Final PPP poll in NC + more NC data
by DPW, Mon May 05, 2008 at 08:50:19 AM EDT
PPP has released its final poll for NC, which shows Obama ahead by 10. The full release can be found here.
The specific base numbers:
Because demographics (specifically age, race, and gender) seem to have more predictive value than anything else, it's worth noting that PPP's sample consists of 35% AA and 56% female. Age distribution breaks down as follows:
Older Than 65........ .18%
And, for those of you who like to predict delegate allocation by congressional district, PPP includes a breakdown of the various area codes. Clinton's strongest area is, unsurprisingly, the 828 area code (which is the appalachian area in western NC). Obama leads rather comfortably every where else.
The poll also includes some interesting info relating to early voting. Specifically, according to PPP's sample, almost 30% of voters have voted early; and, among those who've voted early, Obama leads by 29%. SUSA also provides polling of early voters who've already voted, and in its most recent report, Obama led by 18 among those who have voted. If these numbers are correct, Clinton will have a lot of ground to make up tomorrow. However, because Obama has pushed his supporters to vote early (especially students due to conflicts with exams and summer break), I expect Clinton to perform better tomorrow than she has during the early voting period. This is something to keep in mind when exit polls inevitably leak tomorrow. Exit polls, I assume, will only reflect tomorrow's voting patterns, which may cause exit polls to overestimate Clinton's strength. Just my guess, though.
If you're interested in further information related to early voting, Mark Blumenthal has a nice article over at pollster on NC demographics, in which he shares publicly available data concerning the age/gender of early voters. Specifically, AA voters constitute 40% of the early voting electorate, while women constitute 60%. Age data is not available. Both of these numbers exceed the numbers reflected in every poll model used by pollsters, including the PPP poll cited above. Also, approximately 400,000 people have already voted. That would appear to be at least 1/4 of the primary electorate, but who really knows.
Finally, I highly recommend the Blumenthal article if you're interested in learning more about the demographic assumptions used by various pollsters.