I’ve just sent out an email to paying subscribers about 3 things you should keep in mind when it comes to early general election polling. But I wanted to make a summary of the information available to everyone because I think this is a very important issue. Here it is:
The correlation between polls now and general election outcomes as historically been 0.
State-level polling is probably worse than not predictive and (as I joked online) might even make you dumber?
Polarization and high interest in the election might be making early polling more reliable, but we’ll have to wait to see.
If you want to read more, pleas consider becoming a paid subscriber to my newsletter. You can read the first bits of the mentioned piece and sign up by following this link.