It's also worth noting that the poll has Kamala winning Independents by a sizeable margin: 48% to 44% in the head to head.
However, Kamala is still losing in the poll overall because 1. the sample contains slightly more Republicans than Democrats and 2. A slightly larger percentage of self-identified Democrats are planning to vote for non…
It's also worth noting that the poll has Kamala winning Independents by a sizeable margin: 48% to 44% in the head to head.
However, Kamala is still losing in the poll overall because 1. the sample contains slightly more Republicans than Democrats and 2. A slightly larger percentage of self-identified Democrats are planning to vote for non-Kamala candidates than Republicans are for non-Trump candidates. I'm a bit skeptical that either of these data points will be reflective of the final electorate.
They got a sample with a fraction of a percent more Rs than Ds in the raw data. They then did a bunch of weighting, which seems to involve things like past nonresponse bias, and end up weighting to an R+2 electorate.
I'm perfectly willing to believe that if the electorate is R+2, basically 2014 redux, then we lose. I doubt that's what the electorate will be.
And to be clear, I think they're overfitting. If you get more Rs than Ds, I'm not sure why you'd assume that the nonresponse bias is similar to past polls in which you had trouble reaching Rs.
If Democrats are just as or more excited than Republicans about voting (see also CBS/YouGov battleground state polls) then we probably don't get an R+2 electorate. And if Harris wins independents, even by four points, then she likely wins.
It's also worth noting that the poll has Kamala winning Independents by a sizeable margin: 48% to 44% in the head to head.
However, Kamala is still losing in the poll overall because 1. the sample contains slightly more Republicans than Democrats and 2. A slightly larger percentage of self-identified Democrats are planning to vote for non-Kamala candidates than Republicans are for non-Trump candidates. I'm a bit skeptical that either of these data points will be reflective of the final electorate.
They got a sample with a fraction of a percent more Rs than Ds in the raw data. They then did a bunch of weighting, which seems to involve things like past nonresponse bias, and end up weighting to an R+2 electorate.
I'm perfectly willing to believe that if the electorate is R+2, basically 2014 redux, then we lose. I doubt that's what the electorate will be.
And to be clear, I think they're overfitting. If you get more Rs than Ds, I'm not sure why you'd assume that the nonresponse bias is similar to past polls in which you had trouble reaching Rs.
If Democrats are just as or more excited than Republicans about voting (see also CBS/YouGov battleground state polls) then we probably don't get an R+2 electorate. And if Harris wins independents, even by four points, then she likely wins.