Pornography and sexual violence: Why rates of rape in a country tell us little

There are at least four forms of evidence we can use to test if pornography increases the likelihood of sexual violence:

  1. Correlational studies, testing whether there are statistical associations among porn users and non-users between the use of porn and sexually violent attitudes or behaviours;
  2. Experimental studies, in which people are shown porn under laboratory conditions, with data on their sexually violent attitudes or behaviours before and then after exposure;
  3. Longitudinal studies, which gather data over time among porn users and non-users and see if porn use predicts later sexually violent attitudes or behaviours (while controlling for other factors that might shape these);
  4. Ecological studies, testing for statistical associations between the level of porn use in a particular country or context and the level of sexual violence.

Studies based on the first three forms of evidence consistently show that porn increases the likelihood of sexually violent attitudes and behaviours.

The fourth type of evidence is the weakest and most problematic of the four for testing whether porn has an impact on sexual violence. This is because, above all, rates of sexual violence are shaped by *multiple factors*, and porn use is only one of these. These multiple factors include levels of gender inequality, levels of other forms of violence, the presence of war or civil conflict, strong laws and justice responses, poverty and social disadvantage, and so on. Keep in mind that these factors, depending on which way they go, may lead sexual violence to increase *or* to decrease.

Let’s say, hypothetically, that in CountryX, rates of pornography use have increased *and* rates of sexual violence have increased. I *could not* use this to claim that pornography use had led to increased rates of sexual violence. Because, as I’ve said above, rates of sexual violence are shaped by multiple factors, not just porn use.

Equally, in CountryY, if rates of pornography use have increased or remained steady while rates of sexual violence have *decreased*, you cannot use this to say that pornography use is unrelated to sexual violence.

To give a parallel example, it is well established that smoking causes cancer. In a particular country, rates of cancer may go down, while rates of smoking go up. That does not mean that smoking does not cause cancer. It just means that rates of cancer are shaped by rates of smoking *but also* by other factors.

So, debates over ecological evidence – that is, over levels of porn use and their relationships or not with levels of sexual violence in any particular context – are not very useful, whether for those arguing that porn does lead to sexual violence or those arguing that it does not.

The only way that ecological evidence would be useful would be if you could examine a country or context where porn use increased but *all other* factors which might shape sexual violence remained steady. But such places do not exist, as social change is a constant.

So, debates over ecological evidence are largely a waste of time, while debates over the other forms of evidence described are far more useful.

See here for a summary of the research on pornography's effects, with links to published scholarship. See here for full text journal articles on pornography's effects.