They could be, but it could also be that negative comments about Biden get flagged by users far more than negative comments about Trump do. That is, if the average Trump supporter just ignores or laughs at or replies to negative comments about Trump, and the average Biden supporter flags them as abusive, any automated system will learn that.

To be clear I have no idea what the explanation is. But it doesn't <i>have</i> to be due to explicit actions by Google employees. It can also be due to different approaches by the users themselves.