As a quick trailer, I’m currently working on an article which looks in detail at a ‘fact and figures‘ page published by Gail Dines on her ‘Stop Porn Culture’ website in which, amongst other things, she cites a number of ‘statistics’ sourced from Jerry Ropelato’s ‘Internet Filter Review’ (IFR) website.

Now regular visitors to the Ministry will know that I’ve written about Ropelato and his ‘statistics’ before, but before I give Dines the full once over I think it worth recapping what is already known about one of her main sources of statistical information for both her website and, as you’ll see at the end, her 2010 book ‘Pornland’.

IFR is part of a site called ‘Top Ten Reviews’ which (amusingly) promotes its wares with the trademarked strapline “We do the research so you don’t have to?” and that site, in turn, is part of a company called ‘TechMedia Network‘, which bills itself as one of the “largest, most respected technology media companies in the world”, a company that was founded in 2003 by a Utah-based businessman and entrepreneur, Jerry Ropelato. Regardless of any other qualities he may possess, Ropelato is, by some distance, one of most prolific and successful statistics launderers of the Internet age.

What do I mean by statistics laundering?

Well, to begin with we need to consider the basic business model adopted by ‘Top Ten Reviews’ and other similar site, which don’t actually sell any products direct to the general public. What these sites provide, of course, is product comparison information and recommendations based on conducting group tests of similar products, much as the Consumer Association’s  Which? magazine has been doing in the UK since the late 1950’s. Pre-Internet, the way you’d generate revenue from this kind of activity was almost exclusively by way of publishing your findings in magazines or other subscription publications but with the advent of the Internet came new ways of monetising this kind of activity, including click-through advertising and affiliate marketing schemes, and these were the core of Ropelato’s original business model. People would visit his site, look at the product comparison information and click a button or link under the product that most interested them, taking them to a different website where they could purchase the product in question – and of course every time someone clicked-through to a sales website or, better still, actually bought a product based on one of Ropelato’s recommendations then kerrrr-ching, a bit of commission heads in the opposite direction to the owner of the site that made the recommendation.

This is all well and good if, like the Consumers’ Association, you operate with complete independence from any of the suppliers whose products you’re reviewing, but it does a lot murkier when the company publishing the reviews has a close relationship with one or more of the suppliers whose products it’s reviewing, and if you look at Ropelato’s personal biography at TechMedia Network you’ll notice that it includes the following:

Prior to TechMedia Network, Mr. Ropelato served as Chief Operating Officer and Chief Technology Officer of ContentWatch, Inc. and Chief Technology Officer of iAccess.

ContentWatch Inc, which is also based in Utah, produces and sells internet filtering systems to both home users and commercial businesses and although its products do appear to perform consistently well in independent tests conducted by tech publishers such as ZDNet it is almost certainly no coincidence that from the earliest archived version of Ropelato’s Internet Filter Review site I could track down (December 2004) through to the beginning of 2007 it was ContentWatch’s ‘ContentProtect’ software that held down the top spot in Ropelato’s reviews, a situation which only changed after the company bought out Net Nanny, taking over the brand name for use as its main consumer brand, at which point [unsurprisingly] Net Nanny took over the top spot in Ropelato’s group test, where it’s been ever since.

Irrespective of Ropelato’s prior association with ContentWatch Inc, his group tests are worthless if they don’t generate those all-important revenue generating click-throughs and if you’re in the security business, whether it’s online or in the real world, then by far the easiest and most effective way of doing is that is by raising the anxiety levels of your target audience. Fear sells, and the psychology behind that statement is pretty easy to understand; the extent to which we feel threatened by something depends on the extent to which we perceive ourselves to be vulnerable to that something, whatever it may be, and our perception of the likely consequences should we be affected. In short, “How likely am I to get hurt?” and “How much will it hurt?”. The third part of the equation is control, or at the very least the illusion of control. If the threat is something we feel we cannot do anything about then we’ll do nothing, so having raised the anxiety levels of your intended audience you then have to sell them the idea that the things that’s making them feel threatened can be avoid if only they buy whatever it is you happen to be selling.

Ropelato was, and still is, peddling software that gives parents the illusion of control over their children’s web surfing habits – and I say ‘illusion’ here because, of course, it’s not parents who are really in control here, it’s the software companies who decide what kind of information and which websites their software will block access to – so that part of the equation is covered, but what about the first two parts? There Ropelato relies a considerable extent on a page of supposed ‘internet statistics’ which, right from the outset, had included alarming looking claims such as:

Pornographic websites: 4.2 million (12% of total websites)

And…

Daily Pornographic emails: 2.5 billion (8%of total emails)

Average daily pornographic emails/user: 4.5 per Internet user

And…

Sexual solicitations of youth made in chat rooms : 89%

And…

17% of all women struggle with pornography addiction.

Women, far more than men, are likely to act out their behaviors in real life, such as having multiple partners, casual sex, or affairs.

Yup, those evil pornographers are out to get you and your family and if you don’t do something about it then one day you’ll wake up to find that your kids have been molested and your wife is a raving nymphomaniac.

Ropelato’s ‘internet statistics’ page has been around since he opened up his Internet Filter Reviews website in 2003 and with a little effort and the help of the Wayback Machine internet archive it is possible to track the history of its development and how it’s changed over time. The oldest version I could source dates to December 2004 (pdf) and you’ll see that all the supposed ‘facts and figures’ I’ve given above are already in place. Step forward three years to December 2007 and you’ll see the page has been updated and given both fresh lick of paint and a couple of new sections covering what Ropelato purports to be detailed information on porn-related search traffic and the scale of the global porn industry for the years 2005 and 2006. Look closely, however, and you’ll also find a lot on that page hasn’t changed – it still claims, for example, that there are 4.2 million pornographic website and that this is still 12% of the total number of websites on the Internet.

At the same time that I grabbed these two archived versions of that page (January 2013) I also grabbed what was then the current version of Ropelato’s ‘statistics’ page and again you’ll see that both the site and the page has been tarted up but when it comes to the figures on the page, little or nothing has changed over time. In January 2013, the site was still claiming that there are 4.2 million pornographic website and that this amounts to 12% of all search traffic and although the page has since been updated again, with a lot of flashy bells and whistles, it still today makes that exact same claim.

The other thing that Ropelato added to that page in and around December 2007 is a list of claimed sources for his ‘statistics’.

Sources:

Statistics are compiled from the credible sources mentioned. In reality, statistics are hard to ascertain and may be estimated by local and regional worldwide sources.

ABC, Associated Press, AsiaMedia, AVN, BBC, CATW, U.S. Census, Central Intelligence Agency, China Daily, Chosen.com, Comscore Media Metrix, Crimes Against Children, Eros, Forbes, Frankfurt Stock Exchange, Free Speech Coalition, Google, Harris Interactive, Hitwise, Hoover’s, Japan Inc., Japan Review, Juniper Research, Kagan Research, ICMEC, Jan LaRue, The Miami Herald, MSN, Nielsen/NetRatings, The New York Times, Nordic Institute, PhysOrg.com, PornStudies, Pravda, Sarmatian Review, SEC filings, Secure Computing Corp., SMH, TopTenREVIEWS, Trellian, WICAT, Yahoo!, XBIZ

And yes there are a number of credible looking sources included in that although exactly how many of them might have actually served as sources of information is anyone’s guess because the page doesn’t provide any information or backlinks that might help to identify which figures came from where let alone where and when any of those ‘statistics’ first entered the public domain and in what circumstances. So as validating any of Ropelato’s claims this section is about as much use a chocolate teapot and as we’ll see in another article that I’m putting together at the moment, which I’ll publish later this week, that’s unlikely to be accidental given where and when some of the figures that Ropelato has been publishing as ‘statistics’ about online porn actually originated.

Okay, so the guy’s using a bunch of dodgy and hopeless outdated zombie statistics to pimp his business – does that actually matter?

Well yes, and to illustrate exactly why let’s just one of the ‘statistics’ that Ropelato has been publishing for at least 10 years, the one that claims that there are 4.2 million pornographic websites, and see what we find if we run that claim through Google. as part of the putting this article together I ran two exact phrases searches, one for “4.2 million pornographic websites” the other for “pornographic websites 4.2 million” and together these two searches returned a little over 68,000 results.

Now a lot of those results turned out to be blog posts and social media traffic from Facebook and Twitter where you wouldn’t necessarily expect people to be checking their sources, and of course this ‘statistic’ can also be readily found on a wide range of US-based Christian websites, although it and other figures taken from Ropelato’s ‘statistics’ page are also finding their way on to Islamic websites as well.

That last observation brings me neatly to what is easily most bizarre example of the “4.2 million porn sites” claims I’ve come across to date, this video, with a translation by Memri, of an Egyptian cleric blaming the supposed existence of all these porn sites on a Jewish plot to destroy Islam:

[youtube]https://www.youtube.com/watch?v=s89Jc0o2k0o[/youtube]

Nor should it comes as any great surprise to finds that Ropelato’s figures have gained some popularity with both sex/porn addiction therapists and the content filtering industry – although to be fair that last page dates to 2008 and doesn’t appear to feature at all on their current website.

As for politics, here you can see the Kansas branch of ‘Concerned Women for America’ using the 4.2 million porn site claim in 2013 in support of legislation requiring all Kansas school districts to implement mandatory child porn blocking and the same ‘statistic’ also appears in recently revised policy overview paper published by the Center for Arizona Policy, which was also behind a recent attempt to pass a law allowing business owners and employees to refuse to serve people on religious grounds – no prizes for guessing which minority groups that law was aimed at. Okay, so maybe you expect to see that kind of thing in the US, but what about India as well where, within the last week or so, the claim that there are 4.2 million porn sites was cited by lawyers representing the state government of Kerala in the Kerala High Court in support of an application to block a number of pornographic websites on the back of claims that online porn is responsible for a recent rise in a number of crimes, including rape.

You might think academics would be a little more careful when it comes to checking their sources, but of course you’d be wrong as this 200+ page 2009 report on Internet blocking, the production of which was which was funded by George Soros’ Open Source Initiative, clearly shows. The specific reference to the “4.2 million porn sites” figure is on page 63 of the report but before looking it up, please do take the time to read through pages 3 & 4, which set out in some considerable detail the rather impressive looking credentials of the report’s four co-authors, especially those of Marko Gercke who is also the author of section 1.6 of the International Telecommunications Union’s 2008 Global Strategic Report, which forms part of its Global Cybersecurity Agenda. Now see if you can guess what’s on page 36 0f that ITU report?

16.2.2. Making pornography unavailable to minors

Sexually-related content was among the first content to be commercially distributed over the Internet. Recent research has identified as many as 4.2 million pornographic websites that may be available over the Internet at any time.¹⁰²

…

102 Ropelato, “Internet Pornography Statistics” – http://internet-filter-review.toptenreviews.com/internet-pornography-
statistics.html.

And you’ll also find the 4.2 million porn site claim cited in this 2012 paper from the Journal of Chinese High Technology Letters, in which the authors propose a new method of automatically detecting pornographic images and in this 2013 Russian paper on internet porn filtering, which is published as part of the refereed proceedings of the 15th International Conference on Speech and Computer and, indeed, in this 2008 paper published in the journal Crime and Delinquency, which notes that the 4.2 million porn site claim also appears in a chapter on obscenity and the Internet by Jeff Maahs and John Liederbach in the 2006 book “Current Legal Issues in Criminal Justice: Readings“.

And if you think that’s perhaps a little embarrassing then spare a thought for G David Garson, Professor of Public Administration at North Carolina State University, who boasts some extremely impressive academic credentials:

G. David Garson is a full professor of public administration at North Carolina State University, where he teaches courses on American government, research methodology, computer applications, and geographic information systems. He was the recipient of the Donald Campbell Award (1995) from the policy studies organization, American Political Science Association, for outstanding contributions to policy research methodology and of the Aaron Wildavsky Book Award (1997) from the same organization. He is the author of Guide to Writing Quantitative Papers, Theses, and Dissertations (Dekker, 2001), Neural Network Analysis for Social Scientists (1998), and Computer Technology and Social Issues (1995). In addition he is editor of Social Dimensions of Information Technology (2000), Information Technology and Computer Applications in Public Administration: Issues and Trends (1999), and the Handbook of Public Information Systems (1999). He has also authored or edited 17 other books and authored more than 50 articles. For the last 20 years he has served as editor of the Social Science Computer Review and is on the editorial board of four additional journals.

None on which succeeded in preventing him and his co-author, Mary Maureen Brown of the Center for Excellence in Public Leadership at NCSU, from citing almost the entire contents of Ropelato’s ‘statistics’ page in their 2013 book ‘Public Information Management and E-Government: Policy and Issues‘ – here, see for yourself, they’re on page 57. Brown and Garson even manage to include the utterly fantastical claim that the South Korean porn industry generates $25 billion in revenues per year, which at the time was the equivalent of $526 for every man, woman and child in the country, a feat that would be utterly remarkable (and unbelievable) even were it not for the fact that possession of any kind of pornography in South Korea is a criminal offence and subject to a maximum two-year prison sentence.

If academics can’t be bothered to check Ropelato’s figures properly then what chance is there that journalists might do any better?

Well, not much as evidenced by the 4.2 million porn site claims appearance in both Newsweek and the Daily Mail in November 2011, although here the Mail is recycling Newsweek’s copy in its article. You’ll also find the same ‘statistic’ cited in this 2011 article in Canada’s Financial Post, in this 2013 report on the Jordanian government’s moves to implement porn blocks from the Jordan Times, this 2006 article from the Jakarta Post and even this 2003 article from India’s Economic Times, which helps, of course, to confirm just how long this ‘statistic’ has been floated around unchecked. The Economic Times, by the way, is the second most read business newspaper in the world after the Wall Street Journal, so we’re not just looking at some shabby online tabloid here.

Just two more examples to go, the first of which is to be found in the NSPCC’s consultation submission to the 2011 Bailey Review on the Commercialisation and Sexualisation of Children:

There are currently around 4.2 million pornographic websites. These constitute about 1.5 per cent of all websites. Every day, search engines deal with around 68 million requests for pornographic material. This, combined with the proliferation of sexualised images in online advertising, shows that both pornography and sexualised images are becoming more widely available and easily accessible. (p2)

One thing that’s particular interesting about the NSPCC document is that although it credits Ropelato with the 4.2 million porn sites ‘statistic’ the citation given for the claim that search engines deal with around 68 million porn searches a day (another unsourced Ropelato original dating back to at least 2003) is Linda Papadopolous’ 2010 review of the Sexualisation of Young People, which was commissioned by the Home Office.

And, sure enough, when I checked the Papadopolous review, there on page 45 I found:

Each day, search engines deal with around 68 million requests for pornographic material – approximately a quarter of all searches on the net.²⁰⁷ This, combined with the proliferation of sexualised images in online advertising, suggests that both pornography and sexualised images are becoming more widely available and easily accessible.

…

207. Ropelato (2006)

…

Ropelato, J. (2006). Internet pornography statistics. TopTenReviews.com, internetfilter-review.toptenreviews.com/internetpornographystatistics.html, accessed May, 3, 2006

All of which I find particular interesting because the date on which Papadopolous states that she accessed Ropelato’s site predates the inclusion of any information about sources on that page, inadequate as that information was when it finally appeared, by around 18 months not to mention the publication of her report by around four years.

Oh, and we’ve rather drifted on to the equally spurious claim that 25% of internet searches are looking for porn I should perhaps note that one of the many locations that that claim turns up is the BBC Radio 1 Advice website, which is aimed at teenagers and young adults and which currently informs its visitors that:

Around 25 percent of all internet searches are for pornographic content, and on average almost 30,000 people are looking at porn at any given second.

Shockingly, an investigation by the London School of Economics found that 90 percent of youngsters between eight and 16 have accessed porn online, many without meaning to find it, and most while doing their homework.

That last 90% statistic is one that I’ve now fully sourced and I can tell you for a fact not only that it dates back to 2001 but also that it’s based solely on an off the cuff remark by an academic talking to a newspaper about a qualitative study of just 30 families which took place between 1999 and 2001 plus the usual journalistic hype and amateur stat mangling – but I’ll save the full story for a future post.

And, coming full circle, this is part of the opening paragraph of chapter 3 of Gail Dines’ 2010 book ‘Pornland’.

The size of the porn industry today is staggering. Though reliable numbers are hard to find, the global industry has been estimated to be worth around $96 billion in 2006, with the U.S. market worth approximately $13 billion. Each year, over 13,000 films are released, and despite their modest budgets, pornography revenues rival those of all the major Hollywood studios combined. There are 420 million Internet porn pages, 4.2 million porn Web sites, and 68 million search engine requests for porn daily.

All of these figures are sourced from Jerry Ropelato’s Internet Filter Review website, several date back to at least 2003, and for several of them Ropelato is [so far] the only identifiable source.

So after all that, just exactly what are we to make of the claim that there are, or maybe were, at some point around 4.2 million porn websites?

Well, only last year one of the new content blocking companies on the block, MetaCert, put out a press release and infographic giving their estimates for the amount of porn on the Internet and came out with figures of around 7 million individual domains and 700 million individual webpages of which the United States contributes just under 4.2 million domains and 428 million individual pages.

Now there’s a coincidence for you, at least 10 years after the 4.2 million porn site ‘statistic’ made its first appearance as a supposed global figure, up pops a new content blocking companies which cites almost the exact same figure just for websites operating out of the US and, even more bizarrely perhaps, that company also give a figure for the total number of individual porn web pages that is just 8 million higher than another statistic that’s been appearing on Ropelato’s site since around 2005, albeit one for which I do have a definite source – the 420 million pages statistic is based on figures released in 2004 by another content blocking company, Secure Computing Corporation, which went on to be acquired by McAfee in 2008.

Be that as it may, the single most important figure given in MetaCert’s is buried in the text underneath the infographic.

With over 7 million unique domains (or 700 million pages / 35 billion URLs) classified, it is by far, the world’s largest database of Internet pornography – greater than Norton’s, McAfee’s and OpenDNS’ databases combined.

So, to find these 700 million pages of porn, MetaCert had to check 35 billion individual URLs, which suggests that in terms of total content, porn accounts for just 2% of the Internet even before we take into account the rapid growth in the total amount of online content that’s being driven by social media – and based on the information given in this official blog post, Twitter’s normal daily traffic runs to around 500 million tweets per day, every one of which can viewed through a web browser as a single tweet with it’s own unique URL. In other words, it currently takes twitter less than 36 hours to generate the same number of individual webpages as MertCert’s estimate for the total amount of pornographic webpages on the entire Internet and, according to figures from the web metrics company SimilarWeb, Twitter generates something under a tenth of the amount of online traffic generated by Facebook.

Okay, so that’s the picture for URLS but what about these 7 million pornographic domains?

Well, according to Verisign, by the end of 2012 there were already over 250 million registered domains worldwide, although they did also report that around 21% of those led to single page websites, i.e. holding page, while 15% didn’t point to any kind of working website at all, leaving a balance of around 157 million working websites. That, as figures go, seems to fit pretty well with the latest data from RegistrarStats, which puts the total number of registered TopLevelDomains at a touch over 151 million domains, so even if we limit our calculations to just the numbers of active domains then our 7 million porn sites account for no more than 4.5% of all registered and active domains.

That still, doesn’t take into account domain turnover where the current figures from DomainTools show that just yesterday there were over 164,000 new registrations across just six TLDs – .com, .net, .org, .info, .biz and .us – to go with 132,000 deletions and around 150,000 transfers.

Something that no one ever seems to ask questions about whenever content blocking companies start spitting out claims about the size of their blacklists are the issues of redundancy and data cleansing or, to put it more simply, how many of the domains and URLs on their blacklist are still actually live at the time that the company releases its figures and do the company’s automated web crawlers ever make repeat visits to websites and URLs that have been previously added to their blacklist or does the blacklist just keep on growing over time without anyone ever checking to see whether or not most of its contents even still exists?

To give you an idea of the extent to which this might easily inflate the kind of figures cited by MetaCert and other content blocking companies I used a dedicated porn tube search site which is known to index eight of the top ten ranked porn tube sites to see how many porn videos it listed in its search index for each of those sites and then checked the sites to see exactly how many porn videos they actually contain. For two of those sites I was unable to get an accurate figure for the total number of videos because of the rather idiosyncratic manner in which their browsing facility works but of the remaining six none currently contain more than around 18% of the total number of videos indexed by the search engine site while one site was found to actually contain just under 8% of the total number videos listed for it in the search index.

Based on that admitted quick and dirty analysis it would appear that around 85% of the index entries collected by this one porn search engine over the last six years or so, which is roughly how long porn tubes have been around, are either duplicates they point to videos that have long since fallen down the Internet’s memory hole and we are talking here about an index of porn videos which, amongst everything else, includes several of the largest and most popular porn sites in the world, sites which are inevitably going to be amongst the first to be added to any content blocking company’s blacklist.

Now maybe Metacert does in fact routinely recheck and clean up its blacklist, and maybe it doesn’t – I really don’t know – but at the very least what my own checks show is that any kind of claims made about the amounts of porn that may be out there on the internet at any given time have to be treated with extreme caution. And yet, as we’ve seen by looking at just one unsourced and unverifiable ‘statistic’ the need for even the most basic caution is something that has escaped journalists, academics, NGOs and even people who appear to be widely considered to be ‘experts’ in online security issues.

So ask yourself, who and what can you actually trust?