A new Rasmussen Poll in North Carolina shows Barack Obama leading Hillary Clinton 47-40. We had it at 47-43 last week.
Rasmussen doesn't release its topline demographics but they do say that Obama leads by 53 points among African Americans, while Clinton leads by 20 among white voters. We actually had Obama up 63 points with black voters, but down by 23 with white voters.
What that adds up to is Rasmussen projecting higher black turnout than we are- probably somewhere around 35% rather than the 30% we had it at. We will probably move toward a 35% black turnout model next month but we took our poll before Hillary Clinton's strong performances in Ohio and Texas extended the race and ensured a competitive North Carolina primary.
What impact would 35% black turnout have on the primaries for President and Governor in North Carolina? If we had weighted last week's poll at 35%, Obama would have been leading 50-40. That's a pretty strong indicator of how much polls can vary based on their demographic composition- a 5% increase in black turnout turns a 4 point lead into a 10 point lead for Obama.
35% black turnout would not have a huge impact on the Democratic gubernatorial primary. It would give Bev Perdue a 52-24 lead over Richard Moore, just a one point gain from the 27 point lead she showed with it projected at 30%.