I think I've watched the loop 50 times.

An increase in rape incarceration rates isn't neccesarily indicative of an increase in the incidence of rape itself. That could have occured due to a variety of factors. Some examples include: more zealous prosection of rape crimes, more immediately forthcoming victims, and jury members more willing to pass guilty verdicts.

Furthermore, looking at the graph, the reported occurence of rape continued to decline even after the methodology was changed. Therefore, it would be possible to infer that rates still fell during that time period-- barring the possibility that the change in procedure somehow caused the sampling data to become less and less reflective of the population's actual data over time, which is unlikely.

I'm not exactly saying you're wrong when you say pornography demeans and objectifies women, and causes the men who view it to become more disposed to comitting rape. I'm just saying that this data can in fact provide evidence to the contrary.