@Nick It seems to be someone trying to understand a comment they have recorded in their own lecture notes. But in any case something looks to be missing or perhaps misplaced; I see no obvious sense in which interchanging X and Y in a regression (and all while keeping the same parameters!) has anything in particular to do with causality (or indeed with casuality.)
But perhaps I am missing something. I wonder if a modified version of what they wrote might work for, ... well, something.
