New YouGov Poll: Perdue Leads 47-41.

YouGov has completed their “second wave” of interviews in which they spoke to 108,725 likely voters across America, including almost 1900 here in Georgia. There results show the GOP taking the Senate and here in Georgia, Republican David Perdue leading Democrat Michelle Nunn 47-41 (with leaners added in). Libertarian Amanda Swafford polls 3% and 9% are undecided.

As best I can tell, the weighted results contain about 25% African-Americans.

Let the debate over the accuracy of this poll commence in the comments.


  1. Bono said it best: how long must we sing this song?

    The female vote in 2010 was 55% of the vote. 52% once again understates the likely real number. And with females more likely to vote Democratic so far this year, it causes a problem in the poll numbers: There were 2,533,243 voters participating, and 1,395,425 were listed as female (89,000 voters didn’t have a reported gender code on the SOS site, but this likely wouldn’t move the needle much).

    Second: The percentage of African Americans voting in 2010 (last gov election) was 28.25 (official) but really more like about 30% when the category of “unknown” race is actually added in (most officially unknown voters are minority by all accounts…perhaps two-thirds or more). This poll again has the black vote at 25%, which is likely woefully low. If (as I’m stating) it’s really 30% of the vote instead of the NYT/CBS projection 25%, then their poll essentially likely projects the GOP vote about 5% higher than it is.

    The purpose of my post here is not to discuss whether David Perdue is leading or not — he may or may not at this point — but rather to discuss the quite apparent methodological problems that we continue to see in publicly released polls.

    Also — this was an online poll, not a telephone survey.

    • jh says:

      Appreciate you chiming in Mark. I think Michelle will shock when she wins. This YouGov is rather bizarre, especially the Landrieu numbers. Hopefully Jason can avoid a runoff too.

      • Just for clarity: I do think that both Perdue and Gov. Deal ultimately win. The point that I have been making is that the Georgia electorate has been changing and statewide elections are going to be much more competitive in this new era.

        Due to changing demographics, Metro Atlanta is fast on a path to having 8 of the 10 “TSplost counties” electing Democratic County Commission Chairmen by around 2020. DeKalb, Clayton, Fulton, Rockdale already do, Henry, Douglas are now dead even, Cobb and Gwinnett both had Romney win by a mere 53% and continue to see a larger minority vote (though it’s most obvious in Presidential election years). Fayette will most likely stay GOP past that time, and Cherokee certainly will stay GOP.

  2. Bobloblaw says:

    “”And with females more likely to vote Democratic so far this year””

    I think there is the error in Mark’s reasoning.

  3. NorthGAGOP says:

    I’m not the biggest Roundtree fan in the world.

    I do agree with him that most of the polls are under weighting women, and possibly the AA vote. Most show Perdue and Deal losing women by big margins. I’m not a polling expert, so I can’t tell you if that’s a 5 point swing or a .5% swing. I do think that Perdue and Deal need to figure out a way to improve their support with women.

    • I’m not the biggest Roundtree fan in the world, either.

      But that said, maybe Landmark might have some interesting points to make in Georgia elections for governor…

      Earlier this year our surveys had the GOP candidates leading, but not by large margins. They fell back over the course of the summer while the Democratic statewide candidates campaigned essentially unchallenged and while the GOP fought tough primaries.

      I do enjoy the amusement of Bob (below) as he works himself up to a tizzy in denying a significant gender gap, yet then stating that there will be runoffs (to cause a runoff…the votes must come from somewhere) – and then inventing age range categories such as 18-29 that we never offered as an poll option and didn’t release (attention to detail matters, my friend).

      NorthGA, you are definitely correct that most polls show females currently voting more Democratic. As things stand today, barring a significant change there’s a runoff scenario. It’s happened before in Georgia (1992 Senate and PSC, 1998 for PSC, 2008 Senate and PSC). Historically that helps Republicans, of course.

      • Bobloblaw says:

        I don’t deny a gender gap but your poll is wrong in part because it has too little male support for Perdue.

        I still haven’t gotten you to admit that if the top races are heading for Dem victories why the entire Dem ticket wouldn’t sweep. I want to know who the Olens/Nunn/Carter voters are.

        • That’s because your premise is incorrect, Bob. And I addresses this last time you asked the question in another thread. You pretended at that time not to see it, as well.

          The GOP down ticket races are generally considered safe because they outspend their opponents 10-1; have name recognition from being incumbents unlike their Democratic challengers; and are essentially scandal-free in the mass-public’s eyes. They face no real challenges.

          What is different at the top of the ticket is that the Democratic challengers have matched or even bested the GOP candidates in fundraising (Carter out raised Gov Deal in the second quarter, which you may not be aware), and they face no serious primary.

          Finally, your supposition that I’m saying the GOP candidates would lose is incorrect. I have corrected you previously on this in other posts and threads on PeachPundit, as well, but you apparently missed it. Right now the elections would head to a runoff election. And those elections normally work out pretty well for Republicans.

          • David C says:

            Outside of the offices of Governor/Senator, Republicans never took any of the other statewide offices Democrats held in 2002 except when the Incumbent Democrats retired and they got to go after open seats. Cox and Taylor in 06; Baker, Thurmond, and Irvin in 2010. In ’06 the latter 3 each won by 10-15 points even as Sonny was romping to re-election. So it’s certainly possible for Dems to win Senate, Governor, and nothing else. Republicans did.

    • John Konop says:

      I know Mark he is just giving his best effort via his knowledge……Mark is smart guy….I happen think the election is closer to GOP then the Dems this cycle….Bur Mark is an expert I have tremendous respect for his work and he is willing to comment about his work on the PP.

      • I’ve always enjoyed our debates, John. And thanks for the kind words.

        I think something funny happens to folks who stay for a long time on a regular forum that they don’t own…they begin to laugh a bit more and not take each thing so seriously. It’s an enjoyment to see the long term commentators on PP.

        Yes, even Buzz Brockaway and John Konnop.

  4. Bobloblaw says:

    I contended that the main problem with Landmark was Perdue leading among men by only 7 while Nunn had a near 20 point lead among women. Here the Gender gap is much wider for Perdue. If Mark wants to quibble over demographics how does Perdue getting 59% of the white vote sound? Anyone believe that? That is what this poll stating. So plenty of room for Perdue there.

    There are things Mark can criticize about this poll but then there are assumptions that then offset those assumptions. 59% of the white vote for Perdue, a huge 25 point lead for Nunn among 18-29 year olds (Will they even show up in proportions to those questioned in the poll). Would you rather have a 25 point lead among 18-29 year olds or a 40 point lead among those over 65? 7% of whites undecided but 4:1 lean GOP.

    I think the 70% black vote for Nunn and 59% white vote for Perdue basically cancel each other out.

    Still to me looks like a runoff where the Dems do very poorly historically. So the Senate will NOT be decided on Nov4th. GA and LA will have runoffs.

    • When you factor in leaners the Perdue number was (I think) closer to 64 and the Nunn number was closer to 76.

      Still – in 2010 19/20 Republican votes came from white voters, who made up 66% of the electorate. Assume that’s the same this year and unless they are getting 75% of the white vote, they aren’t winning without a runoff.

      That’s really the only number I’d look at.
      White vote for Republican >65 = runoff. >75 = win without runoff.

      We aren’t anywhere near that category, regardless of whether any pollster can accurately interview blacks.

  5. saltycracker says:

    I’m surprised the spread isn’t much more with Nunns nutty ad campaign of the discredited pillow talk and her Daddy’s buddy. Tiring.

    • Will Durant says:

      You would think with all the millions each side’s PACs could produce a little variety in their attack ads. If the NRSC continues running the repetitive loop of Nunn’s 5 word out takes that she supports the President ad nauseum, it is going to work against them. That is if the 999,999 times it has already played hasn’t already.

      I read the other day that $22.5MM has already been spent on this one seat. Wonder what the over/under should be if it goes to a runoff?

      • saltycracker says:

        Tired of that ad too, we already know she’ll talk riddles & vote the line – ready for substance out of both of them

  6. MikeS says:

    It was an internet online poll over two weeks. Online polls are cheap, but under count minorities especially poor uneducated voters. How much is subject to debate.

    • Will Durant says:

      Also it is opt-in and they even admit that 3% of the respondents had deleted their cookies. Their results could easily be skewed by any 14-year old that is already accustomed to getting around a net nanny.

  7. debbie0040 says:

    I think the GOP’s handling of the Nydia Tisdale event has hurt them with women and with independents. Both Deal and Perdue and the GA GOP should have condemned what happened. Sam Olens did not hesitate to and neither did Dawson County GOP chair. Pumpkingate is getting traction…

  8. debbie0040 says:

    I am not good at predictions but I don’t believe there will be a runoff in the U.S. Senate race. I think Perdue wins handily without a runoff.

    I think Landmark’s polls are the most accurate but they are only a snapshot in time. People still aren’t paying attention to the election and people are simply reacting to the ads/news cycle. Once they begin to think and digest the information, (1st of October) the polls will have meaning.

    All Perdue has to do is to not make any gaffes and wrap Harry Reid, Obama’s immigration, national security and overall policies around the neck of Michelle Nunn and she is toast.

    With that said, the GOP needs to clean up the Pumpkingate mess created by Mr. Bearden and the Dawson County Deputy. They need to condemn what happened . Gov. Deal needs to make a strong statement against the journalist being ejected and assaulted by the deputy. He needs to do this because if I am not mistaken, I believe Bearden worked for him when he was Congressman. Gov. Deal can’t afford to be silent on this..

  9. northside101 says:

    Perhaps everyone can agree that “41” (Nunn percentage in this poll) is pretty much the rock-bottom Democratic base in Georgia, close to the 39 percent Michael Thurmond got in his 2010 kamikaze run against Senator Johnny Isakson. Nunn almost certainly will do better than that, but if she is polling 41 in this 25% black poll, I agree with Mark, she would be higher at 30%, maybe around 45%, which matches what Obama got in Georgia in 2012.

    Would be interesting to see in that poll—and any poll for that matter—the breakdown in metro Atlanta (now 29 counties according to the US Census Bureau) and the rest of Georgia. In recent elections, Democrats have run better in metro Atlanta than the rest of the state. Listed below is how metro Atlanta (28 counties for 2008-2012 elections) and rest of the state voted in various contests (percentages not adding to 100% because third-party voting left out)

    2008 PRESIDENT:
    Metro Atlanta—-51.5% Obama, 47.8% McCain
    Rest of Georgia–41.2% Obama, 58.2% McCain
    Overall (Whole State)–47.0% Obama, 52.2% McCain

    2010 GOVERNOR:
    Metro Atlanta—46.1% Barnes, 49.6% Deal
    Rest of Georgia–38.8% Barnes, 57.6% Deal
    Overall———–43.0% Barnes, 53.0% Deal

    2012 PRESIDENT:
    Metro Atlanta—-49.7% Obama, 49.1% Romney
    Rest of Georgia–40% Obama, 59% Romney
    Overall——45.5% Obama, 53.3% Romney

    Rounded to nearest whole number, metro Atlanta (then 28 counties), accounted for 57 percent of the state’s total votes in 2008, 2010 and 2012. If Nunn/Carter cannot improve significantly improve on Obama’s showing outside metro Atlanta, then the only way they can win is to win convincingly in metro Atlanta—basically, if Nunn/Carter only run slightly ahead of Perdue/Deal in metro Atlanta, Republicans win.

    Looking at the statewide races by congressional districts, Nunn/Carter of course will win majority-black CD 2, 4, 5 and 13. But it isn’t likely either can win by taking just 4 of the state’s 14 congressional districts—probably have to win at least 2 others, maybe 3, to win—and the state’s 10 other congressional districts all backed Romney by at least 10 points last time, so both Nunn/Carter would need some Romney voters to prevail in November.

Comments are closed.