It is actually illegal to process or count early ballots in PA before election day, so at best estimates based on party registration of those returning ballots.
And ABC appears to be comparing performance relative to the unique 2020 COVID election in which Dems had a huge lead. And they still do this year:
View attachment 110371
Edit: sidenote, these early ballots can't even start to be processed until election day, as mandated by Pennsylvania MAGA congress. So the above votes will once again be counted after in-person ballots are processed on election day - in many districts may take a day or two more to open/process all of the mail ballots.
And once again, even if Harris wins by a landslide, with this voting imbalance Trump will still appear to be winning election night (based on in-person voting only).
Edit 2: used better table that includes # requested ballots / return rates. And source/link:
» 2024 General Election Early Vote – Pennsylvania UF Election Lab
election.lab.ufl.edu
Can one assume that different parties have the same ratio of 'early voting' to 'voting in person'? Or is that known from 2020?
In any case, the whole concept of reporting anything about votes already cast, before everyone has voted, seems dodgy to me.
I mean, I don't see any logic in banning people from counting early votes early, but I would entirely agree that nothing should be reported about the outcome of that count (and it's your system so up to you, but even reporting 'party registration' of returned votes seems to me to be a dubious thing to do).