Wow I really hope I did not come off as defending sexual harassers in any way. I think many might have thought this because of my last paragraph which seems troll-like. I can assure you I don't defend that behavior and am not trolling.
I am simply saying that your statistical analysis was flawed, because perpetrators can have multiple victims, and you had suggested, statistically, that one man fired means one woman harassed in one instance.
Also given low rates of sexual assault reporting, it is likely that by the time a report is made, and even more-so investigated, that the perp has been involved in multiple instances of sexual harassment.