SurveyUSA (9/6-8, likely voters):
Kay Hagan (D): 40 (41)
Liddy Dole (R-inc): 48 (46)
Chris Cole (L): 7 (7)
(MoE: ±3.9%)
We’ve seen Kay Hagan surging ahead in other polls, so what’s the dealio here?
Well, this is a weekend poll (which any pollster worth his salt will tell you is a less than ideal time to field a survey) taken immediately after an effective GOP convention. Apparently, McCain’s electrifying convention speech and his selection of moose huntress Sarah Palin has awoken a zombie Republican army, as the poll’s crosstabs reveal a partisan sample of 41% R, 40% D. That’s a dramatic jump from the 46% D, 33% R partisan sample that SUSA found in August.
The findings trickle up the ballot, too: McCain has a monstrous 20% lead over Obama, and Pat McCrory leads Bev Perdue by eight points.
No, I’m not putting much stock in these numbers, although it’s possible that the GOP (especially McCain) received a mini-bump. Our friend Tom Jensen of Public Policy Polling doesn’t believe these numbers, either, and that firm is fielding a new poll starting tonight in order to set the record straight.
this has been a four to six pointin every poll, even closer than that recently. Palin shouldn’t have given that big a boost. What are the cros tabs? I bet they under sampled minorities again, and the D to R ratio is odd.
I’m sorry, but I don’t buy a 76-27 lead among white voters. Not in a state like North CArolina where black voters make up 20-23 percent of the vote totals, yet Democratic presidential candidates have still gotten 44% of the vote. Other odd things in the McCain poll area 59-38 lead among independents, and a full 24% of voters whose self-described philosophy is liberal. I don’t like these one bit, the poll looks screwy, and not because it’s bad news, but because its so out of whack with other polls and the internal numbers don’t ring true to me.
Note: All numbers were paraphrased.
That needs to be re-weighted to make any sense.
I did the math and this is what I came out to:
Hagen (D) – 45
Dole (R) – 42
That is using 46% Democrats/33% Republicans/18% Indies, which is a pretty accurate representation of NC voters.
The result for all of those races?
Hagan 46
Dole 39
Dole dropped 7 points and Hagan gained 5.
Same thing for the presidential and you get
McCain 46
Obama 39
Which shows a bump of 2 for McCain.
As for the governor’s race if you reweigh it based on the August poll then here is the result
Perdue 47
McCroy 39
Which would mean Perdue held steady while McCroy lost 5 points.
So the biggest shift was in partisan ID although McCain did make a slight gain, McCroy a big drop, Dole a big drop and Hagan a big bump no matter which sample you use.
Does anyone know the actual makeup of the NC electorate?
I prefer to call them “Palin-drones.”
41% R, 40% D? That’s oversampling the Rs by 8%, and undersampling Ds by 5%. Blacks are undersampled by 2% as well – pretty unrealistic with Obama on top of the ticket.
NC has added 177,000 new Democrats this year, 127,000 independents, and only 22,000 Republicans. While many people used to register as Dems and vote GOP, those same folks register as ‘Unaffiliated’ now.
SUSA’s likely voter screen is also probably off, not accounting for North Carolina’s one-stop registration and voting period that runs Oct 16 – Nov 1. Election Day runs for 16 days in NC, and you can register right thru Nov 1st.