Our August Presidential poll numbers are in and we have McCain widening his lead from 3 points to 6 points. Full press release here.
The big swing this month is among unaffiliated voters who went from McCain +1 in July to McCain +18 this month.
Obama still can’t seem to crack the low 40s here in any of our polls despite spending upwards of $8 million so far in NC and having like 2 dozen field offices.
And his campaign still continues to struggle connecting with white voters with only 29 percent support.
As I wrote earlier, it’s going to take more than relying on new voters to show up for Obama to win NC. He has to get his share of the white vote up to the upper 30s AND have black turnout hit record highs.
NC will be closer than the Bush/Kerry split in 2004, but barring some big changes in the poll numbers soon, it doesn’t seem like he can win here.
Chris,
Do you have any concerns about how recent voter-registration trends might affect the sampling? I don’t recall how the poll is weighted.
John,
We cast a fairly wide net to try and incorporate the best estimate of likely voters for this fall. To be a part of our sample, you have to have voted in either the 2002, 2004 or 2006 General election (1 out of 3) or registered to vote since the 2006 General.
We then ask a screening question on whether the person we call is very likely, somewhat likely, not very likely or not going to vote. The latter two answers we terminate the call.
This month’s poll is actually heavy on newly registered voters. Registration has grown roughly 6-6.5% since 2006 and our poll has 11% of its respondents as new registrants.
While there has been much made about the relative large numbers of new registrants (approx. 250,000 this year), they still compromise a rather small portion of the total number of registered voters (5.9 million)
Thanks, Chris. Do you see any significant change in party ID of your sample over time?