The Sex Myth: Why Everything We're Told Is Wrong (11 page)

The internet needs to be a common carrier. You do not expect the Post Office to open and vet the contents of your letters. A telephone conversation should not be cut off because
an automated program thinks it picked up something offensive. Treating the internet as an entity completely different to these is a fallacy.

But after such an inevitable lock-step to this point, and with so little in the way of vocal criticism from the entire country’s mainstream media, it’s hard for me to be surprised.
This has been coming for a long time. I’m surprised at the people who are surprised. And with the state of the economy, it’s equally unsurprising that the current government is trying
to make porn and immigration stay above the fold. Bread and circuses, kids. Bread and circuses. Only, wasn’t there a time when these circuses used to be more entertaining?

Readers of the sexualisation reviews should be asking questions throughout. The first is, does the literature show a demonstrable relationship between the availability of
sexually suggestive items and violence towards girls and women? Papadopoulos in the Labour consultation says yes: ‘The evidence gathered in the review suggests a clear link between
consumption of sexualised images, a tendency to
view women as objects and the acceptance of aggressive attitudes and behaviour as the norm.’

One saying you hear over and over in the research world is ‘correlation is not causation’. It’s a good rule to remember, particularly when looking at human health and
behaviours. As with the constellations we give names and stories to in the night sky, the time and place in which we observe something influences whether we think they’re related. Taking a
different perspective may show them not to be connected at all.

What’s the difference between a cause and a correlation? Put simply, when you say one action causes a result, you are saying there is a direct line between that action and the result.
Cause implies that an action always results in a predictable reaction.

Correlation, on the other hand, means that the action and the outcome occur, but may not be related. The amount of reality television programming has risen dramatically since 2000; so have the
application fees for marriage visas in the UK. Neither one caused the other. Happening at the same time does not imply any relationship between them.

Showing causation requires a lot more than that. A reputable report should go much deeper. One way to do so is by demonstrating the way in which an observed outcome might have happened, or the
mechanism of action. Another way is by eliminating possible other factors – what researchers call
confounders –
from consideration.

For example, in the original study that showed a connection between smoking and lung cancer, there was an effort made to eliminate confounders. Risk of lung cancer could conceivably be
influenced by occupation and social background, so the study focused on a population that was (at the time) relatively homogeneous: doctors. A well-designed study should always consider whether
confounders exist and what they might be, so it’s important to see whether this has been done.

To prove the link, there should be ways for other investigators to replicate the study and test for the same result. So, while the initial study of doctors and smoking demonstrated a clear link
to lung cancer, this has been backed up by many research groups looking at all kinds of populations around the world. The result is the same every where: smoking causes lung cancer.

Correlations are not useless as a first step in investigating whether there
could be
a causal role. But they are the first step only. Keeping all this in mind, the
saying should read, ‘Correlation is not necessarily causation, but it might be – proceed with caution and look at the problem deeper.’ It requires looking past a small-scale
suggestion of a connection. Eliminate confounders, see if the results hold for other situations, and evaluate the evidence. If the results can’t be replicated, well, the results may not have
been reliable.

So, while correlation seems to suggest a link between porn and negative effects, looking more closely at the results, nagging doubts about the claims of these reports creeps in. In particular,
there are sweeping conclusions made from preliminary evidence showing mixed outcomes.

One of the papers cited in the Papadopoulos report is a 2000 publication by Neil Malamuth, a prolific researcher into sexual aggression. The paper considers whether there is a causal link
between adults who view pornography and sexual aggression.

Now, the difference between ‘sexualising’ images aimed at children and actual pornography created for adults is apparent, and using such a study as a reference in a discussion of
sexualisation is suspect. When looking at how well a review makes its arguments, it’s essential to ask whether the evidence being presented is not only accurate, but relevant.

But, as mentioned before, it would be almost impossible to conduct a study on the effects of sexualising images on children. You can’t really do it very easily, if at all.

Judging the effects of sexualising images on children by looking at the research into pornography and adults is hard to do. Some would say impossible. However, many people believe the difference
between suggestive images and pornography to be a difference of degree rather than of type. With little in the way of evidence to connect the two in this way, let’s go ahead and accept that
proposition and consider the ‘worst-case scenario’: hardcore pornography and its effects on levels of violence.

In a comprehensive review of data produced by the Malamuth research group, as well as knowledge from other studies, this simple question was raised: ‘Hardcore pornography and sexual
aggression:
are there reliable effects and can we understand them?’
⁵⁰

While there was some overlap in people who viewed violent pornography and violent beliefs directed at women, the conclusion was far more muted: ‘We suggest that the way relatively
aggressive men interpret and react to the same pornography may differ from that of nonaggressive men.’ In other words, the pump is already primed in some people, and exposure enhances that
tendency. But, for nonaggressive men, the very same imagery did not seem to incite negative thoughts.

When the Papadopoulos report cites Malamuth’s work, the relevant context that it is a study of adults is omitted. Details from the research have been presented to imply a connection, and
data showing otherwise – as well as the thoughtful, grey-area conclusion – are also ignored.

Porn and violence are not as intimately connected as is assumed by the Home Office reports. So, if porn does not have this damaging effect, how could it possibly make sense that other, less
sexual and less violent things would?

One big question every parent is concerned about is whether sexually suggestive products lead to young people having sex at an earlier age. We can see what is on the market,
but are these items actually ‘sexualising’ anyone?

One hallmark of drawing constellations is relying on anecdotes, or one-off stories. There are loads of anecdotes that claim kids are having sex earlier, dirtier, and with more people. Canadian
filmmaker Sharlene Azam’s documentary
Oral Sex Is the New Goodnight Kiss
features girls as young as eleven years old talking about sex parties. What shocked the filmmaker the most was
that ‘[t]he girls are almost always from good homes, but their parents are completely unaware.’
⁵¹
The clips certainly attracted a lot of
attention, and were held up in some reviews as representative not simply of the participants, but of an entire generation. But a tiny group of middle-class students is never going to be
representative of the broader picture. And it’s just as likely – as is often the case when a group of puberty-age kids get together – that being interviewed in a group could lead
to some exaggeration of what’s going on. It’s
not a sociological phenomenon. It’s just an anecdote.

Giving a single example and drawing a conclusion does not make the assertion false – but it also does not make a universal truth. Anecdotes are fine when you’re telling a story, or
illustrating something you have more evidence for. It’s less acceptable when making a point about life in general.

For instance, my grandfather smoked a pack of cigarettes a day and lived to a ripe old age. That is true. But does the example cancel out the barrage of medical evidence proving a link between
habitual smoking and early death? No, it doesn’t. If I gathered up a book’s worth of old people who smoked and died from some other cause, would that be enough? It wouldn’t. Just
because some people do not die from smoking does not mean smoking doesn’t kill.

Anecdotes like the story of my grandfather can illustrate a point. However, when they become the point – or worse, when unrelated stories are strung together to
create
a point
– what we have is a constellation in the making. Great for pushing an agenda. Not so great for getting to the truth.

And while the Home Office reports are silent on the matter of the real numbers in teenage sex statistics, a number of other wide-reaching studies show the large-scale trends that are
emerging.

In the US, where most of the ‘raunch culture’ reporting occurs, the proportion of girls having sex before the age of fifteen has hardly changed from the late 1970s to today, and has
decreased among boys.
⁵²
If access to sexualising media is correlated with an effect on age of first sexual intercourse, there are not many data to support
it.

The proportion of people who have had intercourse by the age of fifteen has risen among both sexes in Britain, but remains comfortably under 15 per cent. In a study from 1980 onward, the
proportion of women losing their virginity by the age of sixteen showed a rise until the 1990s, when the numbers stabilised.
⁵³
Meanwhile, in the US, the
proportion of high-school students who have had sex decreased steadily in both boys and girls from 1991 to 2007.
⁵⁴
The author of the UK studies, Kaye
Wellings, was quoted at a press conference as saying, ‘The selection of public health messages needs to be guided by epidemiological evidence rather than by myths and moral
stances.’
⁵⁵
A statement those producing guidance for the government would do well to remember.

The Home Office reports from both the current and previous governments present a one-sided view that lacks the rigour we should expect in policy
recommendations. As a starting point for further investigation, they might have limited merit – provided they weren’t presented as an authoritative summary of all known information on
the topic.

Unfortunately, that is exactly how they were presented. And the methodological faults of these particular reviews might not be so clear had the Scottish Executive not commissioned a similar
report that was released in the very same month as Labour’s sexualisation report.

The Scottish Executive’s
External Research on Sexualised Goods Aimed at Children
was commissioned for the same reasons as the other reports: concern by the government that there
were increasing levels of sexual imagery in products aimed at children, and that there might be harm caused as a result.

The study takes several approaches. First, just like in the Home Office reviews, it begins with a literature review. The key earlier studies, from the American Psychological Association and the
Australian Senate, are covered along with other relevant research. As well as highlighting the main summaries in these publications it considers their drawbacks. Particular criticism is reserved
for the one-dimensional nature of earlier reports. ‘[T]here is no indication [in the APA report] that the media might contain any positive images about human relationships, or that children
might critically evaluate what they see.’

The Scottish review notes that ‘[s]uch accounts often present the sexualisation of children as a relatively recent development, but it is by no means a new issue.’ Historical
phenomena such as changing ages of consent to marry, and the evidence for prostitutes under the age of ten in early Victorian London – and the attendant concerns about childhood sexualisation
– are noted. ‘While the public visibility of the issue, and the terms in which it is defined, may have changed, sexualised representations of children cannot be seen merely as a
consequence of contemporary consumerism.’

Three main criticisms are made of previous reviews (which could
also have been applied to the recent UK reports, though they were not yet published at the time the Scottish
review was written).

The first is ‘a lack of consistency and clarity about the meaning of sexualisation’ in other studies. The actual areas of concern are seldom defined, relying instead on assumptions
to frame the debate, with many researchers using the same words to mean subtly different things.

Another problem is that ‘much of the research suffers from methodological limitations that are characteristic of media effects research.’ In general, media effects studies are
conducted on small groups, which can bias the results but also be inapplicable to larger populations. Any laboratory study has to take that into account. It’s likely that the attention
subjects give things in the lab is pretty different from the attention they give in their natural surroundings – especially if they know they’re going to be quizzed on it later.