Every single poll I've seen that asked that question has shown Harris running well ahead of the party split among early voters. Marist found that in NC, GA, and AZ and Siena and I think ABC found that in their national samples. I'm extremely skeptical of the margins they're finding (which suggest a blowout on the national level that would not remotely match the spending patterns in either the presidential or lower level races) but it this point I'm also skeptical of the idea that it's pure noise with no signal. My prior is that it's about 80% noise, 20% signal but even that would be enough to tip the election to Harris with room to spare.
Thanks. Those margins aren't plausible but when every single serious poll finds the same result there has to be some explanation for it, even if it's an explanation for why it's all noise.
Every single poll I've seen that asked that question has shown Harris running well ahead of the party split among early voters. Marist found that in NC, GA, and AZ and Siena and I think ABC found that in their national samples. I'm extremely skeptical of the margins they're finding (which suggest a blowout on the national level that would not remotely match the spending patterns in either the presidential or lower level races) but it this point I'm also skeptical of the idea that it's pure noise with no signal. My prior is that it's about 80% noise, 20% signal but even that would be enough to tip the election to Harris with room to spare.
I believe it was something like this:
Marist (AZ): H 55-44
Marist (NC): H 55-43
Marist (GA): H 54-45
ABC/IPSOS (National): H 62-33 (8% of sample)
CNN (National): H 61-36 (20% of sample)
NYT/Sienna (National): H 59-40 (9% of sample)
Thanks. Those margins aren't plausible but when every single serious poll finds the same result there has to be some explanation for it, even if it's an explanation for why it's all noise.