Rasmussen Poll Shows Perdue up 5 in Senate Race

Rasmussen released the top line of a poll today showing David Perdue leading Michelle Nunn 46-41.  Their previous poll showed Perdue leading 46-40, so no real change. Expect to see a new Governor’s poll from Rasmussen in the next few days.

24 comments

    • Bobloblaw says:

      I’m still skeptical of an outright win for Perdue. More like 2008 I think. Then Perdue wins the runoff. But by less than the 16 chambliss won it by. If dem voters weren’t inspired to give Obama 60 votes in the Sen in 2008, I don’t see huge dem turn out for a jan 2015 primary.

  1. I like Rasmussen. But I do wish that the standard for republishing numbers would be cross tabs having to be publicly released for public analysis. I tend to doubt they weighted correctly here simply because every other out-of-state pollster has weighted incorrectly during this calendar year.

    **Out-of-state pollsters continue to underrepresent the actual number of African-American voters and female voters in their polls. I’m writing this post to continue to point this out.

    –The AJC’s pollster, Abt SRBI, released poll numbers with just 24% of voters being black (which is silly-low… absurdly low … though, in fairness, they acknowledged their objective was simply to report the poll “as it is today” without accounting for likely demographic results in the election); they also clearly underrepresented female voters at just 51%.

    Female voters make up 54-56% of Georgia general elections. Black voters make up 29-30% of voters in Georgia general elections. Since female voters are currently going for the Democratic statewide candidates by double-digit margins, the reported numbers are skewed.

    With these weighting corrections, their numbers would be almost identical to ours at Landmark.

    –SurveyUSA originally weighted their poll to 26% black voters and 51% female voters (though they essentially acknowledged their weighting was wrong and changed it to 29% black on their more recent survey. SurveyUSA continues to weight incorrectly for female voters at just 51%, far too low given that female voters always make up 54-56% of the vote. This again matters because of the vote differences between males and females.

    –NYTimes/YouGov: perhaps the worst poll. They had the female vote at just 52% and the black vote at 25%. Terrible.

    More accurately weighted based on historical numbers and the actual electorate, Landmark has Gov. Deal 44%, Carter 47%, Hunt (L) 4%; for Senate it’s Perdue 43%, Nunn 46%, Swafford 6%; and for Lt Gov it’s Cagle (R) at 50% to Stokes (D) at 41% (poll Sept. 9-11, poll for WSB-TV).

    • Ralph says:

      Gee, a Landmark polling representative saying his poll is better than anyone else’s.

      Next, we might hear that Deal says he is the better candidate and Carter says he is. Who are we to believe?

      • I know, I know. And Ralph, you are correct to second-guess anyone who puts numbers in the public domain for discussion. That’s why I also raise the point of results being republished without cross tabs being included for the public to read. Otherwise, what precludes someone from releasing numbers like, “Rand 70%, Cruz 30%”? Nothing really. Cross tabs help keep it real.

        Yes, we are stating that there is a more accurate way to project turnout, and therefore a more accurate way to weight polls. There is not an exact number criteria, but a range that I personally find reasonable. Each to their own, of course.

        This has been an on-going issue throughout this summer and fall. And with candidates and political parties possibly basing their decisions on faulty polling, it’s a pretty serious conversation.

              • I just got confirmation that Rasmussen has the black vote at 26% and the female vote at 52%.

                That means that in order to have an accurate gauge of the head-to-head numbers, you must add a net +4 or +5 to the Democratic candidates’ numbers in Georgia for Rasmussen poll numbers. Here is why: add +3 or +4 for the net undercount of the black vote (Rasmussen pegged the black vote at 26%, while 29-30% is the historically accurate number), and add +1 for the undercount of the female vote, which was actually 55% in 2010, not the 52% that Rasmussen has weighted it for. …And note that the Georgia electorate is even more demographically female today than it was in 2010.

                I really don’t like writing these points over and over. But these weighting discrepancies keep coming up in the national pollsters’ work.

    • Bobloblaw says:

      Of course as you know it isn’t just the weights but the margins within each weights. I think that is where your august poll was wrong. But even the Nunn fans here have to be concerned that the only poll showing Nunn leading showed her lead cut in half in three weeks.

      I suspect if women were going for Nunn by the near 20 points landmark was claiming shouldn’t we be hearing stories in the press how so-and-so voted GOP their whole life but this time they’re voting for Nunn. Endorsements from groups who normally endorse GOP candidates. Without a significant number of crossover women supporters Nunn can’t be up 15-20 among women.

      • In a 30% black electorate overall, women are about 36% black and men are only 23% black.

        So…
        36% * 85% = 30.6
        64% * 30% = 19.2 / total 49.2

        36% * 5% = 1.8
        64% * 55% = 38.4 / total 40.2

        And btw Nunn will eventually be at 95 with the 36%.

        • Bobloblaw says:

          that’s not 20 points is it? I can see Nunn winning by 10 among women.

          30% electorate black is pushing it. In 2008, it was 31%. What was it in 2012?

          • 30. And 28.3 in 2010, since when the overall registered electorate is about 3% less white and 1% blacker. Same turnout rate applied to race/gender/age as 2010 to existing registration (as of 8/1) puts it at 29+. No reason at this point to think that isn’t happening.

            • Bobloblaw says:

              Registered are not votes. Most likely those whites weren’t even voting in the first place. With 3% fewer whites and 1% more blacks why your think GA’s population was falling.

  2. Today it’s actually a 15% margin that women are voting for Nunn by a margin of 52%-39%, down slightly from August immediately following the contested GOP primary runoff and conducted at the time when Zell Miller endorsed Nunn. Yet men are 49-39% for Perdue. It’s a large gender gap, and underestimating it has an impact on the results of the out-of-state pollsters, negatively so for the reasons I have stated.

    The value to polling is that one’s feelings and suspicions of how people should be voting is set aside. Your feelings that, if a candidate were ahead, there should be endorsements and stories are a normal reaction.

  3. Bill Arp says:

    52% – 39% = 13%, not 15% ….I know you are pro Nunn but numbers are your game…get it right…..Nunn’s success with woman has matured due to their memory of how awful Sonny Perdue was while Governor. It’s not that they love Nunn, but rather they associate David with Sonny and that alone drives people to Nunn.

    • While I admire your enthusiastic Sonny dislike, I think if anything voters don’t think much of him at all these days. He was a likeable but forgettable governor, much like our existing governor (if he were likeable).

        • David C says:

          Yeah. The dirty secret is that Sonny was pretty crooked too. But unlike Sonny, Deal and co. are really, really bad at it.

          • Dave Bearse says:

            The surreptitious arrangement a $100,000 tax break, perhaps only for Perdue, by using means outside the normal legislative practice, and without consultation with the Department of Revenue, by Perdue’s neighbor, colleague and personal attorney Larry O’Neal, is well-documented.

            It only matter when it becomes so blatant, or gets a lot of attention. Pretty much everything on Balfour’s chiseling was in the public domain months long before he coasted to victory in his penultimate campaign for office.

            Keeping it bipartisan, there’s Crawford Lewis. How many PP readers during the course of their lives have ever purchased three full tanks of gas within 20 miles of their home for one vehicle during one day? Crawford Lewis did it.

  4. Bobloblaw says:

    I think the basic problem isnt Deal’s handling of the winter mess or even his corruption. But that the GOP has a supermajority and hasnt done anything with it. Dems are happy just to be in power, but the GOP base DEMANDs that the GOP reform everything. Taxes, schools. They havent done a thing. I think that is what is hurting.

Comments are closed.