New PPP Poll Favors GOP Candidates

Public Policy Polling is out with updated polling numbers on the Georgia races, and all along the line, voters appear to favor the Republican ticket. Governor Deal leads Jason Carter by a 46% to 41% margin, with Andrew Hunt taking 4%. If these numbers were to hold, Deal would win without a runoff. In the Senate race, David Perdue leads Michelle Nunn by two points, 45% to 43%, with 5% voting for Amanda Swafford. Not only is that poll within the margin of error, but it indicates a January runoff, which polling predicts Perdue would win by 3 points.

The GOP is also ahead in down-ballot races. Casey Cagle leads Connie Stokes by a 48/37 margin in the Lt. Governor’s race, with 16% undecided. In the Attorney General race, Sam Olens leads Greg Hecht, 45% to 36% with 19% undecided. The Agriculture Commissioner race between Gary Black and Chris Irvin has the same 45/36/19 split in favor of the incumbent, as does the race for Labor Commissioner between Mark Butler and Robbin Shipp.

In the three way race for Insurance Commissioner, incumbent Ralph Hudgens leads with 41% of the projected vote, followed by Liz Johnson with 34% and Ted Metz with 10%. 15% are undecided. In the Secretary of State race, Brian Kemp leads Doreen Carter 48/39, with 13% undecided. And in the State School Superintendent race, Richard Woods has a 46% to 40% lead over Valarie Wilson, with 14% undecided.

None of the candidates break the 50% approval threshold a month out from the election.

The poll of 895 likely voters was taken between October 2nd and October 5th. It has a 3.3% margin of error. This poll appears to better estimate the likely turnout of women and black voters than other recent tallies. 28% of the sample is African American, and 53% of respondents are women. While both numbers might be a bit low, they are more in line with recent voter turnout results.

There are plenty of other questions and answers in the extensive poll, including a look at some of the possible 2016 presidential candidates. Based on this poll, only Jeb Bush would defeat Hillary Clinton in a hypothetical matchup:

Candidate H. Clinton Opponent Undecided
Jeb Bush 44% 45% 11%
Herman Cain 48% 45% 7%
Chris Christie 46% 41% 13%
Ted Cruz 47% 41% 12%
Newt Gingrich 49% 43% 8%
Mike Huckabee 48% 45% 7%
Rand Paul 47% 44% 9%



  1. jh says:

    Perdue’s dumb ads aren’t working it seems since Nunn is pulling closer while his numbers are stagnant, and this is before the outsourcing deposition.

    That’s odd about Carter falling behind though, same trend as Rasmussen, but against InsiderAdvantage and SurveyUSA.

  2. xdog says:

    Cain and Huckabee pulling 45 percent against hrc? That’s interesting, if not very believable.

    If you’re going there, why not add Alan West and Ben Carson to the list? Hell, Darrell Issa and Michelle Bachmann.

  3. Will Durant says:

    How many people actually complete this poll? Do they count the partials? In the past when I had a land line I would goof around with the IVR polls in retaliation for interrupting my dinner, but 52 questions? I wouldn’t have stuck around with a machine that long even as revenge.

    • Yeah that’s always been unclear to me with PPP. I try to ask no more than 25 questions to any one poll recipient and that is pushing it. Also they conduct 20% of their poll via the internet. Bravo to them for trying to do something different, but if you look at the crosstabs the internet population will be something like 25% undecided for top of ballot races with like 10% going to the libertarian.

      I presume they are using internet to help bolster youth and non-whites that are hard to reach, the problem with this is that the young people who do show up on election day will not be 25% undecided.

  4. John Konop says:

    GOP better take note it is coming soon if not this cycle….Need a new message…..The “Money Ball” guy (Cohn) is not a someone I would bet against. I give Mark credit to be the first pollster to point this out….

    ………..Because “no other plausibly competitive state has seen a more favorable shift for Democrats in the racial composition of eligible voters over the last decade,” and pollsters haven’t quite kept pace with that change, Cohn writes. Citing data from the Georgia secretary of state’s office, he points out that the white share of registered voters has fallen by more than four percentage points just since 2010. White voters comprised 62.6 percent of registered voters four years ago; they comprise 58 percent today.

    In a tweet earlier this week, Cohn pointed out another number that I found startling: “Since Nov 2013, the number of white registered voters in GA has increased by 84,582. Nonwhite: 125,446.” At a pace like that — 40 percent white, 60 percent nonwhite — it doesn’t take long to create a new electorate. By 2016, the impact could be profound……

    • Thanks for the note, John K.

      We at Landmark ( @LandmarkCommGA ) have worked and refined the Georgia voter file most every day for 24 years. We watch trends and compare notes. I would name the names of those who have contributed to this effort, though they may prefer more anonymity on it.

      The trend to a larger black portion of the electorate has rapidly escalated over the past eight years. The other unseen demographic factor is gender: female voters are a decisive majority in statewide elections (since minorities who are registered to vote are also female by nearly a 2-1 margin) . Females this year are more likely to support Democratic candidates.

      National pollsters are also still missing the female vote when they poll Georgia. Why? Because they are missing the **minority** vote (again, it’s 2-1 female). The female vote has *never* been less than 54% of the vote — NEVER — since and including the 1996 election. Yet the national pollsters keep showing 51% or 52% (Rasmussen did this just last week)

      Check this out on the vote by GENDER (you must remove the SOS’ “unknown” reporting category to obtain these numbers):

      Female %

      1996 55.08%

      1998 54.34%

      2000 55.43%

      2002 54.09%

      2004 56.00%

      2006 54.14%

      2008 56.25%

      2010 55.08%

      2012 56.45%

      • seekingtounderstand says:

        Mark: If the Governor would like to announce that he is giving away grants/tax credits to those citizens participating in the Georgia Grown/ Farm to Table programs this might buy votes.
        If he went all in for helping those he is letting out of prison (very important to the black community that they have help getting skills and a new start in life) by giving them a tax break or grant for the Georgia Grown/ Farm to Table, Organic wellness movement and tourism he could buy a lot of votes. Savannah is booming due to their ability to create a great place to visit. They are inspirational and creative people.
        Gary Black and his team are doing good work but its a very small budget. We are an ag state and its vital that we get jobs going. Feeding the World with the new Savannah port widening is a great way to go.
        I hope those that are well connected know that if they do not start addressing important issues for all of GA, their time in power is very limited.

      • seekingtounderstand says:

        If you running Gov. Deals campaign start highlighting what he has done and the future. Stop responding to his critics. Carter has nothing to offer after watching the debate last night.

      • benevolus says:

        Saw this on Jim Galloway’s column:
        “By way of verification,’s Nate Silver has this gradation of pollsters across the nation.
        Survey USA received an ‘A’. The firm used by the AJC, Abt SRBI, was given a ‘B+’. Public Policy Polling of North Carolina was graded at ‘B-‘. Landmark Communications got a ‘C+,’ and InsiderAdvantage was given a ‘D.’ ”

        I don’t really have a comment, thought others might.

        • I got a bad grade on there but there are only using 3 polls and they are counting the worst one 3 times, so like A+, A+, F, F, F basically when it should be A+, A+, F.

        • Benevolus, that’s a good question. The ratings provided by 538 were impacted by the ‘type’ of polling that is done — not accuracy per se. Their ratings were impacted by factors such as whether live calling is used, cell phones, as well as whether a firm has joined a membership in polling organizations, etc. –Things that are not about accuracy, but other factors.

          Of the 337 polling firms listed in the United States, only 91 had a better grade than ours, and many of those grades were apparently impacted due to the type of polling (538 likes live calling, membership in polling groups, etc).

          Our firm was in some pretty good company on this list, as well: the same grade as Gallup, YouGov, US News and World Report, Fox News/Opinion Dynamics Corp, Penn Schoen, Public Opinion Strategies, and more.

          What was odd about their rating system is that companies with just 1 published poll *ever* could be given an A or B, which is silly. In fact, 30 of the 90 polling firms listed had fewer than five (5) published polls — ever.

          The idea of a rating system is good. But if I was rating their system in return, I’d have to give it a C.

      • John Konop says:


        Since you are the expert I have a question 🙂 How do you guys account for the increase in interracial couples, marriages…..Would that increase the white vote at all for Democrats? Is their an increase of interracial people not accounted for in the polling via not being identified?

        …….Survey found more than a third of adults (35%) say they have a family member who is married to someone of a different race…….

        ……..Blacks are now substantially more likely than before to marry whites.[3] In 2010, 15% of new marriages were interracial.[3] In 2010, 25% of Asians, 25% of Hispanics, 17.1% of blacks, and 9.4% of whites married interracially.[3] Of the 275,500 new interracial marriages in 2010, 43% were white-Hispanic couples, 14.4% were white-Asian, 11.9% were white-black, and the remainder were other combinations……..

        • jh says:

          Off topic, but I’ve noticed in liberal Midtown, I don’t see much interracial dating, besides with gays. In diverse middle class Pleasant Hill Duluth, interracial dating is very common. Not so much in Suwanee or other majority white or majority areas.

    • And I have criticized them for using Nov 2013 as a baseline. Best to look at the differences between Nov 2010 and now.

      They are much, much greater.

      From Nov 2010 to Sep 1 2014…
      White = -131,469
      Black: = +63,816
      Hisp = +15,571
      Other/Un = +138,880

          • NoTeabagging says:

            I re-registered specifically to remove race on my Voter registration and I’d do it in a heartbeat on Census forms. It is trending for those concerned about privacy.

            Although I was recently wrist slapped on PP, that such info never surfaced in official redistricting data *cough,cough*, you can see it is the bread and butter of Political Campaign consultants.

        • Very rough figures:

          “Unknown” isolated as a category was 5% of the vote in 2012; 3% of the vote in 2010.

          Known non-black minority turnout was 2% in 2010, 3% in 2012 (Asian and Hispanic combined, plus voters who self-describe their race as Indian or other known, non-black minorities).

          So I assume the “Other” designation included both of these categories though I can’t say it with certitude.

            • It’s such a small target group with such a bad phone number match that the truth is no one knows for sure — they will self-describe their race if asked in a poll, so you’d have to base the call off of an actual voter file and not just do open dial calling.

              That said, I’d estimate that it’s 2-1 Democratic. About 1/3rd of the list is likely white (we do research on the file, and that’s our estimate based on a series of factors) and 2/3rds non-white. Of the non -whites, perhaps 50-60% are black and the rest are non-black minority.

              You can look at it one of two ways: it’s either a a 50+% increase from 2010 or the vote goes down by 40% in the non-Pres years.

              Again, the unknowns are still black or white or hispanic or asian or some other race, so on a poll when describing themselves they can’t respond “I’m unknown based upon the SOS records…” so, especially when combined with the bad phone matches (very high percentage of this group is cell phone based only), there’s not a really great way to know the deep details. But our estimate is 2-1 Democratic, give or take.

              • John Konop says:

                One could assume as the number grows it will favor Dems…. The trend shows growth….Which is a hidden boggy for Dems….In a tight race could mean something….Once again thanks for your honesty….The problem with the GOP is if you do not truly analysis the numbers you cannot come up with a real strategy….This is the number one reasons businesses fail they lie to themselves….If I was any of the candidates I would CALL your firm ASAP….so you could give them reality shots over Kool-Aid….not only are you guys smart….you shoot it straight….may not what they want to hear…but very important for survival…. Charlie Harper is right the GOP needs guys like you…because the party does not understand it is no longer an internal fight….It is now a general election…a way different game with many rule changes….I would have you and Charlie on my speed dial if I was running 🙂

  5. jh says:!/vizhome/VotersaddedafterMarch12014/Sheet2

    Adding to Mark, AJC reports

    “More 212,000 have been added to voter rolls so far this year as the Nov. 4 midterm election approaches. On average, a county election office may see 50 to 70 pending voter applications, said Chris Harvey, the chief investigator for the Secretary of State’s Office. This year, it’s 2,000 to 3,000, which Harvey called “uncharacteristically high.”

Comments are closed.