As my colleague Dr. Malamuth has suggested, there are no perfect studies. This is an area that many scientists are concerned about and have approached in different ways. There is no perfect study.
What I've done in the last 10 minutes is to review multiple lines of evidence. We pointed out that correlational studies could mean either that people who view pornography are aggressive in a contributory way or that people who are aggressive like to look at pornography, and the confluence model study, which Dr. Malamuth has talked about, is one that we've actually replicated.
That is, people with, broadly speaking, anti-social personality traits who use more pornography report that they engage in more anti-woman aggression. The issue there it that, of course, that remains a correlational study and we don't know which way it's going.
As I pointed out during my remarks, when we actually add in a simple measure of men's sex drive, it knocks out the contribution of pornography in the context of anti-social personality traits, etc., so the science is far from settled.
What I can say is that we've had an incredible natural experiment that none of us asked for, and that involves the onset of essentially unlimited access by every man, woman, and child in Canada with, for example, their anti-social or pro-social or neutral personality characteristics, since about 1995. As I remarked, we look at a number of possible markers of what's going on. These are population-level data and they cannot tell us what's happening to any individual, but they can tell us almost on a policy basis what's going on in the context of unlimited access by Canadians and Americans.
From what we've reviewed, there's been a substantial decline in rates of sexual assault. This is not of sexual assault reported to the police; this is from victimization surveys. This is active surveillance, which the U.S. has continued throughout the era of Internet pornography. There has been no change in rates of sexual assault in Canada. We find that adolescents in Canada are not having sex any more often or at younger ages or with more partners. We find that the rate of divorce per thousand in Canada and the U.S. has continued to decline. And we actually find, in the context of the nationally representative U.S. data that our lab has analyzed, that egalitarian attitudes seem to co-vary or correlate with pornography use.
The issue, broadly speaking, is that the evidence is quite mixed. Every data point that I've talked about with you is from a published study, so that's where we stand.