Furthermore, the correlation with Trump approval is totally fishy. These numbers don't rule out the possibility that men harass women at roughly the same rates regardless, and self-report differently due to their awareness of the issue.
Self-reporting bias would certainly have some effect, but not as much as brute similarity of worldviews regarding women, gender, sex, workplace etc. Like attracts like. Nothing fishy about that.
A lot of things are possible. Maybe Trump supporters are more likely to harass or maybe they surprisingly aren't; the most dangerously effective misogynist is one that adopts the language of political correctness but not the beliefs. This requires further study.