 inclusive. According to the graph, the numbers were relatively static

during the period 2002 to 2004: approx. 1,800 - 1,900 domains/sites; and
during the period 2005 to 2007: approx. 2,800 - 3,100 domains/sites.

The IWF Reports offer no suggested explanation for the approx. 50% increase between those two three-year periods, when the number known to the IWF increased from 1,894 during 2004 to 2,966 during 2005 (according to the 2006 Report).

However, changes in U.K. law effective from May 2004 (Sexual Offences Act 2003) may be relevant. In 2004, the age of a child, for the purposes of determining whether material is an 'indecent photograph of a child', was increased from 16 years to 18 years. The new legislation also created a conditional defence for IT professionals to enable them to legally copy and store, for a limited time, potentially illegal material depicting children found on company networks for the purpose of reporting/providing same to law enforcement officers or the IWF.

From September 2004 to June 2005, the IWF ran advertising/awareness campaigns directed at encouraging IT professionals to report such material, as IWF research had found that over 70% would not previously report same (presumably due to the prior lack of defence in relation to possession/copying) [*112].

The foregoing changes/events would presumably result in a greater number of sites becoming known to the IWF from 2004 because previously legal material had become illegal, and more people (e.g. IT professionals) were likely to lodge reports.

It is also notable that the number of reports continued to increase at the same rate after 2004, at which time the UK's largest ISP British Telecom and subsequently some other ISPs in the UK implemented systems to block accidental access. One may speculate as to whether this is because blocking systems are ineffective because illegal sites frequently move according to the IWF (perhaps the blocking has the effect of warning site owners of discovery), or whether many people lodge complaints about material that is not illegal.

Irrespective of the original basis of the year 2000 estimate of 100,000 websites, such an old 'statistic' cannot have any credibility over eight years later, given the massive increase in, and successes of, international police operations, since approx. 2002, directed to capturing distributors of child sexual exploitation/abuse material and closing down such web sites; and the enactment of criminal offences, or heavier penalties, to deter establishment of new sites, by numerous governments. The recently alleged 'estimates' above, and the IWF 2007 findings, appear more likely to have some connection with reality.

Peer to peer (P2P) networks:

Special Agent Flint Waters (Lead Agent for the Wyoming Internet Crimes Against Children Task Force) statistics.

(Note:
According to the ACMA filtering trial report issued in July 2008, none of the filters were capable of filtering/blocking illegal content on peer to peer (P2P) networks.)

3 October 2007: "Child Sex Crimes on the Internet" [*113], prepared for: House Judiciary Committee, by Flint Waters, Special Agent, Wyoming Internet Crimes Against Children Task Force, Wyoming Attorney General Division of Criminal Investigation:

"The statistics herein come from documented observations of one particular type of technology being used to facilitate child exploitation globally. Therefore, at most, the staggering numbers reported reflect a small portion of the severity of this problem given the growing form of predation facilitated by several types of technology associated with the Internet. Prior efforts to measure the use of technology in child exploitation have proven difficult due to the complexity of the systems leveraged by Internet predators. However, this report is able to provide some clear insight into the use of Peer to Peer networks in this type of crime.
...
Investigators deploying software written by the State of Wyoming have identified a vast network of traffickers who have distorted the original uses of Peer to Peer (P2P) networks to feed their own needs. The tactics being deployed by law enforcement have resulted in the identification of staggering numbers of individuals trading child sexual abuse movies and images.

The details you are about to review originate from a single P2P network, one of many used daily on the Internet. These details relate to just one small corner of the Internet. It applies only to one P2P system where child sexual abuse movies and images were presented to undercover law enforcement throughout the world. This data does not include traders using email, chat, social networks, news servers or paid and free web sites. At most it can be seen as a bare minimum of the trafficking of child sexual exploitative materials.
...
The software used on this particular network maintains a unique serial number for each installed system.
...
In the chart labeled Distinct P2P use we can see that over forty-nine thousand unique systems were found trading child sexual abuse imagery in August, 2007. That number represents the latest statistics available at the time of this report and we can see a continuing trend in the increase of this activity even though law-enforcement has been trying to disrupt this system for three years.

The monthly totals listed only depict unique use during that month. In most cases these users were also reflected in prior months. A review of the complete seven month period reveals 193,626 unique computers in the United States located by law enforcement trafficking child sexual abuse imagery. This ability to track serial numbers was implemented in late 2005. Since that time we have identified 377,044 unique serial numbers related to this activity.
[emphasis added]

16 April 2008: Testimony of Special Agent Flint Waters [*114], Lead Agent for the Wyoming Internet Crimes Against Children Task Force, Wyoming Division of Criminal Investigation Office of the Attorney General, for the United States Senate Committee on the Judiciary Subcommittee on Crime and Drugs - "Challenges and Solutions for Protecting our Children from Violence and Exploitation in the 21st Century":

"We can't blame the peer-to-peer systems. We can't blame chat rooms or social networking sites. We are a society of technological advance. Sadly, some leverage those advances to hurt children. Blaming this problem on peer-to-peer innovation is like blaming the Interstate highway system when someone uses it to transport drugs.

What we have to do is scale our law enforcement, prosecutorial and judicial resources to ensure we, as a society, are prepared to respond to the challenges that come from innovation, that we are prepared to rescue children when the map to their abuse is sitting right in front of us.
...
The identification of unique serial numbers that can be traced to the United States is still on the rise. Currently we only capture the serial number in about half of our undercover communications. We have exceeded 624,000 unique serial numbers that we can trace to the United States.
...
I would like to be clear, I am NOT saying law enforcement isn't doing enough with what they have. I am saying they could do so much more if they only had the resources.

Senators, I would ask you to picture the pile of work you have to leave waiting at the end of your day. While you want to make more progress at some point you have to go home, as it is there isn't enough time for home and family.

Now imagine that in your inbox are hundreds of leads. And as you leave the office to go home, you know you are walking away from dozens of children who are waiting to be rescued. Each of those children must wonder if anyone cares."
[emphasis added]

It should be noted that the population of the U.S.A. is approximately 300 million and that the number of computers identified is equal to less than 1% of the 70 million personal computers sold [in 2007] in the USA [*115].

The foregoing information is included to avoid perceptions along the lines that there is an actual or potential child molester on most streets in the U.S.A. It does not, however, suggest that there is not a major problem in relation to the use of P2P networks for trafficking child sexual abuse images. To the contrary, there obviously is a major problem - probably also in many countries. And it is one that cannot be reduced by ISP-level 'blocking' of web sites.)

See also: Special Agent Flint Waters' response [*116] to Wayne MacPhail's column "How the media can misrepresent the Web" [*117], 28 February 2008

4. Media Misreporting about Web Sites and Police Operations

Contributing to false perceptions about the number of web sites containing child sexual abuse material is the Australian media's tendency, whether through shoddy journalism, technological cluelessness, or some other reason, to incorrectly report that persons arrested had been using web sites to distribute or obtain child sexual exploitation/abuse material, when in fact they had not been using web sites but an entirely different Internet technology.

Examples include, but are not limited to, Australian media reports

in March 2008 about 'Operation Achilles'
(Usenet newsgroups, not a web site as all Australian media outlets incorrectly claimed) and
in June 2007 about 'Operation Lobate'
(peer-to-peer hosted chat room and filesharing network, not a web site as various Australian media outlets and the then Minister for Justice claimed).

Media mis-reporting not only gives the general public and politicians the false impression that use of web sites is the principal means of distribution, it also seems likely to give criminals using other Internet technologies the (false) impression they are safe from detection because it appears from media 'reports' that law enforcement agencies' activities are focussed solely on the use of web sites.

Unfortunately, media mis-reporting about web sites also fuel calls for types of governmental policy and associated taxpayer funded expenditure, such as 'blocking' web sites, that have no effect in reducing child sexual abuse and miniscule, if any, effect in reducing use of Internet technologies to distribute and access material depicting such abuse.

March 2008: Operation Achilles - Australia, USA, UK, Germany, etc.

In March 2008, Australian television and newspaper journalists reporting on the arrest of 22 members of 'a highly secretive and organised network of international child sex offenders' claimed that they had been using a 'highly secret website' (ABC Lateline, 5 Mar), an 'encrypted website' (Courier Mail, 6 Mar), etc., to communicate with each other and trade files, etc.

Meanwhile law enforcement agencies in the U.S., U.K and Germany had announced, and various overseas media outlets reported, that the 'network members' had been using newsgroups.

(The writer has been unable to find any Australian media outlet report which referred to newsgroups).

The U.S. indictment, publicly released, made clear that 'newsgroups' meant Usenet newsgroups. An FBI affidavit/arrest warrant for one of the 22 'network members' stated the network/group had been using PGP to encrypt messages and image files before posting same to whichever Usenet newsgroup they were using to share allegedly illegal material from time to time.

"Members of Vast Child Exploitation Enterprise Indicted in the Northern District Of Florida" [*118], U.S. Department of Justice, Media Release, 4 March 2008
"Eight UK children identified in images seized from international paedophile ring" [*119], Media release, U.K. Child Exploitation and Online Protection Centre, 5 March 2008
"Internationaler Schlag gegen Kinderpornografie: BKA durchsucht Wohnungen von fünf Beschuldigten in Deutschland" [*120], BKA (German Police) Media Release, 6 March 2008
U.S. Indictment (U.S.A. vs. Michael Berger and 11 others) [*121], 21 February 2008
Parts of FBI affidavit/arrest warrant (USA vs. D. Castleman) [*122], 29 February 2008

- - -

(Note:
A sub-operation of 'Operation Achilles' named 'Operation Koala' (EU)/'Operation Pariah' (Qld/AU) discovered and shut down in Jul/Aug 2006 several web sites run by an Italian photographer (who was arrested) and subsequently arrested customers. Those arrests and the completion of 'Operation Koala' were announced by law enforcement agencies in November 2007, and re-announced in March 2008).

June 2007: Operation Lobate-AU, Operation Chandler-UK, Project Wickerman Two-CA

People arrested were alleged to have been using a peer-to-peer ("P2P") hosted chat room and associated P2P file sharing network. However, some Australian media and a then Government Minister incorrectly claimed they had been using a web site to trade files etc.

The chat room, hosted on a personal computer in the U.K., was known as 'Kids the Light of Our Lives'.

(It was a reinstatement of a previously shut down peer-to-peer chat room - known as 'Kiddypics & Kiddyvids' - hosted through the WinMX software on a computer in the U.S.A. [*123])

Reports making the use of peer-to-peer ("P2P") technology clear include:

"Australians arrested after police crack global paedophile network"[124], AM Program, ABC Radio, 19 June 2007
JIM GAMBLE [director of the UK police Child Exploitation and Online Protection Centre]: "These are private rooms, peer-to-peer networks."
"The fight against net crime"[125], Marc Cieslak, BBC Click (UK), 13 July 2007
In June 2007, Operation Chandler saw CEOP infiltrate a paedophile peer-to-peer network. ...
"Our strategy was to identify the co-ordinator of that peer-to-peer network, explained Jim Warnock, head of operations at CEOP."

Incorrect media reports and other claims about 'Operation Lobate' - which wrongly referred to a 'web site' - include but are not limited to:

"Questions Without Notice: Child Protection" [*126], Responder: Minister for Justice, Senator David Johnston, Senate Hansard, 20 June 2007
When speaking about Operation Lobate, the Minister said:
"I can confirm that the Australian Federal Police has been an active partner in the international operation to crack the transnational paedophile ring which exploited and abused innocent children. This operation was code named Operation Lobate.
...
As part of the undercover operations, our agents worked as part of the team that actually took over the website, and gathered information on some 700 people around the world who were accessing this website.
"Online sting nabs 63; Victorian dad on child porn charge", Carly Crawford, Herald Sun, 20 June 2007
Aussies were among those allegedly using a UK-based pedophilia website
[AFP assistant commissioner Roman] Quaedvlieg said the depraved website was not open to the public, but accessible only to those who knew how to find it.
"Man had 34,000 child porn images, court told" [*127], ABC News, 14 April 2008
[The man] was charged in March last year as part of an investigation into a UK-based website known as 'Kids - The Light Of Our Lives' .

June 2008: Operation Centurion

Numerous Australian media outlets reported on this operation in a manner that either definitely did, or was likely to, give the public the impression that 12 million people around the world had downloaded child pornography images from a (cracked/hacked) European web site during three days. In fact the number of people was, at most, approx. 144,000.

The writer considers this type of reporting highly irresponsible because it seems extremely likely to give people engaging in such illegal activity a belief that there are millions of other criminals just like them, thereby 'normalising' such behaviour, and gives the public and politicians the impression that the problem is so massive that it is insurmountable by international law enforcement agencies.

According to ABC News:

"[Operation Centurion] started when a hacker infiltrated a respectable European website and placed 99 explicit images of young girls from eastern Europe, the US and Paraguay.
The site got 12 million hits worldwide in the days that followed and Australian police have traced some of those back to home addresses.
...
Commissioner Mick Keelty says it is hard to comprehend that 12 million people worldwide have looked at the explicit images."
Source: "Some child porn hits may have been accidental: police", ABC News, 5 June 2008 [*128]

It is not only hard to comprehend, it is impossible to believe among technology literate people (which does not encompass the entire population) because 12 million "hits" is not the same as 12 million "people". Nevertheless, according to an audio file of the AFP media briefing, Commissioner Keelty said:

"12 million is a big figure and it shows you just how much this problem has grown over a relatively short period of time. I mean the technology wasn't capable of involving that many people around the world in one offence 10 years ago, 20 years ago certainly."
Source: Australian Federal Police media briefing [Mp3 audio], LiveNews/Macquarie National News, 4 June 2008 [*129].

However, downloading 1 image results in one "hit". Hence, if every person who accessed the images downloaded all 99 images (which would happen automatically if all images were displayed on one web page), then 121,212 people accessing the images would result in 12 million hits on the site. This number is close to the 144,000 "users" reportedly referred to by Croatian police:

"Zagreb - Croatian police raided 187 homes, in one of the country's largest crackdowns on potential paedophiles, after investigations led to suspects in possession of child pornography. The international police action, deemed Sledgehammer, began in August 2007, when the internet site of a Croatian non-governmental organization was hacked and pictures of child pornography were put into its database.
Police stated that in three days, the site tallied some 12 million visits, after which police recovered the information of 144,000 users from 170 countries who accessed the material - almost 300 of which were located in Croatia.
After further investigations, 187 homes in Croatia were confirmed to have accessed the pornographic material.
...
According to Croatian law, a child pornography conviction can entail up to 10 years in prison."
Source: "Croatian police raid 187 homes in paedophilia crackdown", DPA, 22 Feb 2008 [*130]

Moreover, the 144,000 figure would be computer IP addresses, not necessarily 144,000 different users/people throughout 170 countries.

While numerous Australian media reports referred to 12 million "hits" (and a few referred to 12 million people), there appears to have been only one which mentioned the total number of IP addresses:

"The arrests followed a six-month sting into internet child pornography that originated in Europe. More arrests are expected in coming weeks with at least 1500 Australians - using 2883 IP addresses - under investigation.
...
AFP Commissioner Mick Keelty yesterday said the images had attracted 12 million hits - from 144,000 IP addresses across 170 countries - within three days of being posted on the website, described as a legitimate, non-government portal. The owner was unaware the images had been secreted on to the site, believed to be domiciled in Croatia.
Source: "Teacher suicides after child porn raid", Michael McKenna and John Stapelton, The Australian, 6 June 2008 [*131]

Operation Centurion also serves as a recent example of why 'blocking' web sites is an entirely ineffective means of preventing access to such material. As noted above, a legitimate web site was broken into and illegal material placed on it. It was shut down as soon as the site owners/law enforcement authorities became aware of the material. During the intervening three days, tens of thousands of people had accessed the material, reportedly having found out about the hacked site from temporary non-public chat rooms such people use to communicate:

"Mr Colvin [AFP director of high tech operations] said offenders used live chat rooms, quickly set up and taken down, and code words.
People who are engaged in this type of activity ... they know how to follow the links, he said.
It only needs a couple of messages posted on a chatroom to say you should look at this website.
That message spreads around the world like wildfire.
Mr Keelty said people had to prove themselves to become part of a child porn network.
To get in you've actually got to be supplying something yourself and that means more children will be abused, he said."
Source: "90 in court in child porn crackdown", Karen Davis, AAP/News.com.au, 5 June 2008 [*132]

AFP Commissioner Keelty also referred to the unlikelihood of finding illegal images via e.g. a search on Google:

"Commissioner Keelty: You [meaning AFP investigative officers] can tell the difference between an accidental logging into a particular site, you can tell the difference between that and somebody who's gone through a chain of sites to lead to them to where this sort of material sits ...
There's some people you might have some suspicion about but your priority is always going to be for those people you who know have channelled their way through the various firewalls, the various sites, on the Internet to actually get to the one that they want and part of the way that they do this of course is, they're not obvious, I mean, you're not going to put a Google search in for example and find some of this stuff, you're going to find it on unobvious sites, secreted in other images, or accessed by another way that the [child porn] network understands."
[emphasis added]
Source: Australian Federal Police media briefing [Mp3 audio], LiveNews/Macquarie National News, 4 June 2008 [*133]

December 2008: Operation Resistance - AU
September 2008: Operation Carousel - Brazil, Spain

These operations concerned the use of a peer-to-peer file sharing network, notwithstanding that an article in the Sydney Morning Herald, part of which was paraphased by the Australian Media and Communications Authority ("ACMA") in an April 2009 report, incorrectly claimed that 'Operation Carousel' concerned "websites".

ACMA's April 2009 Report claimed (page 69):

In September 2008 ... Investigators from Spain and Brazil employed various technologies including a computer program which detected and provided the name and location details of Internet users as they logged into websites hosting confirmed sexual abuse material.
Sydney Morning Herald (2008), 'Spanish police hold 121 in sweep against child porn'.
Source: "Developments in internet filtering technologies and other measures for promoting online safety", ACMA, April 2009 [*134]

The above SMH article claimed:

"Around 800 police took part in Operation Carousel, the biggest operation against child pornography on the Internet ever carried out in the country [Spain]. ...
Police launched the investigation last year in cooperation with Brazilian police. ...
Investigators have been helped by the use of a programme known as Hispalis, which allows them to detect and provide names and addresses of Internet users as they are logging on to illegal websites."
Source: "Spanish police hold 121 in sweep against child porn", SMH/Associated Press, 1 Oct 2008 [*135]

However, both 'Operation Carousel' and 'Hispalis' concern peer-to-peer ("P2P") networks, not websites.

According to the Guardia Civil (Spanish police), 'Hispalis' is a search engine which uses file hashes to find files on P2P networks containing child sexual abuse material previously identified by police:

"['Hispalis'] is capable of tracking network "p2p" to find computers that are downloading, possessing or share pictures and videos of a pedophile. i.e., able to detect any movement of these files in pornographic networks "p2p". ... an initiative of the Guardia Civil after seven years working in this field, and consequently have found that pedophiles, for concealment, are migrating from conventional networks to the "p2p" trying to escape police action."
Source: "Presentado el 'buscador Híspalis' dentro de la operacion 'Azahar' contra la pornografía infantil", Guardia Civil, 27 Oct 2005 [*136] &
"Alonso destaca el papel pionero de la Guardia Civil en la lucha internacional contra la pornografía infantil en Internet", Guardia Civil, 24 Feb 2006 [*137]

'Operation Carousel' concerned peer-to-peer file sharing according to Brazilian police:

In late 2007, Carlos Miguel Sobral and 14 other Brazilian police investigators were getting ready for another day of fighting Internet crime when one of them suggested looking into peer-to-peer file sharing networks.
...
The recent arrests in Australia and other countries are a direct result of our Operation Carousel, which we set in motion in December 2007 as part of our routine efforts to combat the criminal use of the Internet, said [Carlos Miguel] Sobral, who heads the [Brazilian] federal police department's cybercrime unit.

The investigation focused on men sharing images over the Internet - peer-to-peer file sharing - rather than downloading from an illicit Web site, which is easier to detect."
Source: "Investigators Find 250 People in 78 Countries Involved in Child Porn File Sharing", Fox News, 12 December 2008 [*138]

The related arrests in Australia in December 2008 were the result of the Australian Federal Police's 'Operation Resistance' which the AFP said concerned peer-to-peer file sharing:

"Assistant Commissioner Gaughan says the operation [called Operation Resistance] began last December after the AFP received information from the Brazilian Federal Police.
We, through our international liaison network, received information from the Brazilian police that there was a large amount of males throughout the world in over 70 countries involved in the trading of this sort of information in a peer to peer network, he said."
Source: "QC, childcare workers arrested in child porn sting", ABC News, 11 Dec 2008 [*139]

- - -

"The Federal Police began their investigation after a tip-off from Brazilian police. Mr Colvin [AFP] said the network had exchanged the images and videos through peer-to-peer software. He said the network shared files directly, with some having to prove their credentials by uploading images."
Source: "AFP busts porn ring in three states", David Stockman, Canberra Times, 12 Dec 2008 [*140]

- - -

"Neil Gaughan [Aust. Federal Police]: Peer to peer technology allows Internet users to share files without actually accessing a central network ...
Philip Clark: So filters and all that sort of thing don't work?
Neil Gaughan: They don't work at all. ...
Neil Gaughan: ... we have people operating in undercover capacity who actually get online and ... pose as paedophiles and assist and breakdown those barriers. ... [We] use terminology to make people believe we're actually involved in that trading and as such people will share files with us. Once that's occurred we move forward and take the necessary action."
[emphasis added]
Source: Interview - AFP Acting Commissioner Neil Gaughan re 'Operation Resistance', Philip Clark, 2GB Radio, 11 December 2008 [*141]

Download:

graham_refs_and_endnotes.pdf