Sex and the world wide web

Summary

More or Less, BBC Radio 4

Episode Summary: "Sex and the World Wide Web"

Date: July 6, 2013
Host: Ruth Alexander
Guests: Mark Ward (BBC Technology Correspondent), Brent Schrotenberg (USA Today), Ben Carter (More or Less team)

---

Episode Overview

This episode of "More or Less" investigates the statistical claims surrounding two much-discussed topics in the media: the prevalence of pornography on the Internet and crime rates among American football players, particularly following the high-profile arrest of NFL star Aaron Hernandez. The episode debunks widely cited "zombie statistics," offering insight into the reality behind sensational numbers, and draws context from expert analysis.

---

Part 1: How Much Porn Is There on the Internet?

Main Discussion Points

• Listener Questions: Listeners George and Deborah wrote in with questions about the pervasiveness of pornography online, including how usage is measured (e.g., number of websites, data traffic) and the validity of claims such as "the Internet is 37% porn."

  • Quote: "Figures like this. The internet is now 37% porn. CAN statistics like this really be right?" (Ruth Alexander, 00:42)

• Questionable Statistics:

  • Mark Ward notes that even the filtering software companies propagating big statistics about online pornography usage have backed off their claims.
    • Quote: "Even the company that put the statistic out has now said we don't think that's true." (Mark Ward, 01:27)
  • Motivation for inflation: Companies selling Internet filtering software benefit from amplifying the perceived 'problem.'
    • Quote: "That company makes filtering software... They're in the business of stopping you looking at this stuff." (Mark Ward, 01:35)

• Methodological Flaws:

  • Filtering software often mistakes sexual health sites for pornography, and the cited study was never published with transparent methods.
    • Quote: "They regularly confuse sexual health sites with sexual sites and that study has never been published and nobody really knows where they got their figures from." (Mark Ward, 01:49)

• Web Traffic Estimates:

  • Frequently cited statistic: 30% of all Internet traffic is porn-related.
  • Ward points out flaws:
    • Vast underestimation of total web traffic
    • Overestimation of traffic to porn sites
    • Every website click adds multiple requests—further confusing the count
    • Quote: "The maths done in that study are a bit suspect... at the very least, their estimate of web traffic was a small part of a much bigger total." (Mark Ward, 02:24)

• More Credible Estimates:

  • One more reputable study suggests pornography comprises about 4-5% of all web traffic (02:47).
  • In the top million websites, about 4% are dedicated to pornography.
    • Quote: "They found about 4% of those top million sites were dedicated to pornography." (Mark Ward, 03:04)

• Search Queries:

  • Around 14% of all search engine queries are sexual in nature, though this includes sexual health searches and is not solely "porn."
    • Quote: "That's about 14% of all the searches done on those search engines every day are for sexual subjects of one sort or another." (Mark Ward, 03:23)

• Academic Research:

  • While more detailed studies are emerging, "zombie statistics" persist—numbers cited repeatedly without a reliable source.
    • Quote: "[Academic studies] tend to declare their interests... than you would with all of these zombie statistics that just get bandied around by anybody and everybody." (Mark Ward, 03:48)
    • Quote: "Our thanks to BBC technology correspondent Mark Ward, valiant slayer of zombie statistics, which no doubt will still refuse to die." (Ruth Alexander, 03:59)

---

Part 2: NFL Players: Crime and Statistics

Main Discussion Points

• Context: Following Aaron Hernandez's arrest, headlines suggested a crime problem among NFL players. More or Less investigates how arrest numbers for football players stack up against the general American population (04:37).

• The Arrest Rate Myth:

  • Schrotenberg compiled an arrest database since 2000. There have been about 660 arrests—about 60 per year.
    • Quote: "The NFL arrest rate for active players is a lot less than the US general population arrest rate. Strikingly so." (Brent Schrotenberg, 05:03)
  • NFL player arrest rate: About 1 in 47-50 players per year.
  • General US population: About 1 in 25 people arrested per year.
    • Quote: "It turns out that they're still better behaved than regular society." (Brent Schrotenberg, 05:03)

• Breaking Down the Data:

  • Ben Carter further breaks down Bureau of Justice Statistics:
    • NFL players are all men: US male arrest rate is 1 in 15.
    • NFL players are young: For men aged 21–34, arrest rate is 1 in 8.
    • For a 23-year-old (Aaron Hernandez's age): 1 in 6.
  • Clarification: These are arrests, not individuals—repeat offenders drive the numbers up.
    • Quote: "It's not that one in six 23 year old men in the US get arrested every year. These figures are for arrests, not people." (Ruth Alexander, 06:35)

• Repeat Offenders:

  • NFL player Adam "Pacman" Jones cited or arrested eight times since 2005; such cases inflate overall numbers.

• Football Players Look Good:

  • Adjusted comparisons show about six times as many arrests among young American men as among NFL players.
  • Even US women (arrest rate 1 in 46) are marginally more likely to be arrested than NFL players (1 in 47).
    • Quote: "In general terms, there are about six times as many arrests among the general population of young American men as there are among football players." (Ben Carter, 07:00)

• Types of Crimes:

  • Most common: Drunk driving (about 30% of NFL player arrests).
    • Quote: "By far the most common type of arrest for an NFL player is drunken driving." (Brent Schrotenberg, 07:30)
  • About 20 out of 60 player arrests each year are DUIs—NFL rate is half that of males in their age bracket.
    • Quote: "The drunken driving arrest rate for NFL players is about half that of their age bracket in the U.S. general population." (Brent Schrotenberg, 07:30)

---

Notable Quotes & Moments

• On zombie statistics:

  • "We're very familiar with the problem of zombie statistics on More or Less. In fact, we even like to think we coined the term, gave a name to the beast." (Ruth Alexander, 03:59)

• On the impact of media attention:

  • "A lot of that has to do with the media attention. When an NFL player gets arrested, it's going to get in the news." (Brent Schrotenberg, 05:03)

---

Key Timestamps

• 00:42 – Introduction to the question: How much of the internet is porn?
• 01:27 – Mark Ward on dubious statistics and software companies’ motivations.
• 02:24 – Math flaws in Internet traffic estimates.
• 02:47 – More credible statistic: 4–5% of web traffic is porn.
• 03:23 – Sexual search queries: ~14% of total.
• 03:59 – "Zombie statistics" discussion.
• 05:03 – NFL arrest rate context: footballers vs. general population.
• 05:59 – US male and age-stratified arrest rates vs. NFL players.
• 07:00 – Repeat offenders and comparisons to US women.
• 07:30 – Most NFL arrests are for DUI, rate is below national average.

---

Takeaways

• The claim that "the Internet is 30–37% porn" is greatly exaggerated; more reliable estimates put porn at 4–5% of web traffic and about 4% of popular sites.
• Many much-cited statistics about porn, crime, and other contentious issues are "zombie statistics"—unsupported numbers that refuse to die.
• NFL players, contrary to media focus, are arrested far less frequently than the general population, especially when comparing age and gender, and most offenses are not violent crimes but drunk driving.
• Media coverage amplifies the impression of crime among NFL players, while statistical analysis shows a different, more reassuring reality.

---

For more episodes and detailed investigations of popular stats, visit the BBC World Service's More or Less website.

Transcript

[BBC Announcer]

Thank you for downloading from the BBC. For details of our complete range of podcasts and our terms of use, go to bbcworldserveys.com podcasts

[Ruth Alexander]

this is the short edition of More Or Less, first broadcast on the BBC World Service. Hello and welcome to More or Less on the BBC World Service. I'm Ruth Alexander. In this week's programme, we look at the offending rate among American football teams after what feels to some like a spate of arrests. The latest involving star player Aaron Hernandez, who's made headlines around the world.

[Brent Schrotenberg]

Former Patriots tight end Aaron Hernandez has now officially been charged with murder.

[Ruth Alexander]

But first, many things about pornography are exaggerated, including the statistics quoted to suggest that the Internet is awash with porn listeners. George and Deborah have both emailed us separately at more or lessbc.co.uk to ask about how much porn there really is on the Internet. How can usage be measured? George wonders, through the numbers of sites, data streaming, minutes viewed. And Deborah's been wondering about some of the startling figures she's been reading in the press. Figures like this. The Internet is now 37% porn. CAN statistics like this really be right? She asks. Well, the BBC's technology correspondent, Mark Ward has been looking into all this and he isn't so sure they are.

[Mark Ward]

No, I don't think anyone's sure about that. Even the company that put the statistic out has now said we don't think that's true. I'm not sure it was ever true because, not wishing to cast aspersions, but that company makes filtering software and filtering software that tries to stop anybody looking at things they'd rather not see. Be that.

[Ruth Alexander]

So they'd be in the business.

[Mark Ward]

They're in the business of stopping you

[Ruth Alexander]

looking at this stuff. There's a big problem out there. They are indeed a filter on your Internet.

[Mark Ward]

Exactly. But also I think the filtering systems, that they're better than they were, but they're still pretty bad. They regularly confuse sexual health sites with sexual sites and that study has never been published and nobody really knows where they got their figures from.

[Ruth Alexander]

What about how popular online porn is? One way of measuring that is to look at how much traffic these sites generate, that is how much data is sent and received by visitors to the websites. Some statistics that abound suggest it's a lot. Researchers have found that 30% of all Internet traffic is porn related, but Mark Ward just doesn't buy it.

[Mark Ward]

The maths done in that study are a bit suspect because they vastly underestimate the amount of traffic that's passing across the web. So at the very least, their estimate of web traffic was a small part of a much bigger total. And then they overestimated the amount of traffic that your average porn site gets. So I think if you take those two factors into account and add in the fact that every click on a website generates two or three more, then you have to shrink it down again.

[Ruth Alexander]

In fact, another study, which has more credence, he says, estimated that pornography accounts for more like 4 or 5% of all web traffic. So are there any other reliable statistics which can give us more of an idea about how big Internet porn is? Mark Ward says there are some.

[Mark Ward]

There was a different study done in 2010, and that's quite a long time ago in Internet years. The other study looked at the top million websites and visitor numbers to them, and they found that about 4% of those top million sites were dedicated to pornography.

[Ruth Alexander]

What about the number of people just searching, putting porn or sex into an Internet search engine?

[Mark Ward]

That's about 14% of all the searches done on those search engines every day are for sexual subjects of one sort or another. And I don't think it was ever broken down into whether that's the porn end of it or the sort of sexual health end, but it's about that figure. You know, the interesting thing is there are studies starting to be done of porn, But I think a lot of the academics who study this say it's really too soon to say, given time, and they will show for good or ill what the effects might be.

[Ruth Alexander]

More detailed studies which do define the terms.

[Mark Ward]

Yeah, they tend to be academic and they tend to declare their interests. And you get a much better sense of why they've done it and what they've done than you would with all of these zombie statistics that just get bandied around by anybody and everybody.

[Ruth Alexander]

We're very familiar with the problem of zombie statistics on more or less. In fact, we even like to think we coined the term, gave a name to the beast. Our thanks to BBC technology correspondent Mark Ward, valiant slayer of zombie statistics, which no doubt will still refuse to die. One of American football's top players, Aaron Hernandez, has been charged with murder. Hernandez, who played for the New England Patriots before his arrest, denies the charges. It's shocked the sports world, and Hernandez is by no means the first American football player to be arrested, leading many commentators to ask, does American football have a crime problem? One sports writer, Brent Schrotenberg from USA Today, has decided to investigate. And along with a colleague, he's compiled an arrests database for the National Football League. It lists about 660 arrests since the year 2000. Now, that sounds like a big number, but put it in context, Brent says, and it's actually not.

[Brent Schrotenberg]

They tell us that the NFL arrest rate for active players is a lot less than the US General population arrest rate. Strikingly so. The arrest rate for active NFL players is about one out of every 47 to 50 players in the NFL gets arrested each year. And in the general population, the arrest rate is actually double that. It's about one out of every 25 people. And so that's kind of a surprise right there that you hear about NFL players getting arrested about once a week, it seems. And it turns out that they're still better behaved than regular society. And a lot of that has to do with the media attention. When an NFL player gets arrested, it's going to get in the news.

[Ruth Alexander]

And when you break down the numbers for the US population even further, football players look even better behaved. Ben Carter's been crunching the numbers.

[Ben Carter]

Yes, Ruth, American football players aren't representative of the general population because they're all men. So I've looked at the Bureau of Justice's figures which show the arrest rate for US men is actually 1 in 15, much higher than the arrest rate of American football players, which is 1 in 47. Now, American football players are also all young men. So I looked at the arrest rate for 21 to 34 year old men in the general population and got a bigger surprise. The arrest rate is about one in eight. And for people of Aaron Hernandez's age, he's 23. The number is one in six.

[Ruth Alexander]

Now we should say it's not that one in six 23 year old men in the US get arrested every year. These figures are for arrests, not people. And some people are arrested time and time again. Prolific burglars, for example, or Cincinnati Bengals cornerback Adam Pakman Jones, who's been either arrested, charged, or cited eight times since 2005. Repeat offenders bump up the arrest figures.

[Ben Carter]

That's true, but in general terms, there are about six times as many arrests among the general population of young American men as there are among football players. And even US Women are more likely to be arrested than players, albeit by a small margin. Their arrest rate is 1 in 46 compared to 1 in 47 for the players.

[Ruth Alexander]

Thanks, Ben. And we couldn't let Brent Schrotenberg go until he told us what these players are being arrested for. Obviously, most of the arrests won't relate to serious crimes, as in Aaron Hernandez's

[Brent Schrotenberg]

case, by far the most common type of arrest for an NFL player is drunken driving. It makes up about 30% of all active player arrests during a year. Now There are about 60 NFL player arrests per year, so a little bit more than one per week. And of those 60 in the neighborhood of 20 per year are DUIs or drunken driving arrests. The drunken driving arrest rate for NFL players is about half that of their age bracket in the U.S. general population for males in their 20s. So again, the arrest rate is higher for the general population in drunken driving than NFL players even there.

[Ruth Alexander]

Brent Schrotenborg from USA Today if there are any numbers you'd like us to take a closer look at, do email us. Our address is more or lessbc.co.uk we love hearing from you. Roger sent us a missive recently commenting on our programme which looked at how rapidly Africa's population is expected to grow this century. It won't reach 4 billion in 2050 as predicted, he says. Surely economic and social factors will have changed to avert this long before that. And Yvonne's emailed to say she doesn't believe census figures which revealed the UK and the US are missing thousands of elderly people can be accurate. When she tried to fill the form in for her 90 year old father, she found it difficult to be accurate about his employment history and wasn't sure the nursing home ever even passed the form on. Well Yvonne, you're right that census figures themselves are estimations, but the elderly are the group who most reliably fill in and send off their forms. So the figures are thought to be pretty close to accurate. Let's hope your father was counted too. And you can also sign up for a free download of our program at our website bbcworldservice.com More or less

[Ben Carter]

There

[BBC Announcer]

are dozens of different podcasts now available from the BBC, including news, documentaries, science, business, arts and sport. The details of them all go to bbcworldservice.com podcasts.