InsiderAdvantage Poll: Perdue and Deal Lead

UPDATED – There have been some issues with the crosstabs for this poll.  The original crosstabs were removed from the IA site as of about 5 p.m. due to an apparent clerical error made while entering data into a template.  They have since been replaced and the links below have been updated.

A poll released today by InsiderAdvantage/Fox5/Morris News shows a tight Governor’s race with Gov. Deal up 44%-40%.   The margin of error is 2.9% and the confidence level is 95%.

It also shows a wide gap in the U.S. Senate race with Perdue up 50%-40% and winning without a runoff.  This margin is 6 points higher than the AJC poll released earlier today.

This poll puts the African-American vote at 33%, much higher than the 24% used by the AJC poll.

Further analysis can be seen at the InsiderAdvantage site.   The updated crosstabs can hopefully be found here.  The original crosstabs, which were discussed in the earlier comments, can be found here.

On the issues with the original crosstabs, InsiderAdvantage states:

“In producing the cross tabs for our Georgia poll the numbers were inserted into a set template. The numbers for several of the demographics were the older numbers and have been updated. This does not impact the weighted numbers which were established off of the updated numbers.”

Now that both polls have been released, it is time to discuss/argue about which one is more accurate and what may change before the next poll is done.

Governor IA AJC
Nathan Deal 44% 43%
Jason Carter 40% 42%
Andrew Hunt 7% 7%
Undecided 9%


U.S. Senate  IA AJC
David Perdue 50% 45%
Michelle Nunn 40% 41%
Amanda Swafford 5% 6%
Undecided 5%

Insider Advantage further requested that we add append the below statement to this post:

“Today InsiderAdvantage released a poll for Fox5 and the Morris Newspapers which contained the accurate results, but which unfortunately showed a few demographic categories which reflected the raw numbers from a prior poll. These raw numbers were not utilized in the actual cross tabulation of the data. When translating the new poll to a more readable format, some of these older numbers were left in the template by mistake. While this did not impact the results, we wanted to clarify this to your readers. We appreciate you posting this.”
.Matt Towery, CEO InsiderAdvantage


  1. It is LAUGHABLE to put out a poll with 33% black that a Republican leads by 10 in.

    It is not worth discussing it. Matt Towery/IA just make up their #s and they should release their actual call logs or call center invoices because this BS should be illegal.

    Fox 5 – I know people read this site. This is pitiful. You should be ashamed of yourself.

    • Bobloblaw says:

      Im not a big fan of IA. They were the only ones in 2006 who had the Govs race close and in 2008 Towery said the Sen runoff would be a nail biter….oh well. AJC numbers are consistent with everything else.

      The libertarian vote is too high in both races. Ill bet 3-4%. Outside of maybe MT, no Lib gets 7%. I cant see the Libs doing better with no more Boortz around.

      • Bobloblaw says:

        well that is how it is offset. 33% black must mean then that the % white for the D would have to be low. They cancel each other out. I still dont buy IA

        • jh says:

          That makes no sense. What does black turnout have anything to do with white % for D, unless you think that racist Democrats seeing lots of black people voting means they have to vote the other way?

          • Bobloblaw says:

            They have nothing to do with one another but if IA says it will be 33% black that helps the Dems. Meanwhile the % white for Dems to too low which hurts Dems. So the total Dem vote is right, but for the wrong reason.

  2. ryanhawk says:

    I expect the steady drumbeat of news about nursing home reimbursements, campaign donations, and regulatory appointments to put a dent in Deal’s numbers.

  3. MAYBE someone at Fox 5 can explain to me how…
    522 males
    438 females
    1,167 respondents.

    On multiple calculators I’ve checked with, it equals 960.

    SHAMEFUL. Anyone who publishes a Towery poll – how do you sleep at night?

    • David C says:

      Not to mention

      274 18-29 yr olds
      311 30-44 yr olds
      465 45-64 yr olds
      177 65+ yr olds

      1167 (It equals 1,227, somehow 60 more people than the base total)

      • It is actually possible for your raw weights (as that is sometimes reported) to equal more than your weighted number. For example, too many 65+ answer the phone and not enough 30-. So you have to discard or weigh down some of your 65+.

        So it would be OK to for example call 1,100 people and weigh that down to 1,000 weighted respondents. In some ways it would make your poll a little better than it’s overall margin of error (if for example you have too many white completes and you combine them into a smaller number, the margin of error of that subsample is actually a little lower than what is being baked into the overall number).

        It is never OK to call 950 people and weigh it up to 1,150.

  4. Noway says:

    Chris, you’re talking inside baseball kinda stuff. What do you see as the realistic percentages for each candidate for both guv and senate?

  5. Will Durant says:

    “The crosstabs were removed from their site as of about 5 p.m. due to an apparent clerical error made while entering data into a template.”

    And this is what they previously wanted you to pay to see? Did the “clerical errors” just affect the published crosstabs or also the published poll percentages? Somebody’s got some ‘splainin to do.

    Congrats to the guys here for noticing their slip was showing. If they can’t add I’m not too confident of their multiplication and division skills either. Statistical analysis? Fuggitaboutit.

    • Lawton Sack says:

      From IA:

      “In producing the cross tabs for our Georgia poll the numbers were inserted into a set template. The numbers for several of the demographics were the older numbers and have been updated. This does not impact the weighted numbers which were established off of the updated numbers.”

  6. Bobloblaw says:

    It is clear now that Deal and Perdue arent suffer from changing demographics. All the talk about changing demographics isnt the reason. The entire Dem ticket is polling no better than they were in 2010. The downticket GOP races are at 50% or better. Deal and Perdue are suffer from ugly primaries. There are likely TP Deadender types that are angry that Broun and Pennington didnt win and they arent yet supporting Deal and Perdue.

    If Changing demographics were the reason we’d be seeing a steady improvement in the Dem share of the vote. Isnt happening. We’d also see Cagle, Olens and Kemp suffering. They arent, they are cruising.

      • Bobloblaw says:

        So you tell me why the down ticket GOP is doing fine but Deal and Perdue arent. It isnt demographics and Nunn and Carter arent doing any better than past Dems or the downticket Dems. And if it is demographics, are the down ticket GOP getting 25-30% of the black vote?

        • xdog says:

          Bob, they’ve answered that question before. I’ve read their responses and they make sense to me. Of course you don’t have to believe them but imo it would boost your credibility and show evidence you’re paying attention if you stopped acting as if you’re breaking new ground every time you ask the same question.

        • jh says:

          Thurmond was polling low 30s consistently, Nunn is polling low 40s consistently.

          Barnes polled 38-40%, and only lead in 1 poll. Carter is polling around 44%, and has lead a third of the polls. You are being intentionally obtuse.

        • taylor says:

          The GOP ticket looks better in the downticket races because they have name recognition and the Dems don’t. I would think that’s obvious.

          I’ve seen multiple Nunn and Carter ads today. Haven’t seen a Stokes or Hecht ad yet. And while Cagle and Olens aren’t up, their incumbents with name recognition.

          • Bobloblaw says:

            None of you have refuted my theory.

            Why is it impossible to believe that a contentious primary is the reason why Perdue and Deal are struggling? Do you guys not know who it was who voted for Pennington and Broun? It was “My Way Or The Highway” conservatives. I am seeing the same thing in some other Senate races were TP took on establishment candidates. In KS and NC. In SC Lindsey Graham is running 7 points behind Tim Scott. Graham had a primary, Scott didnt.

            You cant say that Demographics are why Deal and Perdue are struggling but name recognition is why the down ticket GOP is doing ok.

            Also you cant compare Thurmond’s polling to Nunn’s or Carter’s, neither is going to lose by 20. As for Barnes, I dont think Deal and Perdue will win by 10. But there is s consistency to the Dem ticket. It is all polling in the low to mid 40s (name recognition doesnt seem to matter), while on the GOP side the down ticket is polling 5-7 points better.

            • John Konop says:

              You cannot ignore the changing demographics….In Georgia more minorities are voting and women are fleeing. The GOP can look in the mirror or pretend it is not happening….Historically this should be a GOP landslide….unpopular lame duck presidents help the other party….you are a smart guy, take your blinders off….The Phil Kent side of the GOP is a death march….

            • Adam Griffin says:

              Everyone has refuted your “theory”.

              In regards to the disgruntled Pennington/Broun supporters holding out: it doesn’t work that way. These people you speak of would be declaring themselves undecided right now; they wouldn’t be claiming to support the opposite party, only to flip back right before the election. The nail in the coffin is the fact that undecideds at this point in the cycle – across an aggregate of polling – are actually a smaller percentage of those polled than they were in 2008, 2010 and 2012. There is no psephological basis to your “theory”.

              The issue in Kansas stems not from a primary challenge in the Senate race, but two separate matters. Brownback has screwed up the state’s affairs so royally that his toxicity is now flowing across the entire ticket. In addition to that, a strong independent candidate has emerged that is pulling even more support than the Democrat was – which forced the Democrat to drop out of the race.

              And yes, you can say that demographics are having an effect at the same time that name recognition is pulling one candidate or another ahead, because that’s what is happening. Nunn and Carter are doing so well because a) they have name recognition, b) they’re sucking up money like vacuum cleaners, which allows them to plaster themselves all over TV, radio, the internet and other sources, and c) they’re able to tap into white sentiment without sacrificing non-white support. Demographics are changing, but a generic, unknown candidate for Governor or Senator wouldn’t be doing as well as either of them due to the name recognition factor and the ability to tap into old money. These legacy candidates are in effect giving us 2018 conditions for a generic candidate in 2014.

              The Republican down-ticket races are advantage Republican, because a) there is a much bigger discrepancy between the Republicans’ name recognition and that of the Democrats’, b) the Democratic candidates for these offices are relatively weak, and c) all of the Democratic money is being sucked up by Carter and Nunn, which locks these Democratic candidates into said situation of being unknown, generic commodities.

              And you said that the entire Dem ticket is performing no better than it was in 2010, but yet “you cant compare Thurmond’s polling to Nunn’s or Carter’s, neither is going to lose by 20”. Not only are none of these candidates polling as terribly as the 2010 slate of candidates were (even the worst-performing candidate in the race, Doreen Carter, is sitting in polling – with up to 10% undecided – at levels comparable to the average Democratic performers on Election Day in 2010), but they are also going to outperform the election results for each office when compared to 2010. This part is being fueled by the demographic trend. Everything else – for both Republican and Democrat in Georgia – is being fueled by name recognition and money advantages.

              • Bobloblaw says:

                The name recognition argument is bogus. If it was the case we’d see Nunn and Carter doing better then the downticket Dems, but they arent doing better. They are doing roughly the same.

                The demographics argument is bogus unless you can prove that the downticket GOP is getting 20% or more of the black vote while the upticket is getting 5%.
                Prove it and Ill accept it.

                Also I noticed no one is predicting black turnout at 2008 levels for 2014 when Obama got 47% and Martin got 46%.

                I do think ethics is hurting Deal (though he wasnt ethical in 2010), but is isnt hurting anyone else and thus doesnt explain Perdue’s weaker numbers.

                I think the poll we need to see is how much of the “conservative” vote each candidate on the GOP side is getting. My guess is the upticket GOP doesnt have as much “conservative” support as the downticket does.

                I never said Pennington/Broun supporters would support the other ticket. But they might not vote in those races as at. The posters here really have no clue as to just how angry the far right is at the GOP Establishment. Go visit the cesspool know as Freerepublic or even NRO. The posters there are so angry that they are rooting for Grimes over McConnell even if it costs the GOP the Senate. The TP is no longer a political movement like it was in 2010. They are political assassins. Notice the TP isnt attacking Dems anymore. They are spending their resources going after GOP incumbents. I can tell you what the main issue is. It is Amnesty. Perdue is not trusted by people who make being against Amnesty their number one priority. Where as Cagle, Olens and Kemp arent in offices where the issue is a main concern.

                In KS Brownback’s problems arent why Roberts, who never got less than 60%, is polling under 40%. Milton Wolf supporters arent supporting Roberts. In SC Graham is polling under 50, while Scott is polling mid 50s. Hmmmm why is that??? Think Scott is getting the black vote….not really. Scott didnt have a primary is the reason why he is polling 5-7 points better than Graham.

                Youre saying is hat ugly primaries dont impact the general election. But parties seek to avoid ugly primaries for the reason we’re seeing in the top two GA races now.

                So who is it that is supporting Cagle, Kemp and Olens but is undecided in Perdue and Deal races???? There is somewhere between a 5-8 point difference in their polling numbers. These voters arent supporting Nunn and Carter. They are supporting no one at the moment.

                The only explanation is angry conservatives in the Senate races and angry conservatives plus ethics in the Govs races (Perdue is polling ahead of Deal in every poll).

              • georgiahack says:

                There is no engaging him – willful ignorance is just not something you can argue. He shows up on PP, starts throwing around his political “critiques,” claims victory when no one engages him, and then repeats it all on another thread. The guy claims to have such political acumen that he knows he is correct and that the rest of the folks who actually do this for a living just don’t get it. This is the same guy how had to ask what the PSC was last week. Not worth your time.

                • Bobloblaw says:

                  The PSC is an obscure position

                  I deal in facts:

                  1. The bottom of the GOP ticket is at the 50% or better level.
                  2. The top of the GOP ticket is polling 5-7 points less.
                  3. The entire Dem ticket is polling about the same level.
                  4. The top races on the GOP side had contentious primaries.

                  From this you cannot conclude that name recognition is helping the bottom of the GOP ticket when the bottom of the Dem ticket is polling the same as the top of the Dem ticket. We would expect the bottom Dems to poll worse.

                  You also cannot conclude that demographics are the reason for poor GOP top ticket performance unless you also can show the bottom of the GOP ticket is doing much better with minorities.

                  The only conclusion is a bad primary. And in Deal’s case ethics is costing him 2-3 points off Perdue’s numbers.

                  • Will Durant says:

                    Polls are not facts.

                    In the long run as a Georgia consumer, whoever is on the PSC will decide a lot more about your money than the Superintendent of Schools or even a Lt. Governor.

                  • georgiahack says:

                    The PSC is not that obscure. They effect your life and everyone else life EVERY DAY. That being said, no one on this blog would argue that the PSC is not 100% known to the general population, but there isn’t a politico in this state who does not know what it is. Then you came along…. The new kid in the class that likes to show how much smarter he is than everyone else and just ends up being the lonely annoying one. Until they learn to play along.

                    Hey, I disagree with folks here all the time. Sometimes they are right, sometimes I am right. We all have our own take on facts, but you just keep on making up crap (i.e. some of your supposed facts) that professionals, from both sides, have politely tried to show you how wrong you are.

                    Folks come to this site because there is no better place to get inside gossip, talk about the races, see the news you missed when scanning the headlines, and engage other like-minded (read political) individuals in the comments. We go to the political insider to read the articles, but stay away from the comment section because it is complete trash. It is full of people who argue like you do. It’s not fun. It’s not engaging. It plays to the lowest common denominator and adds nothing to the debate. Step up.

  7. Let me just point out one other giant huge thing about this poll that makes me call BS.

    Georgia’s electorate has been becoming rapidly more diverse, and as you would expect, the percentage of the electorate that is white grows as the age of the electorate gets older. Overall 29.8% of active registered voters as of 8/1/2014 are black. But in the 18-24 age group, that number is actually 36%, and as you go up the ladder the number goes down.

    Similarly, among all voters, the white % of registration is 58.5%. Yet among 18-24 year olds, it is only 44%, and as the voter gets older, they are more likely to be whiter.

    By the time you get to 65+, active voters by registration are only 20% black and they are 74% white. In other words, all things being equal, the 65+ electorate should be significantly more white and more Republican. Among those who are actually likely to vote, 77% white, 19% black.

    Compare that to an overall number for projected turnout of about 64% white and 29% black.

    Now in the Towery poll, here is how the vote among 65+ compares to the overall #…
    Nunn – 39.8 (overall 39.8)
    Perdue – 49.6 (overall 50.1).

    That’s right – Perdue is actually doing better among voters under 65 than he is among those who are 65+. We are to believe Nunn is doing the same.

    Carter – 40.1 (overall 39.9)
    Deal – 45.6 (overall 44.4)

    Again – Carter and Deal roughly doing the same with the 65+ set than they are among all voters overall.

    I’m sorry – this is impossible.

    • Adam Griffin says:

      Not to disagree with your broader premise of this poll being bogus, but Georgia’s polarity is much more reflected across racial lines, and barely reflected (based on what little exit polling we have available from the past 10 years) across age groups.

      For example, the difference between 18-30 year-olds and 65+ in the 2008 presidential was 3 points (51% and 54% McCain, respectively). The same difference existed in the Martin-Chambliss matchup (47% and 50% Chambliss, respectively). The difference in the 2006 Governor’s race was a bit starker (51% and 59% Perdue, respectively), while there was a 12-point difference in the 2004 presidential (52% and 64% Bush, respectively).

      Now, the 2014 result may end up being more reflective of 2004/2006 in terms of contrast between 18-30 and 65+, partially due to the (lack of) racial makeup of candidates and the ever-changing demographics that are widening the racial difference between young and old. I would think, however, that we would have seen a widening in the 2008 exit polling between these two groups when compared to 04/06 (due to the demographic trend) rather than a tightening, unless younger white voters are becoming even more Republican than their grandparents. Maybe there’s a bit of truth to the fact that Perdue is doing slightly better among younger voters than he is among the older groups.

      Personally, I tend to discredit polls that show vast differences in candidate preference by age group, as recent history suggests there shouldn’t be much difference.

      • Exit polls certainly aren’t perfect though. I looked at my 2012 polling which was very accurate and there was a clear trend – Democrats winning younger voters, Republicans winning older voters. Just because an exit poll didn’t pick it up, doesn’t mean much.

  8. And to give some further ammo – although the AJC poll is poorly weighted for likely voters by race, I have confidence that they actually do the calls, unlike Towery/IA who I believe makes them up (or inserts fake results to pad the few calls they actually do).

    Here is how the vote for Senate compares by age in the AJC poll to the Towery poll.
    18-39: 45% Nunn, 28% Perdue
    40-64: 40% Nunn, 43% Perdue
    65+: 30% Nunn, 57% Perdue

    Look at how Perdue’s number goes up as voters get older (and whiter) and Nunn’s number goes down.

    18-29: 46% Nunn, 32% Perdue
    30-44: 38% Nunn, 60% Perdue (60 are you kidding me)
    45-64: 37% Nunn, 51% Perdue
    65+: 40% Nunn, 50% Perdue

    So – according to Towery’s poll, the numbers are basically flat among the age groups and Nunn actually does best among 65+ when you look at all voters over 30?


  9. craigh14 says:

    Deal I’m not sure about yet. Per due, yes please! I’m so tired of the fluff coming from Nunn. There’s no substance.

Comments are closed.