Overnight, the news has gotten even a little better for Senator Barack Obama. We're now seeing trends in both IN and NC that reflect the national movement toward Obama, and this could well mean that the sweep I suggested was possible yesterday now is more likely, even if its no cinch.
First, NC. It is, at this point, a lock. Given the amazing turnout here in early voting, about 400,000 (or 1/4-1/3 of the total expected democratic turnout), Obama comes into a today with what looks to be a tremendous lead (12 points according to one analysis at
http://ccpsblog.blogspot.com / before a single vote was cast today). Cap that with consensus between two pollsters (who obviously are using quite different weighting models) that he's been increasing his support through yesterday, and NC looks good for him. Here's NC by the numbers, by pollster, the last few days to get an idea of trends, with the most recent survey from each pollster first:
5/5, Insider Advantage: Obama 47, Clinton 43 (Obama +4)
5/4, Insider Advantage: Obama 48, Clinton 45 (Obama +3)
5/4-5, Zogby: Obama 51, Clinton 37 (Obama +14)
5/3-4, Zogby: Obama 48, Clinton 40 (Obama +8)
5/2-3, Zogby: Obama 48, Clinton 39 (Obama +9)
5/1-2, Zogby: Obama 46, Clinton 37 (Obama +9)
Note that for Zogby to jump 6% in a tracking poll, Obama must have polled a massive blowout number (around a +20) for yesterday only. The MOE for that day alone would be big, but a 6% change in a tracking poll is a big spring any way you figure it.
Nobody else has polled after Sunday in NC, but everybody who polled this weekend had Obama up in NC.
Now INDIANA, which still is a tougher call, especially because of the Zogby problem. There, only Zogby has Obama ahead, and, obviously, we can't base any opinions on Zogby alone (even though Zogby nailed the margin in PA). Here's IN by the numbers the last few days, by pollster, most recent polls first:
05/04-05/05, Zogby: Clinton 43, Obama 45 (Obama +2. 5/5's numbers only: Obama +6, w/an MOE too large to count on.)
05/03-05/04, Zogby: Clinton 42, Obama 44 (Obama +2)
05/02-05/03, Zogby: Clinton 41, Obama 43 (Obama +2)
05/01-05/02, Zogby: Clinton 42, Obama 43 (Obama +1)
04/30-05/01, Zobgy: Clinton 42, Obama 42 (tie)
*Note the undecideds breaking 3 to 1 here FOR Obama, as they used to do earlier in the primary season. IF this holds for IN, it'll break Clinton's recent trend of of winning a majority of late-breakers. In addition, as of Sunday-Monday there were only 7% undecideds left, according to Zogby. IF Zogby's numbers are good for 5/4-5 and have held, then Obama needs only to secure 36% of the undecideds left to win the state.)
05/04, Insider Advantage: Clinton 48, Obama 44 (Clinton +4)
04/30-05/01, Insider Advantage: Clinton 47, Obama 40 (Clinton +7)
*Note the undecideds again breaking FOR Obama, here 4 to 1 over the last few days, similar to how they used to break for him earlier in the primary season. According to I.A., there still were 8% undecideds left on Sunday, meaning Obama will have to win them by a hair over 3 to 1 to squeak out a win (if I.A.'s numbers have held and are right.)
Nobody else yet has a poll for Sunday and/or Monday only in IN, but two other surveys from Saturday-Sunday polling only had Clinton ahead in IN (Suffolk, +6; PPP, +5). So did other pollsters whose polling took place at least in part before the weekend.
Predictive wrenches in IN that could influence voters enough to affect outcome and/or make pollsters look silly:
1. The gas tax holiday battle. Did Obama's message and the media's drift toward a meme of labeling the holiday as pandering stick or not? (This, to me, is why Sunday-Monday polls are most important in IN.)
2. Indiana's new voter ID law is sure to baffle college students who show up with their college IDs, some of which won't count as ID, as well as the older crowd. Plus, it's finals time.
3. Rush L.'s dittoheads voting for Hillary, who have influenced the vote by 2-4% in PA and TX according to this site:
http://www.fivethirtyeight.com/2008/05/indiana-toss-up-...4. Early voting (160,000 in IN). It has been very strong in IN counties expected to break for Obama, but as a whole nothing like the statewide turnout in NC.
I still think it might take too many IFs for an Obama win in Indiana, but we have two pollsters with competing methodologies, one whose sampling favors Clinton and one whose sampling favors Obama, but both of which show drift toward Obama in the most recent polls we have in both NC and IN. It is now almost certain Barack Obama will win NC AND net more delegates today. IF Indiana's support for both candidates follows the trends above, breaks Obama's way even a little, or hardens fairly evenly for both candidates, then it looks like the state could be extremely close, a push, or even a win for him.
http://www.realclearpolitics.com/epolls/2008/latestpoll...http://www.fivethirtyeight.com/2008/05/indiana-toss-up-...http://www.pollster.com/08-IN-Dem-Pres-Primary.phphttp://www.pollster.com/08-NC-Dem-Pres-Primary.phphttp://ccpsblog.blogspot.com /