Pollster.com

The NC polls definitely show an erosion in Obama's lead... Looks like undecided are going for Hillary.

actually, i think the latest polls say the opposite. there was an erosion several days ago, but he seems to have pulled it out.

anyone know if the RNC is running those Wright ads?

When you see numbers in NC that are all over the board (Insider Advantage Clinton + 2, Zogby Obama + 16) it really makes you question the predictive power of any poll. What I do think is predictive are demographics, recent events and history. In a post a couple of weeks ago I proposed that a demographic overlay of all the primary states would be more accurate than any of the polls. I still think that may be true. I also think historic trends play a role (if you can call 4 months history). In past primaries undecideds have gone to Mrs. Clinton. I would guess that will happen in Indiana and maybe in NC. The only real wild card is recent events. A primary is more like a prize fight than anything else. Fighters always know that if they win the last round they can influence the judges and win the decision. What happens over the weekend seems to have a real impact on the outcome. Perhaps the MTP interview will have a real impact. If I were to bet I would take Clinton by 7-8 in Indiana and Obama by 8-10 in NC. I am not sure how the media will interpret those numbers but to me it will look like Obama's problems amongst white voters are intractable. Of course an Indiana win and a NC blow out would show just the opposite. Just have to wait until Tuesday.

per richard above,

I agree that demographics are going to play a better predictive role here than polls. But I also think turnout is key. All of the NC polls in the last two weeks have AA turnout at ~33%. Results from the NC B.of Elections show AA turnout at 38% in early voting. And that has the potential to increase for a number of reasons.

That said, if we use 35% as a baseline for AA turnout and 60% for whites, and 5% for "other" (there are a lot of new hispanics in NC), and demo breakdowns as follows: AA: Obama 90-10, Whites: Clinton 60-40, Other: 50-50, the final result ends up: 58-42, which I think is pretty darn close to where we'll be.

And it makes sense, Clinton hasn't really been able to crack the low 40s in any of these polls. Obama had a similar problem in PA polls and look where he ended up.

Just a thought.

" Results from the NC B.of Elections show AA turnout at 38% in early voting. "

how do you know this?
Link?

Mike: Agree with your premise. Tweak the numbers a bit (say 85-15 and 65-35) and the race gets much closer. I am guessing that after all the money and attention there are really very few voters in play.

tybo,

it was a blog on DailyKos: http://www.dailykos.com/story/2008/4/30/14835/7007
It was 38% AA, 58% white

Also if you can figure out the NC Board of Elections site you can see the hard numbers. I still can't seem to find it. Not sure if they report every day.

and yes richard, you're right, a small tweak either way changes the score. i.e. say if he managed to hold it to 55-45 with whites (NC is more like VA and not PA in terms of white Dems)

Mike: You are right. A few point swing in either's base means the difference between a tie and a blowout. My sense is that this is a more favorable time for HRC than February and she should do better than in VA, but as you said NC is much more like VA than PA. I believe that demographics are the most predictive factor in any primary. NC should be a real test of whether events can overturn demographics.

Mike, thanks for the explain. I can't find it either via the link , but the NC site says the data(gender, race, age) is supplied by the voter by choice and is only available for some counties
(from your link).