Internet Reduces Rape?

Over at is this short news item:

Commentary and Opinion: Porn a catalyst of sexual violence? No

No one would say the current level of violence against women is acceptable. But the enormous progress in recent years is one of the most gratifying successes imaginable. How can it be explained? Perhaps the most surprising and controversial account comes from Clemson University economist Todd Kendall, who suggests that adult fare on the Internet may essentially inoculate against sexual assaults. In a paper presented at Stanford Law School last year, he reported that, after adjusting for other differences, states where Internet access expanded the fastest saw rape decline the most. (11/04/07)


Deceptively short this news report actually serves to hide some worrying ignorances of the world. Firstly even if the report showed that Internet access was causally linked to lower rape rates, the accessibility of sexually violent images on the web cannot mean we can safely presume this is a positive relationship. Have we won the battle if we can say “well now its 1,000 men watching another man (or men) rape a woman?”, is that actually a better position? Additionally Kendall’s position is one based on a “snap-shot” of time rather than, as one must, looking at long-term harm reduction. His findings mostly relate to a lower rate of rapes by 15-19 year old men being reported (and there are all sorts of problems with assuming reporting or arrests (as he does) equate to the “real” rate of rape) yet he doesn’t seek a longitudinal view which asks what happens when these 15-19 year olds at 20, 22 or 25.

Luckily for us Todd Kendall’s paper is available to all here. Kendall begins by making the stunning claim that he can provide empirical evidence that:

I find that internet

access appears to be a substitute for rape. Specifically, the results suggest that a 10

percentage point increase in internet access is associated with a decline in reported rape

victimization of around 7.3%.

(Kendall 2006 p.3)

He bases this claim on six pieces of evidence, each of which I am going to question in turn:

1. States that adopted the Internet quickly saw larger declines in rape incidence than other states (while no similar effect is evident for homicide).

a. Kendall artificially splits the States into quick and slow adopters rather than seeking a continuum-based analysis which would be the only way to show a causal relationship (if states are ranked on there adoption of Internet access and and the pattern there matches the pattern of reduction in rape reporting then there would be better grounds for further analysis and theorising then the rather clunky separation into two generalised groups – sadly whilst Kendall provides the States ranked by Internet adoption he doesn’t give us the data for States by decline in rape arrests).

b. Additionally his data on internet access asks only whether someone in the household accesses the internet and he presumes this is a male. Of course it would be harder to explain the thesis “internet = less rape” is those internet users were women.

c. His “significant reduction” is 2.2 rapes per 100,000 population which, whilst wonderful if true, isn’t to my mind a significant reduction which justifies access to violent pornography on the web – indeed Kendall studiously avoids talking about web pornography and whether this displacement is presumed to be men watching internet porn.

2. this effect is most concentrated among states with a high ratio of male to female population, suggesting that men are substituting pornography for rape most when potential mates are in low supply. Kendall, here, assumes that rape is about a lack sexual mate in terms of supply. There are so many challenges to this not least that criminality cannot be explained this way (taking his assumption as true) unless you assume a biological-evolutionary model, when there are legal ways to solve the problem and therefore he should also have examined whether prostitution rose or fell in the same period (he claims he has but there are significant problems with this – mostly that he’s talking about arrests rather than crimes committed). Additionally rape is not a crime of economic supply (a woman’s right to say no not being equitable with an i-pod or a pair of trainers) and rape is a crime of power which happens even when there is a plentiful supply of female partners. Plus Kendall assumes that men’s sexual needs “need” fulfilling in some life-threatening sense. I’ve not yet seen a man die from a lack of intercourse.

3. using regression analysis with fixed state and year effects to show a negative

correlation between internet access and rape, even controlling for a wide variety of other

factors. But Kendall doesn’t control for, for example, the presence of profilic and repeat offenders in State, STD rates (he only looks at HIV infection rates), campaigns against sexual violence, high profile trials for sexual violence etc. Nor does he examine any data on Police behaviours and whether the drop in arrests is due to (at the most specious level) Police also using the internet, but more seriously there are ebbs and flows in policing of different crimes due to the priorities set. And there are ebbs and flows in crime reporting and rape is a crime much less likely to be reported after large scale coverage of, say, a presumed false accusation. Kendall doesn’t comment on whether any of the states had a high profile rape case where the claimant was presumed to be lying or presented as such in the media.

4. similar analysis evidences neither a statistical nor economically significant effect of internet usage on any other violent or property crime for which reliable state-level data is available. I am not sure why we would presume internet usage has impact on any form of crime and Kendall never explains this assumption.

5. using data on arrests I find a significant negative effect of internet access on rape arrest rates among men ages 15-19 – a group for whom pornography was most restricted before the internet Now here’s the doozy – Kendall is actually making a case for less restrictive controls on access to pornography for the under 20s. But he uses a rather dodgy presumption of causation (between internet access and rape stats) to achieve that. But this claim is worrying – it’s essentially saying give 15-19 year old access to internet porn (which we know is usually more violent and abusive and less controlled in terms of how material is produced than print porn) and they reduce their sexual offending behaviours. But Kendall’s research is short-term – what happens when the 15-19 year olds grow up? I’d question what new patterns of escalation would appear if 15-19 year olds, who tend towards relatively petty offences which then grow in magnitude, postpone their criminality to young adulthood and postpone it whilst they access some of the most violence and abusive porn available through any medium?

6. Finally, I also provide evidence on the correlations between internet adoption and several other measures of sexuality, including teen birth rates, prostitution arrests, marriage and divorce rates, and HIV transmission. The results generally imply that internet usage has had significant effects on sexual behavior more generally, and thus they lend credibility to the claim that the internet may impact sexual assault to the degree. Sadly in this paper Kendall doesn’t provide that data except to show a general decline in reported rape associated with every fact he looked at – which sure leads us to questions whether this actually shows us a general decline in rape reporting connected to some other factor rather than rise in internet use (or beer drunk or any of the other factors he associates).

In the end what Kendall shows us is that rape reporting declined between two points in history – the reasons why are an entirely different matter.