According to the recent poll by Verne Kennedy of Marketing Research Institute (MRI) of Gulf Breeze Florida, he is. Absolutely not, says the Vitter’s campaign and some of his followers--the Kennedy poll simply cannot be believed.
Welcome to the battle of the polls.
On Friday, the Kennedy released his latest poll. Various Newspapers across the state were abuzz over the prospects of a remarkable comeback by Angelle, who only months ago was in polling in single digits.
In trying to report this interesting development and the various issues circulating around the story, I felt it necessary to investigate, talk with the pollster and to those contesting the results and decided to abandon the strict journalistic narrative to present my own opinions in the format if I were being interviewed.
So here’s my questions and my responses. Enjoy the the Q&A’s and let’s hear back from you:
WHO IS CURRENTLY LEADING IN THE LOUISIANA GUBERNATORIAL ELECTION?
Honestly, that is hard to say at this time. We have seen a number of polls in the past few months and in particular, over the last couple of weeks. The one that came out this past Friday was commissioned by the Advocate owner John Georges and roughly 15 and 20 business persons and was conducted by Verne Kennedy. That poll indicates, based upon the polling responses, that Vitter, right now, has topped out and has been overtaken by Angelle. This poll claims Angelle leads the four-person race with 25 points, Vitter comes in second with 21, John Bel Edwards closely follows with 20 and Jay Dardenne trails the rest with 12%. By the way, “undecided’s” are very competitive with 21%.
However, the polling doesn’t end there. After I received and reviewed the poll, I talked with the pollster, Kennedy. He explained that since this election is statewide, after obtaining the raw numbers and percentages, he redistributed those results based upon traditional African-American voting. Kennedy says Blacks support Democrats with 90% of the vote, traditionally. The result, as you see in this chart below, John Bel Edwards skyrockets to 34%, Angelle and Vitter tie at 21%, Jay Dardenne remains at 12 and those “undecided” drop to 12. However, note the* at the bottom of the chart’s slide, the undecided are white voters. Thus, in my view, one would assume that they might favor the republican.
HOW DO THESE NUMBERS COMPARE TO THE PAST VERNE KENNEDY POLL
Unquestionably, the trend in these Kennedy shows Scott Angelle gaining significantly. For example, as you can see, in the June poll (below), before the African-American redistribution, David Vitter and Edwards were neck and neck with 28 to 27% respectively. Scott Angelle received 15% of the poll tally. From other polls, he started out in single digits. After the African-American redistribution, Edwards climbed 11 points to 38%, Vitter dropped slightly to 27 and Angelle drop two points to thirteen.
WOW, IF ACCURATE, THE DIFFERENCE BETWEEN JUNE AND JULY NUMBERS ARE SIGNIFICANT.
Indeed. Looking at the numbers before the redistribution, Scott Angelle jumped 10 percentage points and David Vitter plummeted by seven. After the African-American distribution, Vitter and Angelle are tied at 21% and John Bel Edwards again spikes, this time by 11 points to 38 percent.
SO WHAT WOULD A RUNOFF SCENARIO LOOK LIKE?
First, all we know right now would be the poll results from a Vitter-Edwards match. According to Verne Kennedy’s poll, which is below, the Democrat John Bel Edwards would lead Vitter 46 to 41 and after the African American redistribution he would likewise be leading 42 to 38. Also, the undecided votes would go to 20. So, the redistribution shaves off four from Edwards, three from Vitter and nine from the “undecided”. Given the fact that Kennedy assigns a 4% plus or minus, you can argue that Vitter and Edwards are in a statistical tie. You can also argue that theoretically, Edwards could be up to 46 and Vitter down to 34 if you give Edwards those four points. In my opinion, what is significant however is the undecided numbers, which after the African-American redistribution, makes up 1/5 of the remaining vote. Personally, it would seem logical to me that Vitter would get at least the majority of the vote, but how much of a majority is uncertain. Therefore, the actual margin of error would be important if Vitter would get a good part of the undecided vote. Of course, much depends upon voter turnout. Kennedy has factored in black turnout at 26% which he says is conservative. Regardless, at this point in time, assuming the accuracy of the African American redistribution and other factors, based upon this poll, the election could be tight.
SO, WHAT ABOUT A DIFFERENT RUN-OFF SCENARIO, SAY BETWEEN ANGELLE AND VITTER OR EDWARDS AND ANGELLE?
Kennedy said he did not anticipate the dramatic gain by Angelle, therefore, did not question or poll those matches so right now, it would all be total speculation.
INDEED. NOW I WOULD ASSUME THAT THE DARDENNE CAMP AND ESPECIALLY THE VITTER CAMP AND SUPPORTERS WOULD BE DISPUTING THESE RESULTS, AM I CORRECT?
Absolutely. Dardenne has been quoted to say that his campaigns’ internal numbers show him in a much more favorable light. The Vitter supporters and campaign are taking issue with this poll. First, they question the historical accuracy of the pollster. Second, they question these poll numbers and therefore, his conclusions.
HOW ARE THEY QUESTIONING HIS ACCURACY AS A POLLSTER, IN GENERAL?
They cite one of his polls with then State Representative, African American Democrat, Cedric Richmond vs. former Republican Congressman Joseph Cao and claim that Kennedy had Richmond losing to Cao by 15 points in a African American dominated Congressional district. Richmond won by over thirty. They also point to another poll, actually, the US Senate race in 2010 against Vitter. They say that a Kennedy poll initially had Chet Traylor only ten points behind David Vitter but Vitter beat Traylor by 80 points in the primary.
OK. HOW ARE THEY QUESTIONING HIS ACCURACY IN THIS POLL?
They are pointing to other polls that tell a different story. On Friday, I received a response from a prior email from Vitter’s campaign manager, Kyle Ruckert. By e-communications, we discussed the poll. First, Ruckert sent me this chart that he apparently created:
His point was well taken. The next morning, by phone, I discussed with Kennedy, the poll and the objections. He responded that the first months of the Angelle ad buys were introductory to get him name/face exposure and that the recent July month, when there was a significant drop in TV commercials, but the ads reinforced Angelle’s presence. In addition, Kennedy said the month of July was marked by a sharp decline in Vitter’s favorable. He said, normally when voters move off of a candidate, they fall into the uncertain category. Kennedy noted that for some reason, the normal switch from “candidate-to-uncertain” did not occur. Instead, his poll revealed they moved directly from Vitter to Angelle without going to undecided.
IS THAT THE ONLY POINT OF DISAGREEMENT?
Not at all. On Sunday, I received more information from Ruckert. He took issue with the basis of the calculations. This is what he wrote to me:
On point, he said, “ 2 other points in Verns survey which were incorrect and led to false result: Verne's poll under samples the New Orleans media market by 6 points, Vitter strong hold. In Verne's poll, NOLA represented 28% of respondents. In the 2014 US Senate election, NOLA represented 34% of the electorate. In the 2012 Presidential, NOLA represented 34.1 and in 2010 it represented 34.7%.
Shocker, Verne's poll also over sampled Lafayette by 3 points...Angelle's stronghold. In Verne's poll, Lafayette represented 17% of respondents. In the 2014 US Senate election, Lafayette represented 14% of the electorate. In the 2012 Presidential, Lafayette represented 14.5 and in 2010 it represented 14%.
Well, I then forwarded those comments to Kennedy for his elucidation. To those points, his response said:
“It depends on how they define media markets. Our survey is based on media markets as we have traditionally used, but more importantly every parish in our survey is balanced based on its voter turnout, much more accurate that any definition of media markets. WE take parishes and put them into media markets after the survey is completed. Our survey is balanced by parish voting , much more accurate than media markets which differ in how one defines them. In the morning I shall be glad to send you our survey population by parish. Media markets are not important if you balance your sample by parish.
Steve, their whole attack that Vitter launched is so full of deceptions that it is hardly worth answering. Some of the email blast they send out Saturday takes results that are not in any survey I have conducted, made up data. In other cases, I think they quote a survey taken as long as 18 months before an election and compare it to election results. Don’t be deceived by their attack. They are only attacking because they know my survey is accurate and don’t want voters to know the truth. In one case they took a question which measured Mayor Landrieu’s strongly committed voters and claimed that I said that is all the vote he would get. David Vitter is losing support each month and he fears that if voters know that even more will abandon him.
Even if their claim were accurate, which it is not, Vitter is still losing support. Why would they spend so much time and energy in attacking one survey unless they know it tells the truth which they do not want out. Consider the fact that we used the same definition of media markets in every survey, May, June, and July, and things changed with Vitter losing support. Our definitions stay the same but the results changed. Regardless of media market definitions, Vitter lost support.
Ask them to send you their definition of media markets based on parishes. I can then compare it to our definition.
Update: 12:56 PM
"I did just receive data back on the New Orleans Media market and it was balanced exactly based on the 2011 election at 28.1% of statewide vote. That election we believe is a more accurate because it has the same races we have this year. I should receive Lafayette later and shall send you the results. "
The latest issue is the claim by the Vitter campaign that the survey under represented the Orleans Media Area by 6% and over represented Lafayette by 3% based on the 2014 U.S. Senate Election. Compared to the 2011 election which is exactly the same as this year with all statewide officials and the entire legislature up for vote, Orleans had 28.1% of respondents, exactly like the 28% we surveyed. Because more voters in the Lafayette area told us they were definitely going to vote in the upcoming primary, Lafayette was over represented compared to 2011 by 2%. I contend that the adjustment for Lafayette was the correct thing to do, but even if Lafayette was 2% lower, ballot results would have shown Angelle one half percent lower than the survey depicts. Vitter’s numbers would not have changed.
OMG. THIS IS GETTING MORE INTERESTING THAN THE ACTUAL CANDIDATES’ DEBATES.
Yep. Afterwards, I received another email from Ruckert.
“Also, among african americans Vern has Angelle at 21, JBE at 37 and Vitter at 8. No republican has come close to double digits in a competitive race with a democrat on the ballot. Angelle should be in the 5 range. Other polls I have seen have all the Rs in single digits and JBE anywhere from 51-67 with the AA vote.
I forwarded this comment to Kennedy this morning for his response and here is his (temporary) response:
I forget that my data processing company is in CA and they just opened so I shall get the data I need from them later this afternoon. I do not believe the Vitter criticism is correct and only considers 2014 not 2011 which is the identical race we have this year, but even if they were correct I calculated what it would mean. It would add 1% to Vitter and take away 1% from Angelle. Get back with later todayl
THANKS. YOU ALSO MENTIONED OTHER POLLS THAT CONTRADICT KENNEDY’S. COULD YOU EXPLAIN FURTHER?
Sure. Ruckert pointed out two polls conducted in mid and the late July 2015. Those polls claim that Vitter is holding strong but steady. The polls, conducted by a group named Triumph Campaigns linked up with conservative publication The Hayride to conduct polls. The one in mid-July had Vitter ahead of Jon Bel Edwards 32 to 31 yet nineteen points ahead of Angelle and 20 percentage points in front of Jay Dardenne. The second Triumph-The Hayride poll has Vitter at 33, Edwards at 32, Angelle at 12 and Dardenne at 11. So, you can see there are some significant discrepancies between the Kennedy and the Triumph Campaign polls.
SO THE BIG QUESTION, WHICH ARE YOU SAYING ARE THE MOST CORRECT?
I can’t say. I know that Triumph is an electronic poll that various pollsters believe are not as reliable as those with real people calling the voters. Then there is the bias issue. Verne Kennedy does polling for John Georges and a group of roughly 15 to 20 other politically-oriented “movers and shakers” business people who want polling insights. Part of the problem is some critics of the Kennedy poll claim that Georges is considering jumping into the race and so therefore the Kennedy poll is biased. On the other hand, I have heard others say that the Hayride Triumph operation favors Vitter. In my view, other than this most recent sharp conflict in the Triumph vs. Kennedy polls of approximately twenty-percent, most of the polls results, so far, have been pretty close to one another. I think one should note that the Hayride Triumph poll shows that Vitter is in low thirties. Regardless of Angelle’s actual numbers which are in dispute, this is not that great for a person with almost universal name recognition. But, if anything, the low numbers also downplay the bias charge against the Hayride Triumph poll. Last, the difference between the Kennedy and the Triumph poll is really over ten percent with margin of errors of 4 percent in the Kennedy and 2 percent in the Triumph poll. Ten percent is not really a lot. Let’s say Vitter dropped ten points as the Kennedy poll claims, well, essentially as Kennedy’s poll would indicate, that ten points is made up by Angelle.
Other points of interest. The Triumph poll does not include any cross tabs, so, it is difficult to deep drill into and criticize, as the Vitter Campaign has done with the Kennedy poll. Also, Ruckert is citing another poll, this one by MarblePort, whose proprietor has been closely associated with Bill Cassidy, who was strongly supported by Vitter last year in the US Senate race. This one, called The Hayride MarblePortOne, which was published last week has no supporting information or original documentation to make any comparisons. That poll, reportedly taken early last week has Vitter at 31.1, Edwards 30.6, Angelle 14.1, Dardenne 13.4 and undecided 10.8%.
For what it is worth, The Hayride states:
“Vitter has shed three points while everyone else has gained a little, but the problem for Angelle and Dardenne is that neither has been able to substantially gain on the other – and thus they are currently cancelling out each other’s chance to push into the runoff.
So, honestly, in many respects, we are talking comparing apples to oranges to pears. We don’t have enough information to truly make accurate comparisons.
AND WHAT ABOUT THE KNOCK AGAINST KENNEDY AND GEORGES?
Well, I can say this—based upon my conversation with various people about the poll, the other business people are paying for and participating in the polling group, it is not just Georges. There are a number of business people, some favoring various candidates who have ponied up to obtain this information from Kennedy. In fact, already, we had two separate incidents over the past couple of months in which members favoring different candidates appear to have selected certain information favorable to their candidates and this information was released to the public. The poll was not intended to be released to the public. So, while Kennedy surely is Georges’ pollster, given the fact there are other financing the poll who have their different favorite candidates, I don’t see how Georges plays a role in influencing these poll results.
WELL DIDN’T YOU TELL ME THAT THERE ARE RUMORS THAT GEORGES MIGHT ENTER THE RACE?
Yes, I have even heard complaints that since Georges is using his pollster, and has inserted questions in the polls asking the viability of his running for governor, therefore, the poll is not accurate. I believe he is thinking about that possibility just as four years ago, he thought about running for governor, but didn’t. I have also been told that he polls himself on a variety of issues, not just his prospects of winning an election. But, here’s my main point--how would Georges benefit by these results? Edwards is actually looking like he could possibly win, assuming all remains the same as today and everything remained equal. Even if one would assume the accuracy of the Kennedy poll, with Angelle and Vitter essentially tying, how would that help Georges? I can’t see how a strong democrat and a strong republican against a strong US Senator, who is essentially the incumbent in this race, would warrant the need for Georges to jump in. If Edwards was doing poorly and if Angelle or Dardenne were doing middling or also poorly, and if Georges’ numbers indicated that he could beat Vitter, then, I could see him jumping into the contest. To me, that scenario does not exist here. Surely not now. Who knows what happens between now and qualifying but to me, the numbers just don’t make sense for Georges to run, other than the fact that the other candidates don’t have the money really needed to compete with Vitter. I am not saying he could not win for he surely would be able to match Vitter’s spending, but these current poll numbers seem to argue against that scenario.
OK. SO, LET’S TALK ABOUT THE POLL AS IT RELATES TO THE FUTURE. FOR THE SAKE OF DISCUSSION, LET US ASSUME THE KENNEDY POLL IS THE ACCURATE ONE. CAN WE ACTUALLY SEE THE PROSPECTS OF AN EDWARDS-ANGELLE OR EDWARDS-DARDENNE OR EVEN AN ANGELLE-DARDENNE RUNOFF?
We can, but, in my view, don’t count on it. Vitter has the one thing that his three current competitors do not have, the political mother’s milk, money. Over nine million in the bank says a lot. Now, of course, roughly half of that money comes from his SUPER PAC which supposedly is separate and out of the control of Vitter or his campaign. So, let’s assume Kennedy’s numbers are valid, showing Vitter with high negatives and total recognition, therefore, not having much room for growth. Let us also assume that the incumbent is at best, in the low thirties. With his state campaign money, he will present himself positively whether by TV, Internet or mailings. Because of his existing negatives, he cannot risk his campaign running the negative ads against his opponents, thus, you can bet his Super PAC will be bombing away at the candidates. This would likely bring his own numbers down somewhat, but not as much as if his campaign were running the negative commercials. In military terms, think of Dresden. You can bet the assault of attack ads will raise his opponents’ negative numbers. We will see if the Vitter campaign or his Super Pac takes the Kennedy poll seriously, based upon their TV ad responses, if any.
So, personally, even assuming the Angelle marked improvement and further assuming that Dardenne has not even gotten going yet with his campaign, and has a couple of million to spend to get his own message out, Vitter’s Super PAC has over four million dollars to remind anti-Common Core voters that Dardenne favors it. He has plenty of money to define Angelle and to pin Jindal’s tail upon Angelle’s brow. You can bet the negative ads will be steady, often and powerful. Even with Vitter at low 30%, I’d rather be in Vitter’s cash situation than in any of the other candidates’s spot. Polls are great right now because of the inherent controversies. This one by Kennedy should help Angelle rev up the campaign money flow. But, when one candidate has a roughly 3 to 1 advantage over the rest of his opposition combined, what happens today is not necessarily even close to what might occur tomorrow.
I believe that Vitter is not only in the catbird’s seat, he is ready to devour the bird. We shall just wait and see.
THANK YOU, THOSE WERE INTERESTING AND GOOD RESPONSES.
Well, thank you. Gee, almost as good as your questions.
Talk about the article below