How does the "southernness" of a state affect the Democratic Primary?
Clinton has been doing well in "The South." Of course, defining what "The South" is is pretty tricky. I divided up the states by "Deep South" vs. Other, so all the usual orginal deep south states count as "southern" except Florida. You know what they say about Florida. "The farther south you go, the farther north you get." Also, Texas is not deep south in the traditional sense.
Using this rough division, Clinton wins all the time in the "Deep South" and Sanders wins some and loses some in the other states, as shown in this handy dandy graphic:
This is good news for Sanders, because the three big states coming up are New York, Pennsylvania, and California, and they are not in the deep south. Of the remaining states, only Kentucky is southern. So, this biases future primaries, given this one variable, towards Sanders.
Being "deep south" vs. not is in and of itself a bit of a meta-variable, associated with other factors with in the Democratic Party, mainly ethnicity.
And if if break out "western" voters, you will find they generally prefer Sanders.
Just as Texas is not southern, it in not western either; the saying is that the line between the south and the west runs between Dallas and Fort Worth.
Given that we see Bernie as the candidate of the Occupy movement, we tend to forget that he is Jewish. I would be surprised if good old-fashioned Southern antisemitism did not account for a significant portion of his weakness in the south.
Not to be a Debbie Downer, but the fact is Sanders is still the guy for whiter and wealthier states. While it's true his percentage of African American voters goes up as you go north, he needs to get past the ~30 percent mark. He might do that in NY, but I don't see a path for that elsewhere.
Now I know that minorities do vote for Sanders. So I'm not going to listen to that "you're erasing minority voters who like him" line. There's a difference between saying that he can get minority voters to vote for him -- clearly he has some -- and saying whether enough will do so to put him over the top. We can go back and forth as to why that is all day, but for now it's an empirical fact that in more diverse states he simply hasn't done as well as he needs to.
(Fviethirtyeight has an interesting analysis that notes Sanders isn't doing as well in states that actually look like the Democratic electorate).
If Sanders way outperforms the polls in NY he'll split it. But that won't be enough. I do think it will be very interesting to see if any native son effect works for him in Brooklyn. If Clinton underperforms seriously in NY that will change the media narrative of the campaign -- New Yorkers did (at a statewide level) vote her into office twice, after all.