j reacts to west virginia polling, and what a surprise sanders win *actually* implies

there's some polls coming up in west virginia, and they're kind of splitting the difference - the racist models are suggesting he should win in a landslide (because there are few blacks), where as i actually think a better projection ought to be clinton by about 10-15. the polls i've seen are in the margin of error, suggesting a split, but they're also small samples and wonky sources...

forget how many blacks are in west virginia. it's a boneheaded way to go about this. instead, let's look at how she did amongst whites in the neighbouring states - all of which were open primaries.

virginia: 57% of whites voted for clinton
ohio: 53% of whites voted for clinton
tennessee: 57% of whites voted for clinton
kentucky: hasn't voted.

so, given that west virginia is mostly white, why exactly would you expect sanders to win?

?

but, i'm data driven. so, i'm going to suggest a way out, and it's this - shit happens. people change.

again: i think some skepticism is in order with the polling. we don't have anything really solid right now. but, if sanders does win west virginia, what i would suggest is that voting intentions in the south have shifted since march.

this is likely a fruitless request: if you were to do some rigorous polling in tennessee or virginia, today, what would the results be? because if sanders wins west virginia (after winning indiana), and is then at least competitive in kentucky, it's hard to see how the results from the early primaries can be current.

again: if you want to take the information you have and apply it to west virginia, you should get a big clinton win. if sanders wins, it strongly suggests that southerners, or at least white ones, are actually abandoning clinton.

let's not get carried away, though. i actually still expect clinton to win in west virginia. and kentucky.