Final SurveyUSA Poll has Deal Up by Five; Perdue Up by Three

Governor Nathan Deal appears to be closing in on re-election without a runoff, according to a new SurveyUSA poll out this morning. The poll is one in a series commissioned by Atlanta’s 11Alive.

Deal leads Jason Carter, 47% to 42%, with Andrew Hunt and undecided voters each at 5%. Deal leads every age group except millennials, where he trails by 15 points. Deal and Carter split the female vote, with 45% each, and among those who have already voted, Deal has 51% to Carter’s 45%. The Libertarian, Hunt only gets 3% of those already voting, another sign that libertarian leaning Republicans may be voting GOP in order to avoid runoffs. Acording to the polling memo issued by SurveyUSA,

[S]hould Deal win by 5 points, and/or should Libertarian Andrew Hunt be held to less than 5% of the vote, a runoff will be avoided, and Deal will win his next term tomorrow, 11/04/14. Should Deal win by fewer than 5 points and/or should Hunt get 5% or more of the vote, a January 2015 [sic — runoff date would be December 2nd] runoff will follow. A Carter win would be an enormous upset, given these poll results.

Things are a bit closer in the Senate race. There David Perdue leads Michelle Nunn 47%-44%. Amanda Swafford has 5% and undecided voters make up the final 4%. Roughly the same number of people in each age group support Perdue as support Deal. Similarly, support for the two GOP candidates is similar among whites and African Americans. The difference making the Senate race closer than the Governor’s race is women voters. In the Senate race, Nunn leads Perdue by six points. Is a runoff likely?

Were Perdue to win the seat by 3 percentage points, and were Libertarian Amanda Swafford to get less than 3% of the vote, a runoff would be avoided and Perdue would win the seat outright tomorrow, 11/04/14. Were Perdue to win the seat by less than 3 percentage points, and/or were Swafford to receive more than 3 percent of the vote, a January 2015 runoff would be required. Were Nunn to win the seat, and only a narrow Nunn win would be consistent with these results, that would be a significant upset.

In down ballot races, Casey Cagle leads Connie Stokes in the Lt. Governor race, 54% to 38%. Brian Kemp is ahead of Doreen Carter, 52% to 40%, and Sam Olens has 51% to Greg Hecht’s 39%. In the race for School Superintendent, Richard Woods holds a narrow lead over Valarie Wilson, 47% to 44%.

The poll of 591 adults was conducted from October 30th to November 2nd. The poll’s 4.1% margin of error keeps the Senate and State School Superintendent’s race tied within the margin of error. View the crosstabs here.


  1. Stefan says:

    And the cross tabs reveal African-American projected to be 26% of the vote. That seems low, given early vote at 33%. By my math, AA election day turn it would have to be 23% or lower to give that result. Very unlikely.

    • Bill Dawers says:

      Black vote almost certainly too low — and probably low by at least 3 points. Other issues: Nunn and Carter only getting about 80% of the black vote. Women only 52% of the electorate. In short: another poll that seems to be based on bad assumptions.

  2. Stefan says:

    Wow, more good news for Dems in the cross tabs. Michelle winning 29% of the white vote, Jason getting 28%. Those numbers are good enough for Dems to think about victory.

    • TheEiger says:

      Bless your heart. You sound like me in 2012 trying to show everyone how every poll showing Romney losing was wrong.

    • John Konop says:

      I think a huge swing could be the other vote category, depending on what is in that group….If it is multi culture….that could be a few point pick up for Dems….I still think the Dems will come up short….But 16 could be a big breakout year if you guys run the right candidates and the GOP cannot control crazy talk…

      The one strange factor if the election goes into a run-off is the Ferguson ruling comes out before the run-off election….that could throw things off via who comes out in a run off….Which I bet is why the Dems played that card in the general….

  3. northside101 says:

    John, the answer is “yes”—you can vote in any general election runoff even if you do not vote in the “first round” (which of course is tomorrow, November 4). I would suspect, though, that those not voting on November 4 (or in early voting) are not likely to come back (should there be one or both) a December 2 runoff or January 6 runoff. Turnout of course drops significantly in such runoffs. In 2008, 3.752,413 votes were cast in “Round 1” of that year’s Senate race (Chambliss, Democrat Jim Martin and Libertarian Allen Buckley). But in the subsequent runoff 4 weeks later, just 2,137,956 returned to the polls–a drop of about 1.6 million votes in about a month. Going way back to 1992, our first general election runoff, about 2.2 million voted in that year’s Senate contest between Democrat Wyche Fowler, Republican Paul Coverdell and Libertarian Jim Hudson. Only 1.2 million returned for the runoff.

  4. saltycracker says:

    I hope the “undecided” and the rest of the “we are over it” that would like to answer their phones will send the Nunn/Carter “children of the corn” away.

    The GOP is getting the “it could be worse” vote as expectations are not much is going to improve in managing revenue, tax reform or continued spending on anything but public infrastructure.

    • Bobloblaw says:

      which is why the poll is likely understating the Dem vote by 1-3 points. But that doesnt get Carter and Nunn to the winner’s circle.

        • Bobloblaw says:

          The Democrats denial of polling is bordering on Nihilism. When Nunn went ahead, I didnt deny it. I thought the Outsourcing ads were working. But clearly they have had an effect but a limited effect. I think nominally Libertarians, due to the closeness of the races, are going GOP. We’ll see if the libertarians get their 4% or if it is more like 2%.

            • Bobloblaw says:

              For this many polls to be wrong would be unprecedented. GA isnt that different than other states. Are pollsters also unable to poll LA? It has 31% black population too. Obama never did better in GA than the polls indicated, neither did Martin.

              Here is the bottom line: Perdue and Deal are ahead. Deal might win it outright, Perdue likely will not. There appear to be a small number of black and women voters voting for Deal but not Perdue (ill bet Kasim Reed is one).

                • Bobloblaw says:

                  Landmark’s final poll is close to this. It is looking more and more like Deal will win outright. Deal is also getting about 2 points more of the black vote than Perdue is, which makes sense.

                  Also Perdue is polling about where Chambliss was in 2008. So somewhere between 49-50%, unlikely to go over 50%.

                  Landmark shows a decline in white support for Dems. Could it be the idiotic Ferguson fliers?

Comments are closed.