Tom Schaller argues in the New York Times
that Barack Obama should abandon the South, or at least the South
outside of Florida and Virginia. I don't agree with the entirety of
Schaller's reasoning. In particular, while I see the same inverse
correlation that Schaller does -- the greater the number of black
voters in a Southern state, the fewer white voters tend to vote for the
Democrat -- I don't necessarily see a causation. Our regression model
seems to do a pretty good job of explaining the Southern vote without
any reference to some sort of racial interaction effects, by focusing
instead on things like the number of white evangelicals in the state
and income levels.
Nevertheless, I do tend to agree with
Schaller's conclusion: the South (again excluding Florida and Virginia)
is fairly likely to disappoint the Democrats again. We have a number of
polls today to back that up.
Let's focus first on the results in Georgia, Louisiana and North Carolina. Obama did get a bounce from Strategic Vision's
last poll of Georgia, where he had previously trailed by 14 points. But
he remains 8 points behind, and while Bob Barr is pulling 3 percent of
the vote away from McCain in Georgia, it doesn't appear to me that Barr will have the resources
to improve that number significantly. In North Carolina, Obama has been
stuck at 3-5 points behind John McCain for quite a long time; the PPP poll today confirms that conclusion. And in Louisiana, Southern Media & Opinion Research has Obama 16 points behind, just as it did in April.
It
seems to me that Obama's numbers in states like North Carolina and
Georgia are liable to come in within a relatively narrow range. He'll
do better than a Democrat like John Kerry did there, with substantial
support from blacks (although Schaller is right that African-American
turnout has not been
particularly low), students, information-sector workers, and new
migrants to the region -- as PPP notes, Obama is leading by 6 points
among people who have moved to North Carolina from outside the state,
but trails by 13 among people who were born and raised there. But where
Obama is disliked in the South, he tends to be disliked a lot; his
"very unfavorables" tend to be pretty high in the region. There just
aren't that many swing voters in the South, and the Democrats are left
watching the paint dry and the demographics gradually become more
favorable to them in states like North Carolina and Georgia. North
Carolina could be 2012's Virginia, and Georgia could be 2016's, but
it's probably too soon for a non-Southern Democrat to be winning states
in the interior of the region.
Florida, however, remains its own demographic entity, and there an Obama win is more plausible. PPP has him leading by 2 there
-- a big move upward from their only previous poll of the state, which
had Obama 11 points behind in March -- although Strategic Vision has
him trailing by 6. The fact is that Florida remains something like
Obama's Plan C or Plan D for winning the election, but any state with
27 electoral votes and where the polling appears to be this volatile
will need to be closely monitored.
And once we move entirely outside of the South, Obama appears to be doing quite well. SurveyUSA
now has him 20 points ahead in New York -- up from 10 points before --
and he's holding onto a 5-point advantage in both major national
tracking polls. From everything we can tell, Obama's post-primary
bounce has plateaued, but not peaked, though it does appear to be
concentrated in particular regions, some of which (like the Rust Belt
states of Ohio and Michigan) have been quite helpful to Obama, and
others of which (like New York and California) are fairly superfluous.
Finally,
one quick methodological aside: the Strategic Vision polls were
"leaked" today by Political Wire with a limited number of details. I
have filled in my guesstimates of survey dates and sample sizes based
on their typical patterns, but we will correct those tomorrow as needed.
--Nate Silver