It is. The extent to which it is, is still in dispute, with the Mayo Clinic suggesting somewhere between 30|%-80%. Part of the issue with researching pedophilia/child molestation is that the majority of subjects we have to question are/or have been in the criminal system. So selection bias is inevitable in the studies.
Here's an interesting paper, that has links to a fair amount of the research, concerning the same issue when it was dealt with in the U.S., while Ashcroft was running Justice:
http://scholarship.shu.edu/cgi/viewcontent.cgi?article=1040&context=student_scholarship
Basically, the arguments come down to 1) child pornography doesn't inhibit and may exacerbate a pedophile's need to molest and 2) child pornography(drawn and real) is used by pedophiles to draw in children for predation.
You may not agree with those, of course, but there is some available science to suggest the first, and the second is undeniable, as we have multiple records of child molestors using child pornography to attract children.
Bans like these are tricky, because you get over eager enforcement causing problems, and creatives who want to push the envelope, so court cases tend to come hot and heavy for a few years, before they settle down and people accept the new standard. It's not like Japan doesn't have a history of this, and dealing with evolving standards of obscenity. They already have laws about mosaics and what you can and can't show legally. This would be an extension of those existing laws.