In truth, no. I've heard this stated by various people, among them Tim Pritlove, who said it in one of his podcasts (I think it was NSFW, but I'm not sure). Nobody I've heard it say has pointed to evidence. It is one of those things that appears plausible to a lot of people from personal experience. This of course means, that it might be wrong.

There seem to be some people trying to collect data about the correlation between availability of pornography and the level of some crimes, but I'm still waiting for something conclusive.

Not that it's easy. Quantitative sociology is, due to the interconnectedness of the system in question, a mess.