Given the America-centric nature of things...

13%. 0.1%. ??%. Black. Trans. Pedophile.

Even without considering "co-variables", you're already at a tiny, TINY percent of the population before you hit the ??%, since you're starting at 0.013% of the population before applying criminality to it.

Without "trans" reducing the numbers to near-zero in the first place, there's likely a gang initiation element correlating in ephebophilia and black populations, that is, what the other posters theorized on middle and high school girls being involved in gangs, willingly or unwillingly, in sexual manners. But I don't know if that would also map onto pedophelia. Probably some correlation, but the strength of it is unknown. It is possible, but it seems unlikely any research would ever be done on these matters because of how politically charged they are. But the "trans" element could act contrary to that variable, since I'm not sure if trans people are welcomed into gangs at the same rate.