More On The Latest SurveyUSA Mayoral Poll

Everyone is just cold foaming at the mouth about the latest mayoral poll from SurveyUSA. The newest data give Greg Fischer a 48% to 42% lead over Hal Heiner – up from 47% to 45% three weeks prior. And A Kentucky Newspaper, along with the Fischer campaign, is spinning this as tremendous momentum for the Democrat.

But here’s reality: if weighting of Democrats and Republicans hadn’t been adjusted by SurveyUSA, Heiner would probably be in the lead 46% to 44%. In the poll three weeks ago, turnout was predicted at 51% Democrats and 40% Republicans. The latest poll predicts turnout will be 57% Democrats and 34% Republicans. Quite a difference that can’t be ignored. Even WHAS11 failed to mention the turnout prediction change (I didn’t realize it at the time, as crosstabs hadn’t yet been released.)

The newspaper’s assertion that a change in partisanship weighting by the pollster is “momentum” is incompetent journalism at best and an intentional misrepresentation at worst.

The real momentum seems to lie with Heiner. He’s gained among Democrats and has taken a ten-point swing among Independents during the past three weeks.

(Heiner’s still gonna lose because he’s becoming increasingly afraid of the press and his campaign staff are dead set on allowing ignorance to prevail.)

26 thoughts on “More On The Latest SurveyUSA Mayoral Poll

  1. Heiner ought to call a MAJOR press conference and challenge the CJ on this. They’ve done this kind of crap for decades and no one EVER challenges them. For example, just 3 weeks ago they had an ONLINE/LIVE STREAMING Editorial Board interview w/Berman, then the NEXT morning their banner headline turned the Editorial Board meeting into a ‘news event’ and the following Sunday there was a triple sized editorial supporting Berman. This STUNT was totally violative of Canon 5 of the American Ass’n of Editors Canons of Ethics — but they did it anyway. The time has come for the CJ to be brought to task for its mind-numbing effort to control the public debate around here.

  2. They’re not trying to make it look good for Greg.

    SUSA adjusted the weight, like with the U.S. Senate poll, and none of the media outlets reported such as the reason for Fischer’s gain.

  3. None of the media outlets reported the weighting of the poll as the reason for Fischer’s gain because that would be incorrect.

    The SurveyUSA poll is not “weighted” and so obviously there is no adjusting.

    I talked to the pollster before doing my story to make sure I understood how this works.

    In a nutshell:

    They do not ask how the responder is registered to vote. They simply ask if the person considers themselves a Democrat or Republican.

    In this way, they are able to take a more realistic snapshot in time of how people view themselves, and how they will vote.

    In other words, they want to know how people are thinking and feeling on the day they talk to them…not what they told the county clerk 12 months ago, or whatever.

    There were 900 Jefferson County adults interviewed in the poll. Of them 596 were determined to be likely voters. All 596 were calculated in the poll results without regard to whether they identified themselves as D or R.

    and PS: The “Fischer” headline was a simple mistake made by an editor. And it wasn’t up for 3 hours. It was up for about 20 minutes before it was corrected.

    There is no conspiracy here. And unlike the blogs, we correct our mistakes.

    Dan Klepal
    Metro government reporter
    The Courier-Journal.

  4. The Highlander has no idea what he’s talking about. The American Society of Newspaper Editors Code of Ethics (which I assume is what he is talking about) says in its fifth article: “To be impartial does not require the press to be unquestioning or to
    refrain from editorial expression. Sound practice, however, demands a clear distinction for the reader between news reports and opinion. Articles that contain opinion or personal interpretation should be clearly identified.”
    What The Highlander appears to be arguing is that we shouldn’t report on things upon which our editorial page editorializes. That’s ludicrous.
    We occasionally write about what goes on in editorial board interviews on the occasion that a reporter is present or when the interview is recorded.
    Other newspapers do this as well. You may recall the 2008 Hillary Clinton editorial baord interview with a South Dakota newspaper, in which she was forced to appologize after defending her decision to stay in the race by saying, “My husband did not wrap up the nomination in 1992 until he won the California primary somewhere in the middle of June, right? We all remember Bobby Kennedy was assassinated in June in California. You know I just, I don’t understand it.”
    That newspaper wrote about that interview in its news pages as well.

  5. Klepal: Yeah, I never correct mistakes or offer updates, ever.

    Neither your paper – nor WHAS11 – reported a difference in the number of Democrats and Republicans responding. That’s kind of a big deal. And I’m not the only person who thinks so.

    “Weighted” is just a word I chose to use and I’m pretty sure most readers comprehend that.

    it’s ridiculous to suggest SurveyUSA has no control over how many Democrats and Republicans are surveyed. And if they wanted an accurate representation of expected turnout, they’d keep calling to get the numbers.

    The Fischer headline was up for nearly three hours. I waited, out of courtesy, before saying anything. I started receiving screenshots right after the story went live. The particular screenshot I clipped was received at 12:43 P.M., more than two hours before that story went live and just about three hours after I started getting the screenshots.

    See for yourself:


    Personally identifiable information has been redacted.

    P.S. Glad to know A Kentucky Newspaper’s newsroom spends 24/7 refreshing this site.

  6. Well aren’t we all of a sudden on the defense CJers.
    At least in the reporting of JCPS, there is no separation between what is reported and the support of SB by the editorial board. When important issues, such as the unanimous vote to review the student assignment plan, are not reported, it does a disservice to the community. I feel sorry for Ms. Konz.

  7. Jake, you of all people certainly know that “weighted” is not just a word you choose when talking about polling. It has a very specific meaning — increasing or decreasing the value of polling data based on preset demographic ideals.

    The problem with that is if you use only historical norms to determine what the R and D turnout is, you ignore what is going on in reality — in this case, pollsters nationally are seeing what is called an “enthusiasm gap” since Ds seem to be on the defensive. In a situation like the current political environment, I’m not sure anybody has a good model about who will turn out.

    That said, I think allowing the respondents to tell us if they are going to go vote may be the best way to determine turnout.

    And weren’t you just defending SUAS recently as the #2 most accurate pollster according to Nate Silver. The polls methodology has not changed.


  8. “Weighted” is a word I chose to use as it most closely represented what I meant – the fact that the number of Dems & Repubs had changed from the previous poll.

    I have repeatedly written about the enthusiasm gap and the reality that pollsters and campaigns need to account for the increased teabagger turnout this year.

    It’s tough to predict turnout, but historical data is really all we have to compare to. You can adjust from there based on a number of factors if you choose to.

    I routinely say SUSA is the most accurate poll in Kentucky. And it is.

    But you can’t ignore the change in Dem & Repubs.

  9. One more response and I’m getting back to work. The poll that came out at the beginning of September that showed a smaller Dem-Rep gap came at a time when there was almost daily news about the economy that made you want to groan and Conway hadn’t gone on the offensive yet against Paul (or he at least didn’t have the ad with Paul saying he opposed laws banning anything but violent behavior) It’s not hard to believe that as the economy seemed to stabilize a little and if the Conway ads were working, Dems might be a little more enthused than they were before.

    That, combined with normal shifts within the margin of error could very well explain the changing demograpic in the poll.

    That’s it. I’m getting back to work.

  10. Messrs Gerth and Klepal are TOTALLY incorrect. It is an established FACT that it is UNETHICAL for a newspaper to CREATE the news in an editorial (which it did) the report the editorial as NEWS (which it did) then editorialize supporting the news THAT THE EDITORIAL BOARD CREATED. It is mind boggling what they did. Furthermore, just out of my own personal interest, I called several noted journalism/ethics personages and to a one they advised me that what the CJ did with the Berman story was ‘highly unethical’ and one even added that it was irresponsible.
    So Gerth and Klepal can say what they want and defend their bosses to the hilt — but the CJ isn’t going to win any more Pulitizers with that kind of journalism.

  11. For purposes of disclosure, I am a field advisor to the Fischer campaign, so everything which follows may be tainted by that relationship. But this is not a repsonse from the campaign. These are simply my personal observations only.

    I refer Jacob’s readers back to a headline of his from a few days back entitled “Jeff Noble is right about those numbers.” Here is the link:

    In that entry Jacob commented on a post on my blog ( explaining the numbers used in a SUSA poll for Congressman John Yarmuth, numbers which I believed overstated expected Republican participation and understated expected Democratic participation in the 2010 General. Stuart Perelmuter and I worked out the numbers for 3rd District participation going back to 2000 and that was the basis for my post.

    I’m glad to see SUSA adjust its numbers for this latest poll for the Fischer campaign. It is a move in the right direction and represents, using this current adjustment, a reflection of what might happen if the Democrats underparticipate to some extent from their usual numbers and the Republicans overparticipate to some extent from their usual numbers. In short, Jacob is right in his comment above at 2:36 pm. You can’t ignore the change in Democrats and Republicans participating. This poll covers the much ballyhooed “enthusiasm gap” as it has departed from the historical norms for Jefferson County in a manner which is advantageous to the Republican candidate for mayor.

  12. Regardless of the methodology used or weighting or any other assortment of pollster cliches, I took the point of the post to be that the reporting of the results was misleading. And it was. You can’t honestly say Fischer has momentum when one week you take the opinion of 5 Ds, 4 Rs, and 1 I and the next week you take the opinion of 6 Ds, 3 Rs, and 1 I. Basic math indicates the D is going to have a bigger lead. That’s not momentum; that’s counting differently.

    Perhaps the newspaper guys will ask Stuart Rothenberg (noted political commentator), when he’s in town a week from tonight to speak at the GLI Legislative Dinner, what he meant when he said on his blog Monday in reference to the Senate poll (although the same principle applies to the Mayoral poll):

    “What isn’t fine is for a media organization – whether a newspaper, a TV station or a TV network – to act as if it’s impossible that its poll numbers are wrong.

    At the very least, the Courier-Journal should have acknowledged that possibility and noted that the dramatic change in the Senate head-to-head poll question was difficult to believe.”

    If the numbers 3 weeks ago were wrong, acknowledge the fact and attribute the change to that. Anything else does a disservice to the public.

  13. Curious: You are EXACTLY CORRECT. Disregard the flacks and the cj reporters. There is no justification for this kind of reporting by a reputable newspaper.

  14. Two quick points:

    No one here acts as if the polls can’t be wrong. I don’t know if they’re right or wrong. But they are news.

    What I do know is that there were 11 candidates in the mayoral primary, and the last SUSA poll got them all within the margin of error except for David Tandy. That’s a pretty good track record.

    Secondly, Jake is trying to say Fischers 1-point gain in our third poll from our second is becasue six percent fewer Republicans and six percent more Democrats were in the sample.

    What he’s not telling you is that Fischer gained 2 points in our second poll from our first. The sampling difference there? Republicans up 5 percent; Democrats down three percent.

    So, Fischer gained ground even when the sampling should have favored Heiner.

    And I will just repeat, the pollsters do not randomly call people until they get a predetermined percentages for D and R. They call people randomly until they get a large enough sampling to make the poll valid, and then the D-R breakdown is whatever it is.

    That’s why you see differences in those percentages from poll to poll.

    You can agree or disagree with the method, but it is what it is. And as Jake has said in the past, it seems to work pretty well based on actual election results.


    Dan Klepal
    Metro government reporter
    The Courier-Journal

  15. Klepal: You may not act like polls can be wrong – but your inaction and lack of discussion about WHY the numbers changed tells an interesting story. You can think it’s no big deal, but perception matters when you work for an organization that professes to have journalistic integrity. To suggest otherwise is silly. I’m not the only individual making that point. Other publications are doing the same.

    I can count the hundreds of times I’ve give you folks shit and you’ve never reacted. But three reporters lose their mind when we talk about a poll that you admit can be faulty. Why is that? Because of money the dying Gannett stack had to spend? Because Rothenberg is also critical?

    Your reaction is hysterical – just like the last time you melted down in the comments.

    Readers: What A Kentucky Newsapper’s reporters aren’t telling you is that privately they complain about their bosses, the editorial board and Gannett. But they put on a completely different show when they’re at work and getting paid.

    Most of them have stories – particularly those on the local front – about editors not allowing them to include information in news stories because it could negatively impact a supported candidate or cause discomfort on an advertising level.

    And privately, one of the three reporters commenting here, called me last night to complain about editorial influence in the news room.

    Meanwhile, it’s all out damage control on the SurveyUSA front.

  16. Jake,
    I don’t believe you. I think you are lying. If it was I who called you last night to complain about editorial influence, you have my full permission to disclose my name.

    Dan Klepal said he also gives you permission to release his name. And Deborah Yetter also gives you permission to use her name.

  17. Really? One of you is lying.

    But I appreciate you taking the time to comment from Gannett – making me money – while you’re at work.

    Jonathan Hurst tells me you have made similar comments to him. Was he, too, lying when he told me that?

  18. It is quite interesting to see how the folks from the Once Great Newspaper are circling the wagons on this issue.

    I wonder why…?

  19. I have absolutely no idea what you’re talking about.

    But now is time to come clean. You have our permission. Tell us who it was. Provide us with a screen grab from your iPhone to prove to us you took that phone call.

  20. Interesting. Hurst said otherwise.

    Call didn’t come to my personal number, it rolled into my office on our Vonage account. But here you go:


  21. Jake: This is terrific. I, too, wonder why the onslaught of cj reporters blogging in here. Seems as though to me that you’ve hit them in a ‘real sore spot.’
    I have another good question of Klepal and Gerth who seem to be journalistic ethicists. Do your OWN calling around the country and ask the journalistic ethicists about my INITIAL point on this thread. I believe you’ll find the 2 our of 3 will tell you (minimally) that the FAR better practice of a n’paper is to NOT have an editorial board meeting live streamed, then report it as NEWS in a banner – across the top – headline the next day and THEN ON SUNDAY’s editorial page present a 2-3,000 word editorial supporting your position and QUOTING FROM THE NEWS STORY YOU CREATED. Maximally, they will say it is UNETHICAL AND HIGHLY IRRESPONSIBLE JOURNALISM. Welcome to the cj’s effort to control the thought of the Lemings of Louisville.

  22. While we are on the topic of the CJ editorial board influencing the news division how about the recent 32 point headline warning of the Kennedy Bridges structural defecient rating. The truth is a 75 (up from 73 a few years ago) rating is merely indicating that a bridge is due for maintenance and is nowhere near the 40 required for a bridge to be quickly replaced. I understand that your goal is to sell papers and a scary headline works in that regard but the people at the CJ must understand that the average reader skims the headlines. The average reader would come away thinking that the Kennedy bridge desperately needs to be replaced and we must immediately build the downtown ORBP. The truth is that the Kennedy bridge is an integral part of the terribly designed downtown ORBP and is not scheduled to receive an overhaul as part of the project. The CJs coverage of the Bridges issue has been terrible biased from the beginning but I must admit that it has improved slightly (terrible to mediocre). Now if the CJ has any integrity they will write an article about the traffic projections when this project was approved in 2003 and the realized traffic today,

    Projected trips in 2008: approx. 145,000 Actual Kennedy bridge trips: approx 120,000

    Ask any highway engineering expert and they will tell you that the ORBP record of decision predicted traffic volumes are impossible. By the time the ORBP is completed in 2024 the ROD assumes traffic volumes of nearly 180,000.

  23. That prediction of 180,000 I-65 bridge trips includes an east end bridge that would handle much of the cross river traffic. The predicted traffic volumes defy logic.

  24. Thanks Stunoland for this information. The CJ has become a travesty. I”m not sure that they haven’t deliberately endeavored to keep this community ‘dummed-down’ by their DEMAND that a pitiful public education program be continued — utter stupidity of a populace allows something like the CJ to be more effective in moving the masses to their point of view. This is particularly egregious when ‘their point of view’ is DEMONSTRATED on the news page. Yes — it’s a travesty.

  25. Stunoland articulation of what is wrong with the C-J’s coverage of the bridges matter couldn’t be more spot on.

    What I don’t get is how the C-J editorial board doesn’t realize how obvious slants like this leave the impression with almost anyone that nothing they report can be taken seriously any longer.

    The C-J may as well be called now.

Comments are closed.