There is also an argument that some portion are "older", but again, they may have done it decades ago. There is no "when".
Basically, he's trying to play up the number of "predators".
Anyway, here's another one "they know someone who has done the same thing". This revelation apparently completely changes things... Except it doesn't...
As the data here (different data set) shows, 50% of teens in the late 80s, and 40% in 2013 had sex with each other, therefore the probability that someone "knows" someone who has done this approaches 100%.
If there's a correlation between that, and "perversity", or whatever, it might be they're more willing to talk about their private sex lives with each other. Who knows.
Still, it's easy to think of non-conspiratorial explanations.
Saying they're more likely to be "perverts" doesn't change this and isn't relevant.