Pornography and rape

The Criminal Justice & Immigration Act has been passed, among other things banning the possession in the UK of extreme pornography. Despite its origins, most justifications for this measure, even in Parliament have been based on simple distaste for pornography, something that people are of course entitled to feel.

The provisions were put into the Bill following the death of Jane Longhurst at the hands of a man, Graham Coutts, who was obsessed with images of women being strangled. Longhurst was asphyxiated. Her mother collected more than 50,000 signatures in support of the measure and her campaign was taken up by Martin Salter MP. It's plain that Mrs Longhurst believed there to have been a causal link between Coutts' collection of such images and his murder of her daughter. There seems to be no strong evidence either way in this particular case, but Coutts was concerned, before the murder, about his obsession and told his doctors he feared he might commit a crime.

A campaign called Backlash has failed to prevent this law being put into effect. Backlash fears it will be used to criminalise people who look at depictions of sado-masochistic sex, and this is a large minority of people. The majority probably view their activities with distaste, but then that probably remains true of homosexual acts, especially male ones, and it was certainly true when homosexuality was illegal. Despite the odd voice raised in support of Backlash, and the existence of an online petition (with very few signatures so far) calling for these measures to be repealed, this is not becoming a popular issue. It should. I fully support Backlash's campaign.

This is an interference in the private sexual lives of private citizens, and is deplorable simply for that reason. But that's not the main reason.

The interpretation of this act can be expected to creep, as was pointed out by:

Edward Garnier, an MP and part-time judge, who questioned the clause when it was debated in the Commons.

"My primary concern is the vagueness of the offence," says Mr Garnier. "It was very subjective and it would not be clear to me how anybody would know if an offence had been committed."

But that's not the main reason either.

It's a thought crime, the criminalisation of people who have and will never harm anyone else, as was:

... noted by Lord Wallace of Tankerness during last week's debate in the House of Lords.

"If no sexual offence is being committed it seems very odd indeed that there should be an offence for having an image of something which was not an offence," he said.

But again, that's not the main reason.

The main reason is that, on the balance of available evidence, this Act of Parliament will kill women.

The relationship between pornography and sexual crime is hard to track, and about the only really serious attempt to study it based on actual hard, empirical statistical evidence was made a couple of years ago by Dr Todd D. Kendall, now an Assistant Professor of Economics at Clemson University. This is the abstract of his most recent, 2007 draft of his paper Pornography, Rape, and the Internet (pdf):

The arrival of the internet caused a large decline in both the pecuniary and non-pecuniary costs of accessing pornography. Using state-level panel data from 1998-2003, I find that the arrival of the internet was associated with a reduction in rape incidence. While the internet is obviously used for many purposes other than pornography, it is notable that growth in internet usage had no apparent effect on other crimes. Moreover, when I disaggregate the rape data by offender age, I find that the effect of the internet on rape is concentrated among those for whom the internet-induced fall in the non-pecuniary price of pornography was the largest – men ages 15-19, who typically live with their parents. These results, which suggest that pornography and rape are substitutes, are in contrast with previous laboratory studies, most of which do not allow for potential substitutability between pornography and rape.

When Kendall's paper first became generally known, in 2006, there was some criticism of it. The best informed was this piece, which concludes:

None of this is to say that what Kendall is saying is necessarily wrong, though personally I'm sceptical about the argument that access to porn is a substitute for rape (in Kendall's terms, I'm firmly in the camp of those who believe that rape is about power rather than about lust). It's simply to suggest that Kendall's results and the interpretation he puts on them are extremely questionable, and we should be careful before drawing any real conclusions from them.

I wondered whether Kendall had worked further on this area, so I emailed him recently and, with his permission, copy below his reply:

Peter,

I had not seen those comments, but they do not surprise me because I have heard them before. With respect to the issue of multicollinearity, I am not sure just what econometric model one has in mind by which including additional variables can spuriously increase a coefficient or decrease a standard error; nevertheless, the current draft of the paper available on my website (http://people.clemson.edu/~tkendal) includes lots of robustness along these lines: basically, I show that there is *no* combination of covariates for which the relationship between internet usage and sex crime is positive. True, not all correlations are statistically significant, but one has to be somewhat humble in terms of what one can expect to get out of the limited data available. There are only four years of data and only 51 states (inc. DC). Nevertheless, I think the patterns in the data are suggestive and all point in the same direction. Perhaps better data would show that the true effect is smaller than the one I estimate (that is always a possibility), but the "extreme bounds" analysis in the latest draft of the paper suggests to me that it is highly unlikely that better data would reveal an effect of the opposite sign.

Also if multicollinearity is really causing a spurious finding for rape, why is there no similar result for any other non-sex crime? And why is this supposedly spurious finding only relevant for young men, precisely where my (Posner's, really) theory suggests the effect should be? And so on, following the arguments in the paper. My main method of identification of the model is really based on a comparison between sex crimes (rape and prostitution) and other crimes. In general, I think this method shows the robustness of the results to a lot of omitted variables, functional form assumptions, etc., but again, I fully admit the total amount of data available is small and therefore the results should be taken with some caution.

With respect to some of the arguments related to how the internet could affect reporting, I think we don't really know much about non-reporting of sex crimes so, in theory, I guess anything is possible. Nevertheless, I think a natural prior is to believe that, if anything, internet access is probably associated with more reporting, not less. I make the case for this in the paper.

Finally, these commenters (and many others) seem to want to say that my paper shows that rape is about sex instead of power. As I try to make very clear in the paper, I make no such claim. People who say this are implicitly assuming that pornography is about sex, not power. I am no gender studies theorist, but I don't see any reason why, if rape can be about "power," pornography couldn't also be so. My only claim is that there is some evidence of substitutability between them; I never make any claim about precisely which deep preferences consumption of these goods satisfy.

Dr Kendall's paper is now undergoing peer review, but at the present state of knowledge, it seems we must conclude the following:

• Some sexual crimes might be prompted by pornography
• Some sexual crimes might be prevented by pornography, which acts as a substitute for the crime itself
• The aggregate of these two possibilities is that the availability and, therefore, possession of pornography reduces sexual crime
• This new UK law will criminalise innocent and harmless people because of their consensual sexual preferences

It feels harsh to say this, but the result of Mrs Longhurst's campaign is likely to be that more women suffer the fate of her daughter than would otherwise have been the case, and that's too high a price to pay for the satisfaction of her grief.

On the other hand, I noticed earlier today at the blog A Dirty Martini three cartoons taken from porn magazines that are not considered extreme, Playboy, Hustler and Penthouse, that - staggeringly - almost celebrate rape. "Encore", shouts a raped woman, slumped half-naked against a fence in Penthouse. Playboy gives us: "Well, I'm a consenting adult, and Charley here is a consenting adult, so that makes two out of three". Hustler showed a billboard offering a helpline for women who "have been raped or would like to be raped". These cartoons, in magazines that have become entirely mainstream, could only contribute to a normalisation of rape.

It is, however, much easier to criminalise the consensual activities of law-abiding people than to tackle creeping cultural coarseness. And what is the British government for, if not to enact knee-jerk, populist and repressive legislation?


Go on, sign the petition. I'm about to do so now.