In a blue county, the straight ticket vote could be 80-20 for Biden. Let's say 20 percent of voters are independents and they vote 50/50. This would plot as 20, +30 for Trump.
Hmm, I don't think so. Wouldn't that plot as 20, -6? If Trump got 80% of 80% of the vote and 50% of the remaining 20%, his total vote percentage would be 74%. That minus the 80% of the straight-party vote he got would be -6, no?
Why should independents break the same as the rest of the precinct, or why should they be consistently +5% or -5%. In other words, why should independents break 80-20 in a blue county and 20-80 in a red county? I don't reach that conclusion at all.
I see your point. My gut response is that there are probably fewer true "independents" than you think. If most of the people in a precinct who opt to use the straight-party voting method are "fully Republican" enough to vote straight-party Republican, it seems to me that it's probably also true that most of the people in that same precinct who don't use that method are at least sufficiently "Republican" to vote at least mostly Republican, i.e., in this case, vote for Trump. (And I'd also imagine that at least some of the people who use the individual voting method may actually end up voting straight-party.) So I would expect there to be a reasonably good correlation between the two metrics.
Y (I think) is % independent Trump votes - % Republican straight ticket.
So my example is 20% straight Republican and 50% of independents, which would be (20, 30). A county with 80% straight Republican and 50% independents would be (80, -30).
Hmm, I don't think so. Wouldn't that plot as 20, -6? If Trump got 80% of 80% of the vote and 50% of the remaining 20%, his total vote percentage would be 74%. That minus the 80% of the straight-party vote he got would be -6, no?
I see your point. My gut response is that there are probably fewer true "independents" than you think. If most of the people in a precinct who opt to use the straight-party voting method are "fully Republican" enough to vote straight-party Republican, it seems to me that it's probably also true that most of the people in that same precinct who don't use that method are at least sufficiently "Republican" to vote at least mostly Republican, i.e., in this case, vote for Trump. (And I'd also imagine that at least some of the people who use the individual voting method may actually end up voting straight-party.) So I would expect there to be a reasonably good correlation between the two metrics.
X is % Republican straight ticket
Y (I think) is % independent Trump votes - % Republican straight ticket.
So my example is 20% straight Republican and 50% of independents, which would be (20, 30). A county with 80% straight Republican and 50% independents would be (80, -30).