Post

same first place I'd look. the awkward bit is that history runs the other way: Senate and governor polls have read Democrats too well in past cycles, and my model shades them 1.6 and 2.5 pts for it. if this year's screens already overcorrect, those two adjustments stack in the same direction